OFM 2009 Forecast Analysis Fundamentals H A5

OilField Manager Forecast Analysis Fundamentals

Workflow/Solutions Training Version 2009

Schlumberger Information Solutions June 4, 2010

Copyright Notice © 2010 Schlumberger. All rights reserved. No part of this manual may be reproduced, stored in a retrieval system, or translated in any form or by any means, electronic or mechanical, including photocopying and recording, without the prior written permission of Schlumberger Information Solutions, 5599 San Felipe, Suite100, Houston, TX 77056-2722.

Disclaimer Use of this product is governed by the License Agreement. Schlumberger makes no warranties, express, implied, or statutory, with respect to the product described herein and disclaims without limitation any warranties of merchantability or fitness for a particular purpose. Schlumberger reserves the right to revise the information in this manual at any time without notice.

Trademark Information Software application marks, unless otherwise indicated, used in this publication are trademarks of Schlumberger. Certain other products and product names are trademarks or registered trademarks of their respective companies or organizations.

Table of Contents

About this Manual Forecasting Using OFM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Forecasting Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . User Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ratio Forecasts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Learning Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . What You Will Need . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . What to Expect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Course Conventions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Icons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Provo Project Workflow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 1 2 2 2 2 3 4 5 6 7

Module 1: Brief Review of Decline Curve Analysis Theory Learning Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Lesson 1: Principles of Decline Curve Analysis . . . . . . . . . . . . . . . . . . . . . . . . . 9 Arps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Curve Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 Scalability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Assumption . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Forecasting Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

Module 2: Simple OFM Forecasting Prerequisites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Learning Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lesson 1: Forecast Set Up . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Identifying Completions of Interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Setting Up Forecast Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Setting Up a Forecast Association . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lesson 2: Forecast Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Historical Regression (History Match) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Setting Upper and Lower Limits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Excluding Data Points with Range Limit . . . . . . . . . . . . . . . . . . . . . . . . . . . Toggling Working Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reset . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

OilField Manager Forecast Analysis Fundamentals Workflow/Solutions Training, Version 2009

17 17 18 18 19 21 23 24 25 29 31 32

i

Lesson 3: History Match Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fit Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . b Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Manual Fit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Future Performance Forecasting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Viewing Historical Regression Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . Viewing Forecast Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Single Entity (Individual) Forecast Settings . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 1: Saving the Forecast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Recalling a Saved Forecast or Working Forecast . . . . . . . . . . . . . . . . . . . . Exercise 2: Editing the Forecast Graph . . . . . . . . . . . . . . . . . . . . . . . . . . . . Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

32 32 33 35 36 36 37 39 43 45 46 53 53

Module 3: Additional Forecasting Features Learning Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lesson 1: Phases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 1: Moving Between Gas and Water Phases . . . . . . . . . . . . . . . . . Exercise 2: Forecasting with User-Defined Phases . . . . . . . . . . . . . . . . . . . Lesson 2: Additional Scenario Settings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 1: Configuring a Two-Schedule Forecast . . . . . . . . . . . . . . . . . . . Control the Start and End Times . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reserve Type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Adding Default Entries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Decline Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 2: Creating a Forecast with User-Defined Parameters . . . . . . . . . Lesson 3: Re-Initializing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 1: Simulating the Effect of Rate Acceleration Techniques . . . . . . Lesson 4: Additional History Matching Data Reduction Features . . . . . . . . . . . Exercise 1: Digitizing Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 2: Using a Calculated Variable to Exclude Data Points . . . . . . . . . Exercise 3: Task 1: Forecasting Oil for High-Volume Producers . . . . . . . . Lesson 5: Group Forecasting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 1: Task 2: Forecasting Group Oil . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 2: Task 3: Forecasting Group Gas . . . . . . . . . . . . . . . . . . . . . . . . Forecasting Categories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 3: Task 4: Forecasting the Performance of the VK 1 Pool . . . . . . Lesson 6: Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 1: Setting Up a Default Case . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 2: Adding New Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 3: Changing the Current Case . . . . . . . . . . . . . . . . . . . . . . . . . . . Compare Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ii

55 55 56 59 62 62 64 66 66 67 67 70 70 72 72 75 77 77 78 81 82 83 84 86 87 88 89


Creating New Futures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 Comparing Cases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 Exercise 4: Task 5: Forecasting Multiple Cases . . . . . . . . . . . . . . . . . . . . . 93 Exercise 5: Task 6 Comparing all Three Forecasts Graphically . . . . . . . . . 94 Extract Forecast Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 Exercise 6: Task 7: Comparing Group and Sum Forecasts . . . . . . . . . . . . 95 Lesson 7: Automated Forecast Calculated Variables . . . . . . . . . . . . . . . . . . . 100 Using the Automated Calculated Variables Feature . . . . . . . . . . . . . . . . . 101 Saving New Gas Forecast Calculated Variables . . . . . . . . . . . . . . . . . . . . 103 Lesson 8: Automated Forecast Reports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 Using the Automated Forecast Report Feature . . . . . . . . . . . . . . . . . . . . . 103 Forecast Parameters Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Forecast Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 106

Module 4: Ratio Forecasts Learning Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lesson 1: Fluid Cut Forecast Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 1: Creating a Water Cut Forecast . . . . . . . . . . . . . . . . . . . . . . . . Other Fluid Cuts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Accessing Automated Forecast Calculated Variables for Ratio Analysis . Accessing Automated Forecast Reports from Ratio Analysis . . . . . . . . . . Lesson 2: Calculated Forecast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 1: Creating a Calculated Forecast . . . . . . . . . . . . . . . . . . . . . . . . Lesson 3: P/Z Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Setting the Z Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Performing a P/Z Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise 1: Task 8: Forecasting P/Z for Single Compartment . . . . . . . . . . Review Questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .


109 109 109 112 112 114 115 116 119 121 123 125 127 127

iii

iv


Schlumberger

About this Manual

About this Manual Welcome to OilField Manager (OFM) Forecast Analysis Fundamentals. In this course, you will learn how to use the full range of capabilities in OFM to forecast production and retrieve forecast information for further studies.

Forecasting Using OFM The Forecast module in OFM has undergone continuous enhancements, both to improve the usability of the application and to facilitate more sophisticated analyses. The Forecast module offers four major forecasting techniques: Empirical, Fetkovitch, Locke & Sawyer, and Analytical Transient solutions. The last three of these are type curve techniques, and while they bring additional understanding of the well’s current and future performance, in practice they are used very infrequently, and so they will be covered in the Advanced OFM training course. The majority of forecast analysis is done using simple Empirical techniques: the physics are well understood, and it is often the preferred method by which organizations estimate their official reserves. For this reason, we will focus on the Empirical methods for our analyses.

Forecasting Scope Forecasts in OFM can be done at the single well or completion level. Equally, they may be performed at any category level (for example, reservoir and fault block), or on a group of completions. The group in this instance is a single entity. If you compare the group entity forecast with the sum of the forecasts of the individual completions which make up the group entity, it is no surprise that you should expect a different answer. This issue (‘forecast of a group’ vs. ‘group of the forecasts’) will be dealt with in the course.


1

Schlumberger

User Control While OFM fully automates the process of history matching and predicting well performance, you have complete autonomy over the degree to which that process takes place. You can influence every aspect of a forecast session. The majority of your work will be based on liquid (oil) forecasting techniques, which are equally suited to oil or gas well analysis. However, OFM has several methods which are unique to gas. The Well Deliverability analysis enables you to analyze data from Flow-After-Flow tests, Isochronal tests, and Modified Isochronal tests. You will also explore the classical P/Z approach to gas forecasting.

Ratio Forecasts In addition to single phase forecasts, you will explore the ability of OFM to work with ratio forecasts, where cumulative phase production is the basis of interpretation, rather than time. By doing this, you can provide a whole new layer of predictive capability.

Learning Objectives After completing this course, you will know how to: •

set up a forecast scenario

•

use forecast features to create forecasts

•

work with the results of forecasts.

What You Will Need In this workflow, you will need the following hardware and applications:

2

•

OFM properly installed and licensed

•

Training datasets


Schlumberger

What to Expect In each module within this training material, you will encounter the following: •

Overview of the module

•

Prerequisites to the module (if necessary)

•

Learning objectives

•

A workflow component (if applicable)

•

Lessons, which explain about a subject or an activity in the workflow

•

Procedures, which show the sequence of steps needed to perform a task

•

Exercises, which allow you to practice a task by using the steps in the procedure with a data set

•

Scenario-based exercises

•

Questions about the module

•

Summary of the module

You will also encounter notes, tips and best practices.


3

Schlumberger

Course Conventions Characters typed in Bold

Reference to dialog names and application areas or commands to be performed. For example, ‘Open the Open Asset Model dialog.’ or ‘Choose Components.’ Denotes keyboard commands. For example, ‘Type a name and press Enter.’ Identifies the name of Schlumberger software applications, such as ECLIPSE, GeoFrame or Petrel.

Characters inside <> triangle brackets

Indicate variable values that the user must supply, such as and .

Characters typed in italics

Represent file names or directories, such as ‘... edit the file sample.dat and...’ Represent lists and option areas in a window, such as Attributes list or Experiments area. Identifies the first use of important terms or concepts. For example, ‘compositional simulation…’ or ‘safe mode operation.’

Characters typed in fixed-width

Represent code, data, and other literal text the user sees or types. For example, username/password or 0.7323.

NOTE: Some of the conventions used in this manual indicate the information to enter, but are not part of the information. For example, quotation marks and information between brackets indicate the information you should enter. Do not include the quotation marks or brackets when you type your information. Instructions to make menu selections are also written using bold text and an arrow indicating the selection sequence, as shown: 1. Click File menu > Save (the Save Asset Model File dialog box opens.) OR Click the Save Model

toolbar button.

An ‘OR’ is used to identify an alternate procedure.

4


Schlumberger

Icons Throughout this manual, you will find icons in the margin representing various kinds of information. These icons serve as at-a-glance reminders of their associated text. See below for descriptions of what each icon means.


5

Schlumberger

The Provo Project Workflow This manual follows on from the work completed in the Introduction to OilField Manager Fundamentals course. In that course, an OFM project, named Provo, was built and a number of simple engineering workflows were completed which allowed you, as a member of the Provo team, to quickly and effectively assess the performance of the wells. In this new course manual, we will continue the process. You have been assigned the following additional responsibilities, all of which will require an understanding of decline curve analysis theory and OFM forecasting techniques: 1. Forecast the future performance of the following completions: • Oil rate for all oil completions which historically have produced more than 40,000 bbl. • Gas rate for all gas completions which have produced more than 25 MMcf. 2. Group the oil completions in step 1, and create a single oil rate forecast for the group. 3. Group the gas completions in step 1, and create a single gas rate forecast for the group. 4. Forecast the performance of the VK 1 pool. 5. Prepare two new oil forecasts for completion HOGL2_1129, assuming: • a short workover on this completion in June, 2010 will be successful, and that production will return to its historical decline curve • the work will be delayed a further 6 months. 6. Graphically compare all three forecasts (steps 1 and 5). 7. For the oil completions, compare the group oil rate forecast with the sum of the forecasts of the individual completions. 8. Identify a collection of completions in a single compartment and create a P/Z forecast for the group. As in the OFM introductory course, all of these activities and more are possible. As you progress through the material, you will complete each of the above tasks.

6


Schlumberger

Summary In this introduction, we have •

defined the learning objectives

•

outlined what tools you will need for this training

•

discussed course conventions that you will encounter within this material

•

introduced the project-specific workflow you will complete.


7

Schlumberger

NOTES

8


Schlumberger

Brief Review of Decline Curve Analysis Theory

Module 1 Brief Review of Decline Curve Analysis Theory In preparation for the work you will do in OFM, a refresher of the underlying principles behind Decline Curve Analysis may be helpful.

Learning Objectives After completing this module, you will know about: •

Exponential decline

•

General Hyperbolic decline

•

Harmonic decline.

Lesson 1

Principles of Decline Curve Analysis

Decline Curve Analysis techniques are used to estimate ultimate recovery and forecast deliverability rates by evaluating historical production and projecting production to an economic limit. These techniques graph production data to reveal performance trends, model those trends with an equation, and use the equation to project the trend to an economic limit. The biggest advantage of Decline Curve Analysis is that it is independent of the size and shape of the reservoir or the actual drive-mechanism, which avoids the need for detailed reservoir or production data. The only data requirement for Decline Curve Analysis and extrapolation is production data, which is relatively easy to obtain for a large number of fields.

Arps Arps (1945) created the foundation of Decline Curve Analysis by proposing simple mathematical curves (exponential, harmonic and hyperbolic) as a tool for creating a reasonable outlook for the production of an oil well when it has reached the onset of decline.


9


Schlumberger

His original approach has been modified over time, but is still used as a benchmark for industry for analysis and interpretation of production data as a result of its simplicity.

Exponential Decline Exponential decline for flow rate q t is defined as follows:

where qt = Instantaneous production rate at the end of the month qi = Instantaneous initial production rate D = Nominal monthly decline rate t = Time in months The above exponential equation assumes a constant decline rate. It is made linear by taking the natural log of the equation, which produces the classic straight line on a semi-log plot.

General Hyperbolic Decline Hyperbolic decline for flow rate q t is defined as follows:

where qt = Instantaneous production rate at the end of the month qi = Instantaneous initial production rate b = Hyperbolic b value Di = Initial nominal monthly decline t = Time in months from time zero A hyperbolic decline is not a constant percentage decline. The decline rate decreases during the life of a well, extending its life compared to an exponential decline. The rate at which the decline decreases depends on the hyperbolic b value.

10


Schlumberger


Harmonic Decline Harmonic decline for flow rate qt is defined as follows:

where qt = Instantaneous production rate at the end of the month qi = Instantaneous initial production rate Di = Initial nominal monthly decline rate t = Time in months from time zero Note that the exponential and harmonic curves are only special cases of general hyperbolic decline (b = 0 and b = 1, respectively). Modifying the shape parameter alters the shape of the production rate function, and this can be used to determine what kind of decline curve is suitable for fitting against empirical data. The value of the decline parameter governs how steep the decrease in production will be. The exponential decline curve is by far the most convenient method to work with, and it agrees well with actual data. Hyperbolic and harmonic decline curves involve more complicated functions and are less practical. The disadvantage of the exponential decline curve is that it sometimes tends to underestimate production far out in the tail part of the production curve, as decline often flattens out towards a more harmonic and hyperbolic behavior in that region.


11


Schlumberger

Curve Identification By varying the format of the presentation of the data, each of the curves can take on a straight line shape, making the process of matching and forecasting easier to accomplish as shown in the figures.

12

Figure 1

Log rate versus time

Figure 2

Rate versus time


Schlumberger


Figure 3

Rate vs. cum

Figure 4

Log rate vs. cum


13


Schlumberger

Scalability Decline curves of various forms can be used to create either reasonable outlooks for fluid production of a single well or an entire field. However, bear in mind that, in many field cases, a single curve is not sufficient to obtain a good fit and it may be necessary to use a combination of curves to obtain good agreement. The importance of individual wells diminishes as the total number of studied wells becomes large, and generalized well behavior can be identified. In such cases, a simple decline curve can be successfully used to forecast total production from a large set of wells, as under- and overestimations for individual wells cancel out each other in the long run. Decline rate analysis and decline curves can be a convenient tool for identifying long-term trends and projecting reasonable production behavior into the future.

Assumption It is worth pointing out another underlying principle: Decline Analysis is based on the implicit assumption that the factors which governed performance (for example, of a well or field) in the past will continue to govern in the foreseeable future. Anything which disrupts or violates that assumption, such as a workover at some point in the near future, nullifies the projection or forecast.

Forecasting Limits The simplest decline curves are characterized by three parameters - the initial production rate q, the decline rate D, and the shape parameter b. If production is allowed to continue, you can use these parameters to calculate the ultimate cumulative production of the decline phase. You can then calculate the ultimate recovery of the well or field. However, production is stopped when some form of technical or economic limit is reached, such as when keeping the equipment running requires more money and/or energy than it yields. This cut-off point can be introduced into the calculation to derive a technical or economic recovery estimate.

14


Schlumberger


Review Questions •

Is the Decline Curve Analysis independent of the size and shape of the reservoir or the actual drive-mechanism?

•

The value of the decline parameter does not govern how steep the decrease in production will be. True or False?

•

Decline Analysis is based on the implicit assumption that the factors which governed performance in the past will continue to govern in the foreseeable future, such as a well or field. True or False?

Summary In this module, you learned about: •

Exponential decline

•

General Hyperbolic decline

•

Harmonic decline.


15


Schlumberger

NOTES

16


Schlumberger

Simple OFM Forecasting

Module 2 Simple OFM Forecasting Here, you work through a typical forecast analysis using OFM. Rather than try to cover every possible permutation, you explore only the more common features of OFM. A more comprehensive discussion is included later in the training.

Prerequisites To successfully complete this training, you must have knowledge of these aspects of OFM: •

Creating filters

•

Grouping completions

•

Using categories

•

System functions

•

Calculated variables and calculated fields

•

OFM plots and reports

In addition, you should have completed the OFM Fundamentals training, as this course builds upon the results obtained in the OFM Fundamentals training. If the OFM Forecast Analysis Fundamentals training was not immediately preceded by the OFM Fundamentals training, you can retrieve a clean copy of the completed project at: Module 3 Project Creation\Lesson 2 Flat ASCII Text Files\The Completed Provo Project\Provo.ofm.

Learning Objectives After completing this module, you will know how to: •

set up a forecast scenario

•

edit a graph

•

work with forecast features

•

change the fit type

•

work with multi-schedule forecasts

•

perform a manual match.


17


Schlumberger

Lesson 1

Forecast Set Up

A specific analysis requires preparatory work before beginning the process. In addition to selecting the completions to work on, you must ensure that default settings for your subsequent analyses are effective and save you as much time as possible

Identifying Completions of Interest One of the first tasks in performing forecast activity in OFM is assessing which completions to work with. Bear in mind that you will perform a group forecast at some stage, so now would be a good time to identify which completions will be grouped. You will restrict yourself to a short list of the best oil producers in the Provo project, which is a list of producers that have made more than 40,000 bbl of oil. To identify completions of interest: 1. Start OFM and open the Provo project. A 'clean' copy of the completed project may be found at: Module 3 Project Creation\Lesson 2 Flat ASCII Text Files\The Completed Provo Project\Provo.ofm. 2. From the Filter pane, create an OFM Query filter with the expression @last(CumOil)>40000.

3. Name the query Best Oil.

18


Schlumberger


4. Activate the filter by checking the box next to the new name.

As the Filter pane indicates, there are 8 completions which meet the criterion you set. This will be your working dataset for the next few exercises.

Setting Up Forecast Variables Before OFM can perform a forecast, you must prepare the variables you will need. You will forecast oil rates, so you will be asked to provide two variables - daily oil rate and cumulative oil. TIP: The second variable is needed if you wish to switch to a rate cum analysis.

Daily Oil Rate OFM requires a source of daily oil rate data. In the Provo project, you do not have raw daily data, so you will create similar data using a calculated equivalent. Note that OFM performs its forecast calculations on a monthly basis, so a single value for the oil rate will be required for each month. This requirement also allows you to calculate a daily equivalent from monthly data.


19


Schlumberger

There are two ways you can define the daily equivalent: •

A calendar day average rate that is equal to the volume produced divided by the number of calendar days in the month.

•

A producing day average rate that is the volume produced divided by the number of days on production).

The preferred method is the producing day average rate. Because of the way in which a group average is calculated, this must be accomplished using a calculated field. In this instance, the expression for the calculated field is IIF(HOURS <> 0, OIL*24/HOURS, NULL) WARNING: Remember to set the Grouping option to Sum when creating this variable. For reference, the corresponding calendar day average would be a calculated variable with the definition: PRD.OIL / @dom(Date) TIP: 'Days on production' data may not always be available, but it is always possible to calculate a calendar day average.

Cumulative Oil You should already have a calculated variable in your Provo project for this variable. If not, then you must create one. To create a calculated variable: 1. Select Database > Calculated Variables. 2. Click Add Variable to add a new variable. 3. Enter the expression @cuminput(PRD.OIL). 4. Name this variable CumOil. NOTE: On the Units tab, set Output Multiplier = M. 5. On the Plot tab, change the plot name to Cumulative Oil.

20


Schlumberger


6. Click OK to save changes. When forecasting oil, the cumulative variable must be set to provide volumes in thousands of bbl or thousands of m3 if metric. Setting M is valid if the project Unit Multiplier Style is set to Field; if the style is Metric, you should set K. The style is set by selecting Tools > Settings > Units. When forecasting gas, the multiplier must be set to provide volumes in millions cf or m3. The field multiplier style setting in this case is MM; the metric multiplier style setting is M.

Setting Up a Forecast Association You begin the first phase of the forecasting process by selecting the variables to use in your forecast session. This is done interactively in an OFM forecast dialog. To set up a forecast association: 1. Select Analysis > Forecast or click

on the main toolbar.

2. If this is the first time you have used the forecast module, you are prompted to set up the forecast variables. Click Yes. If you are not prompted, then continue with step 3. 3. Make the selections, as shown in the figure, to set up an oil forecast. NOTE: OFM confirms the M output multiplier (arrow) which you provided. If this is not the same as the multiplier in the label to the left (Cum Oil, Mbbl), exit the dialog and modify the multiplier in either the calculated variable or the calculated field.


21


Schlumberger

4. Click OK to save your changes and exit the dialog.

22


Schlumberger


5. Choose completion HOGL1032. A forecast window displays. Working Forecast Phase Case Name b Di qi ti te Final Rate Cum. Prod. Cum. Date Reserves Reserves Date EUR Forecast Ended By DB Forecast Date Reserve Type

100

50

Parameters : Oil : Case1 :0 : 0.00656545 M.n. : 0.827588 : 01/31/2009 : 01/31/2019 : 0.376444 : 50.3675 M : 12/31/2008 : 2.09151 M : 01/31/2019 : 52.459 M : Time : Not Saved : None

Producing Day Oil Rate

10

5

1

0.5

0.1 198081 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99200001 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19

Lesson 2

Forecast Analysis

Now that the preparatory work is complete, you can begin the actual analysis. You will do that by conducting the following: 1. Determine which region of the historical data can be considered representative for the purposes of predicting future performance. 2. Remove data points from the analysis which do not belong to that representative region.


23


Schlumberger

3. Incorporate the rate and time at which your forecast is to begin. 4. Determine which factors will control the end of the forecast.

Historical Regression (History Match) You should see several things in the window after completing the previous exercise: •

OFM has graphed your selected rate variable (Producing Day Oil Rate) against time on a semi-log basis. This is the default display setting.

•

All historical data is plotted. The data points are represented by red dots, and are connected by light green lines.

•

A straight line fit (match) has automatically been made through all of the historical data points.

•

A forecast has been proposed, based on the automatic match and based on default settings for the forecast.

•

Results of the proposed forecast appear in a legend.

All of these things happen automatically according to default settings, and they can all be controlled. Your first task is to create a match, or fit, to the historical data. Remember that the match will control the forecast, so time spent in this phase will be rewarded with better quality results.

Unwanted Data Points in the History Match Analysis OFM provides several tools for ensuring that your match is made using data points considered representative of the past performance of the completion. By extension, this is considered to be representative of its future performance. For completion HOGL1032, there are several observations you can make that will determine how to proceed:

24

•

The region from 1980 to 1991 exhibits a very clean straight line on this plot, which is indicative of exponential decline. However, it is no longer representative because work was done in this well during the 1990s, which invalidates the earlier data for inclusion in the analysis.

•

Abnormally high data values during 2004 are not representative, and may even be erroneous.


Schlumberger

•


The region from 2000 to 2007, while noisy, is considered to be the most representative for forecasting purposes.

For all these reasons, you need a way of excluding the unrepresentative data from your analysis. You do this using the Limits feature, which has six options. You will explore four of them.

Setting Upper and Lower Limits Upper and lower limits perform similar tasks, but with one acting in the 'opposite' way to the other. Together, they eliminate large regions of data by scribing a partition or dividing line across the plot. Any data point on the 'wrong side' of the line is excluded.

Setting a Lower Limit Line To set a lower limit line: 1. Right-click on the graph and select Limits > Lower. 2. Using the left mouse button, digitize a series of points on the graph. These are shown in the figure with arrows. Your goal is to define a line that represents the lowest values you are willing to accept. In other words, a lower limit line is placed below the data points you wish to preserve, but above the points you wish to eliminate.


25


Schlumberger

3. After clicking the last point, right-click again and select Done. The entire region up to mid-1991 has been excluded, as has the single data point at the end of 2006. OFM automatically re-matched the remaining points and updated the results legend. TIP: When a Lower Limit line is confirmed, OFM removes all data points below the line, but it also eliminates all data points beyond the ends of the line. The figure illustrates a fast way to eliminate every data point before mid2000 and after 2009, as well as the spike at the end of 2006.

26


Schlumberger


Setting an Upper Limit Line This same principle can be applied to the Upper Limit. To set an upper limit line: 1. Right-click on the graph and select Limits > Upper. 2. Using the left mouse button, digitize a series of points on the graph, as shown in the figure, to eliminate from the analysis the spike and possibly erroneous values in 2004.


27


Schlumberger

Your goal is to define a line which represents the highest values you are willing to accept. An upper limit line is placed above the data points we wish to preserve, but below the points you wish to eliminate.

3. After clicking the last point, right-click and select Done. The bad data points have been removed from the analysis. The fit recalculates and the results are updated.

28


Schlumberger


Excluding Data Points with Range Limit This feature works by excluding data points outside a range of either the x-axis (date values), the y-axis (oil rate) values, or both. To set a range limit: 1. Right-click on the graph and select Limits > Range. 2. Set the values shown in the figure and click OK. All data prior to 2001 has been removed. NOTE: Remember that data points are being removed only from the analysis; the data itself is not affected.


29


Schlumberger

3. Using the Zoom tool, focus in on the newly reduced area. Working Forecast Phase Case Name b Di qi ti te Final Rate Cum. Prod. Cum. Date Reserves Reserves Date EUR Forecast Ended By DB Forecast Date Reserve Type


5


1

0.5

0.3 1998

30

99

2000

01

02

03

04

05

06

07

08

09

10

11

12

13

14

15

16


Schlumberger


Toggling Working Points This tool works like an eraser, allowing you to selectively remove individual points by wiping their selection (active status) off the screen. To toggle working points: 1. From the main menu, select Edit > Toggle Working Points.

2. On the graph, click and wipe the cursor across the points you wish to remove. The data point symbol changes as you erase the data point. Use this feature to remove some of the outlier points – points that are abnormally high or low. 3. When you are finished, select Edit > Toggle Working Points to apply the changes. You have been introduced to several limit capabilities, and you are free to use them in any order you wish but you do not have to use all of them. In many instances, a single limit option may achieve everything you need.


31


Schlumberger

Reset There is another feature that allows you to restart the data point removal process. Right-click and select Limits > Reset to restore ALL data points.

Lesson 3

History Match Settings

Up to now, all fitting was accomplished automatically, based on selected data points. In this section, you explore how these activities can be controlled. The settings which control the primary shape of the fit are located on the Properties pane. Take a moment to note the default settings, as shown in Figure 5.

Figure 5

Properties pane

Fit Type The Fit Type is set at Best Fit, which allows OFM to decide if the best statistical fit through the included data points is exponential, hyperbolic or harmonic. In some circumstances, you may wish to allow OFM to determine the fit this way.

32


Schlumberger


For example, it may be that exponential decline is predominant in your own project. For this completion, you may wish to force OFM to find the best exponential fit through the data. To have OFM find the best exponential fit: 1. Choose the setting Fit Type. 2. On the drop-down list that displays, choose Exponential. While you continue to work on this completion, OFM will find the best straight line through the semi-log plot, which is the definition of exponential decline. This can be confirmed by looking at the results legend. The Arps ‘b’ value is zero, confirmation of an exponential decline fit.

Figure 6

Results legend display

Alternatively, if you wish to force OFM to always find the best harmonic fit, choose Harmonic from the Fit Type property list. Again, you can confirm this on the results legend, where the ‘b’ value should be 1.

b Value The Arps ‘b’ value is an important parameter in decline analysis. In OFM, the value is automatically calculated when you conduct history matches. This is visible in the Properties pane, where the History Match b Value Method is set at Optimize.


33


Schlumberger

It may be, however, that you have more detailed knowledge of the performance of the wells in your project. You may, for example, know that the b values typically range from 0.2 to 0.4, or you may be reasonably confident of assuming the b value is 0.3. For both of these situations, you can adapt your OFM forecast. To constrain the value to a range: 1. Set Fit Type to Best Fit. 2. Set Method to Range. A new property element is displayed.

3. Enter the minimum and maximum values, for example 0.2 and 0.4, respectively. OFM will not allow the subsequently calculated value to fall outside this range. To constrain the value to a single number: Set Method = User. A new property element is displayed.

4. Enter the single figure you wish to use, for example 0.3. OFM will find the best fit through your data, subject to a b value of 0.3

34


Schlumberger


Manual Fit When OFM performs a match to your data, it is a statistical fit through a series of x-y data pairs, but no petroleum engineering analysis is occurring. You may decide the statistical fit is good, but that an engineering 'eye' can improve it. OFM allows you to do this by means of the Manual Fit feature. On the graph, notice the control points on the plot. These are the two blue dots at the ends of the line. To initiate Manual Fit: 1. Using the left mouse button, click and drag the upper blue control point and take note of several things: • The upper control point is moveable. • The control point drags the fit line. • The lower control point remains fixed. • The forecast moves with the fit. • The results legend updates continuously.


35


Schlumberger

2. Release the mouse button. Note that the entry for History Match Method on the Properties pane has changed to Manual. WARNING: When Method is set to Manual, it will remain Manual until you change it back to Auto on the Properties pane. Using this feature, you can place the fit anywhere you like on the screen. The forecast and the calculated results will reflect your changes so be prepared to justify your action. Your fit is now a personal judgment; and is no longer supported by the calculations. NOTE: In this example, there are two control points because two points are sufficient to define straight line exponential decline. If your existing fit was hyperbolic, a third central control point would also display, allowing you to control the curvature of the line. Notice also that, although the harmonic fit is also a curve, its curvature is defined by b = 1. As a result, the curvature of a Harmonic fit cannot be manually modified.

Future Performance Forecasting Now that you have fully applied the limits and fitting options to build the history match, you are ready to begin forecasting the performance of the completion. Each time you choose an entity to forecast, such as a completion, OFM applies a series of default attributes to the analysis.

Viewing Historical Regression Attributes To view and modify default historical regression attributes: 1. Right-click on the forecast graph and click Scenario Manager. The Scenario Manger dialog is displayed. 2. Click Edit. The Edit Scenario: Default Scenario dialog displays. 3. Click the Historical Regression tab.

36


Schlumberger


4. Observe that the default analysis is a Rate-Time analysis. Observe also that the analysis Method is set to a default of Best Method. 5. Optionally, you can choose any of the other options as the default method. TIP: If wells in your OFM projects typically exhibit, for example, exponential decline, you may wish to set Method = Exponential as the default. This will save you time later, as no manual adjustment is needed. 6. Where appropriate, the Hyperbolic selection allows you to set a specific (known) b value as the default. 7. Observe that you may also set minimum and / or maximum b values.

Viewing Forecast Attributes To view and modify default forecast attributes: 1. Click the Forecast tab. The starting flow rate for a new forecast is set to be equal in value to the last historical data point. This appears as Start Rate = Last Historical Rate. There will be occasions when the last data point is not a good representative point. It is preferable to consider the forecast to be a simple extension or continuation of the fitted history match.


37


Schlumberger

2. Set Start Rate to From Fit.

3. Click OK to return to the Scenario Manager dialog. 4. Click Apply to save the changes. A confirmation dialog displays prompting you to confirm. 5. Click Yes to apply these settings to the current case. 6. Click Close to close the dialog and return to the graph.

38


Schlumberger


7. Observe the change in the behavior of the fit and forecast; they appear as one single line or curve. NOTE: All entities which do not already have a saved forecast will adopt this and the other default settings.

Working Forecast Parameters Phase : Oil Case Name : Case1 b :0 Di : 0.0117753 M.n. qi : 0.809686 ti : 01/31/2009 te : 01/31/2019 Final Rate : 0.197119 Cum. Prod. : 50.3675 M Cum. Date : 12/31/2008 Reserves : 1.5834 M Reserves Date : 01/31/2019 EUR : 51.9509 M Forecast Ended By : Time DB Forecast Date : Not Saved Reserve Type : None


5

1

0.5

0.2 1998 99 2000 01

02

03

04

05

06

07

08

09

10

11

12

13

14

15

16

17

18

Single Entity (Individual) Forecast Settings You modified the default settings for all entities, and now you will explore how to control the forecast of each entity individually. To modify individual entity forecast settings, choose an entity and double-click on the forecast graph. The Edit Scenario dialog is displayed.


39


Schlumberger

The dialog is split into five parts: Schedules

A schedule can be considered as a period in time during which the remaining settings apply. A forecast can be made up of as many as 20 schedules, each with its own settings.

Start settings

Controls the position of the first point on the forecast profile.

Limits

Controls how the forecast will end.

Decline parameters

Allows you to disconnect the forecast from the history match by applying your own values for the type and decline rate.

Advanced settings

Additional control options

You will not investigate every permutation in this dialog, but you will work through a few common options. (Refer to the online Help for full details.) 1. You will create a 5-year forecast for this one entity, subject to two additional constraints - an operational economic limit of 0.5 bopd and a reserves estimate. For completion HOGL1032, use the table to specify the following values for Schedule #1. Start Time

Months from End = 0

Start Rate

From Fit

Reserve Type

Proven-Developed

End Time

Months from Start = 60

End Rate

0.5

NOTE: Months from End means months from the end of the project, not just the entity in question. OFM will look for the last date for which there is production from any entity. This will be 'time zero' for all forecasts.

40


Schlumberger


NOTE: The Proven Developed setting is simply a label. It has no effect on the analysis, but is useful as a future filtering criterion. 2. Click OK. 3. Observe the effect. According to the results legend, the forecast ended ('te') on July 31, 2010, when the 0.5 bopd limit was reached. This is confirmed on the legend by 'Forecast Ended by = Rate.' Observe also that the volume produced by the forecast (Reserves) is 0.816 Mbbl. NOTE: Depending on how you limited your data, your results may be slightly different to these.

4. Double-click on the graph to display the Edit Scenario dialog.


41


Schlumberger

5. In the Schedule Reserves area, type 0.7 and click OK. Observe that the forecast has ended sooner (November 2009), and the final flow rate (Final Rate) is 0.543 bopd. TIP: Observe that, in addition to being able to type into the Schedule Reserves and End Rate boxes, the boxes are actually also drop-down lists and you may choose any project variable to represent your parameters. For example, if you have a table listing all the calculated economic limits for each completion, you can select the table as the source of the Schedule Reserves value. OFM will look up the appropriate value from the table and provide it to the forecast analysis.

42


Schlumberger


6. Edit the scenario again, this time setting End Time = 20100630 (June 30, 2010).

7. Click OK and observe that the forecast ended because you reached the time limit before either the rate or reserves limit was reached. The volume recovered (reserves) is reduced to 0.38 Mbbl.

Exercise 1

Saving the Forecast

For this training, assume the forecast for this completion is acceptable. Save the forecast to the project by either selecting File > Save or clicking

.

Your forecast for this entity is now committed to the OFM project, and you will see three changes to your OFM session:


43


•

Schlumberger

The red working forecast profile on the graph is now superimposed with a series of light blue symbols, indicating the profile which is now in the project database. You may continue editing the working (red) forecast, but the blue line will remain fixed until you overwrite it by saving the display.

• The results legend indicates that today was the last date on which a forecast was saved for this entity.

• The Forecasts pane now shows an Oil forecast was saved for completion HOGL1032 under Case1. (Cases are discussed later in the training.)

44


Schlumberger


NOTE: Although you have worked specifically with oil production data, the same processes are used when forecasting gas or water. The only difference is your choice of the appropriate variables (rate and cumulative, respectively).

Recalling a Saved Forecast or Working Forecast There is one more feature worth mentioning now that you have a saved forecast – recalling the data used to create the saved forecast. To recall the information: 1. Click View. 2. Toggle ON the Database Forecast view by selecting Working Forecast.

You can now see • data points used in the history match • the match itself • the subsequent forecast. 3. To restore the working forecast, repeat steps 1 and 2.


45


Schlumberger

Exercise 2

Editing the Forecast Graph

Before proceeding, now would be a good time to improve the appearance of the forecast graph. There are two areas you can influence - curves on the plot and graph headers (titles). Just as the analysis of a forecast is set by defaults, the display of a forecast window is set with defaults. You can control two aspects of the display: •

which curves are displayed

•

the graphic attributes of the curves on display

Controlling Which Curves Are Displayed To control which curves are displayed: 1. Select Edit > Graph, or right-click on the graph and select the Graph option from the menu. The Edit Plot window in the Forecast module is similar to the Edit Plot window in Plot module.

46


Schlumberger


You will notice that the curves are all grayed out. This means you cannot modify their definition, but you can disable them by clearing the box next to the appropriate curve. 2. Clear the Selected Historical (Line 5) and click OK. The result is shown in the figure.

3. Reverse the change before proceeding. 4. The forecast plot is a special plot, but it is still an OFM plot. You have the ability to add curves, add axes, add more graphs, and more just as in a conventional plot. 5. Select Edit > Graph, or right-click on the graph and select the Graph option from the menu.


47


Schlumberger

6. Click Add Curve and choose the Water.Cut calculated variable. Click OK.

7. Observe the new curve. This can be a useful 'visual aid' for your forecasting work. For example, you can display another curve as a guide to assessing where the wellhead pressure may be unstable. In this case, you may wish to remove the data points in that region.

Controlling Graphic Attributes To control the graphic attributes of the curves on display: 1. Select View > Enable Curve Select. OR Right-click on the graph and select Enable Curve Select. 2. Click on any curve to choose it.

48


Schlumberger


3. Use the Properties pane to alter the color of the curve, the thickness of the line, and more.


49


Schlumberger

Editing Forecast Graph Headers You can introduce headers on a forecast graph in exactly the same way as for a conventional OFM plot. To edit forecast headers: 1. Locate the Headers section of the Properties pane. 2. Click Setup to open a conventional Headers dialog.

3. Click Add. A new row is inserted into the Headers window. 4. Click Assist. The Edit Header window displays. 5. Enter "Forecast "+@Name(). Note the space after the word 'Forecast.' 6. Click OK. The Edit Header window closes and you are returned to the Headers window. 7. Click the Font button and apply the following settings: • Style = Bold Italic • Size = 14 • Color = Red 8. Click OK to return to the Headers dialog.

50


Schlumberger


9. Add another header that reports today’s date. The definition is @fmtdate(@today(), "Mmmm dd yyyy")

10. Click the Font button and apply the following settings: • Style = Bold • Size = 12 Your Headers window should resemble the screenshot in the figure.


51


Schlumberger

11. Click OK. The Headers window closes and the graph displays with the new titles. Forecast HOGL1032 100

January 20 2010

50

Producing Day O il Rate

10

5

1

0.5

Working Forecast Parameters Phase : Oil Case Name : Case1 b :0 Di : 0.0117753 M.n. qi : 0.809686 ti : 01/31/2009 te : 06/30/2010 Final Rate : 0.663421 Cum. Prod. : 50.3675 M Cum. Date : 12/31/2008 Reserves : 0.378075 M Reserves Date : 06/30/2010 EUR : 50.7456 M Forecast Ended By : Time DB Forecast Date : 01/19/2010 Reserve Type : Proven-Developed

0.1 1980 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99200001 02 03 04 05 06 07 08 09 10

Observe the Analysis pane. When you began working on this forecast, a new node displayed named Forecast1. Now that the graph looks reasonably complete, you may wish to rename the node, for example, Semilog Oil v Time, so that its format can be recalled. NOTE: The Forecast nodes which appear in the Analysis pane store information about how you like to view the analysis session. They do NOT contain the forecast results.

52


Schlumberger


This node represents a simple and quick pass through a forecast analysis for one completion in OFM. Your filter is comprised of eight completions and, as you progress through the others, you will explore additional capabilities.


What are Working Forecast and a Database Forecast?

•

Is it possible to delete a Working Forecast?

•

What is the physical meaning of the b factor in a Forecast curve?

•

We cannot select specific set of historical production points in a forecast. True or False?


setting up a forecast scenario

•

editing a graph

•

working with forecast features

•

changing the fit type

•

working with multi-schedule forecasts

•

performing a manual match.


53


Schlumberger

NOTES

54


Schlumberger

Additional Forecasting Features

Module 3 Additional Forecasting Features Up to this point, you have created a complete but relatively simple forecast that introduced you to a few of the more common forecasting utilities. In this module, you will explore these utilities in more detail and using the remaining completions in your filter as candidates.


work with phases

•

apply additional history matching data reduction features

•

perform group forecasting

•

work with cases

•

extract forecast results

•

automate forecast calculated variables and reports.

Lesson 1

Phases

There are three primary phases used in forecasting – oil, water and gas. This lesson shows you some additional possibilities for working with phases. Your work so far has been based on oil forecasting. You will now briefly learn how to work with the other two primary phases, gas and water. In addition you will learn how to create your own primary phase for forecasting in OFM.


55


Schlumberger

Exercise 1

Moving Between Gas and Water Phases

OFM has the default ability to allow you to switch easily from oil to each of the other conventional primary phases - gas and water. To move between phases: 1. Double-click on the forecast graph to display the Current Scenario dialog. 2. In the Phase / Analysis section, click Gas. 3. Observe that the three variables necessary to proceed with the analysis have changed. Set these parameters: • Time (Date) = Date • Cum Gas, MMcf = CumGas • Gas Rate, Mcf/d = Prd.PDGas NOTE: The variables for cum gas and gas rate must be set with Output Multipliers MM and M, respectively. When you choose these variables, OFM displays their multipliers next to the selection as confirmation that they are correct. If they do not match, you must stop the process and correct the discrepancy. 4. Click the Forecast tab. 5. Make the following changes and click OK. • Start rate = From Fit • End Rate = 0.01 Mcf/d

56


Schlumberger


Forecast HOGL1032 January 20 2010

10

PDGas, Mcf/d

1

0.1

0.01

Working Forecast Parameters Phase : Gas Case Name : Case1 b :0 Di : 0.00338535 M.n. qi : 0.0952202 Mcf/d ti : 01/31/2009 te : 01/31/2019 Final Rate : 0.0634344 Mcf/d Cum. Prod. : 1.8999 MM Cum. Date : 12/31/2008 Reserves : 0.285785 MM Reserves Date : 01/31/2019 EUR : 2.18568 MM Forecast Ended By : Time DB Forecast Date : Not Saved Reserve Type : None

0.001 198081 8283 84 8586 8788 89 9091 92 9394 9596 97 989920000102 0304 05 0607 08 0910 1112 13 1415 16 1718 19

6. Observe the effect. With a small number of changes, you can now work exclusively with gas production data. 7. Assuming the fit and forecast are acceptable, save the forecast by selecting File > Save.


57


Schlumberger

The Forecasts pane changes to indicate this completion was saved as a gas forecast, under Case1.

8. Double-click on the forecast graph to display the Current Scenario dialog. 9. In the Phase / Analysis section, click Water. 10. Observe that the three variables necessary to proceed with the analysis have changed. Set these parameters: • Time (Date) = Date • Cum Water, Mbbl = CumWater • Water Rate, bbl/d = Prd.PDWater NOTE: As before, the variable for cum water must be set with Output Multiplier M. If it does not, you must stop the process and correct the discrepancy. 11. Click the Forecast tab. 12. Make the following changes and click OK. • Start rate = From Fit • End Rate = 0.1 bbl/d

58


Schlumberger


13. Save the forecast. The Forecasts pane changes to indicate this completion was saved as a water forecast, under Case1.

Exercise 2

Forecasting with User-Defined Phases

OFM 2009 introduced the concept of user-defined primary phases. Now, you can add your own project-specific phases and use the entire range of functionality to create forecasts of the new phases. Scenario: Completion AOC634 has condensate data, and you would like to forecast it. To forecast with user-defined phases: 1. Create variables for Cumulative Condensate and Producing Day Condensate Rate. The source table variable for both is Prd.Cond. Use the corresponding calculated variables for cumulative oil and producing day oil rate as a guide. 2. Select Tools > Additional Phase Setup. 3. In the new dialog, click Add. 4. In the Name box, type Condensate and click OK.


59


Schlumberger

Leave the phase Type set as Liquid.

5. Click OK again to save your changes and exit the dialog. 6. Select completion AOC634. TIP: You must clear the existing filter to access the completion. 7. Double-click on the forecast graph to display the Current Scenario dialog. 8. In the Phase / Analysis section, click the new entry Condensate. 9. Observe that the three variables necessary to proceed with the analysis now reflect condensate values. Set these parameters: • Time (Date) = Date • Cum Condensate, Mbbl = CumCond • Condensate Rate, bbl/d = Prd.PDCond NOTE: As before, the variable for cum condensate must be set with Output Multiplier M. If it does not, you must stop the process and correct the discrepancy. 10. Click the Forecast tab. 11. In the Condensate Schedules area, click Add to create a new schedule. 12. Make the following changes and click OK. • Start Time = Last Historical Date • Start rate = From Fit • End Rate = 0.1 bbl/d

60


Schlumberger


Forecast AOC634 January 20 2010

1

Working Forecast Parameters Phase : Condensate Case Name : Case1 b :0 Di : 0.00936056 M.n. qi : 0.0290489 ti : 06/30/1991 te : 12/31/2000 Final Rate : 0.00998639 Cum. Prod. : 0.1419 Mbbl Cum. Date : 06/30/1991 Reserves : 0.0619851 Mbbl Reserves Date : 12/31/2000 EUR : 0.203885 Mbbl Forecast Ended By : Rate DB Forecast Date : Not Saved Reserve Type : None

0.5

0.1

PDCond

0.05

0.01

0.005

0.001 1980 81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99 2000

13. Save the forecast. The Forecasts pane changes to indicate this new completion also has a saved condensate forecast under Case1.


61


Schlumberger

Lesson 2

Additional Scenario Settings

The future performance of any entity may be influenced by of a series of events. For example, a well may be subjected to a workover in a year’s time. Two years later, all wells may be shut in for a field maintenance program. OFM allows you to acknowledge these known events by building the forecast from a sequence of time-periods. Those individual time-periods are represented by OFM through Schedules. A Schedule is a time period during which a set of assumptions will apply.

Exercise 1

Configuring a Two-Schedule Forecast

Completion HOGL1132 is not performing well and decline since 2006 has accelerated. Untapped sand was identified in the wells in this area, and a perforation program was begun in early 2010. Early results indicate that a 20% improvement can be expected. Completion HOGL1132 is to be included in the perforation program. It is therefore reasonable to conclude that the future of this completion will be in two distinct parts: • The period leading up to the perforation activity during which it will decline as normal. • The period following the workover, when you anticipate that it will begin flowing at a rate 20% higher than the rate immediately before the work. Finally, as this represents new production, we anticipate that the new decline will be parallel to the previous decline. This is a good example of the need for a multiple schedule forecast. To set up a two-schedule forecast: 1. Select completion HOGL1132. (You may wish to clear your filter and reselect the Best Oil OFM Query filter.) 2. Switch to Phase = Oil. 3. Apply a Range Limit to the data by excluding all data prior to January 1, 1998. 4. Double-click on the graph to display the Current Scenario dialog.

62


Schlumberger


5. Click the Forecast tab. 6. The settings on the page are for Schedule 1. Allow the completion to decline until the date of the perforation job, so set the following: • Start Rate = From Fit • End Time = Date (20100301) • End Rate = 0.01 (This will ensure that rate does not constrain the schedule) 7. Return to the top of the dialog, where you can enter the second schedule, which will model the effect of the perforation. 8. In the Oil Schedules section, click Add to create Schedule 2. The Start Time setting of Months From End in this instance will mean from the end of Schedule 1. So the default value of zero is correct. 9. Start Rate defaults to Previous, which is the rate at which Schedule 1 finished. This is not what you want. You need to enter the effect of the work, so set Start Rate = Change By Percent to 20. 10. Click OK and observe the effect. OFM has honored the increase in productivity in March 2010 as a result of the perforation.


63


Schlumberger

Forecast HOGL1132 Working Forecast Parameters Phase : Oil Case Name : Case1 b :0 Di : 0.00796844 M.n. qi : 1.40232 ti : 01/31/2009 te : 02/28/2010 Final Rate : 0.583459 Cum. Prod. : 62.8225 M Cum. Date : 01/31/2009 Reserves : 4.09441 M Reserves Date : 03/31/2020 EUR : 66.9169 M Forecast Ended By : Time DB Forecast Date : Not Saved Reserve Type : None

January 20 2010


3

1

0.6 2002

03

04

05

06

07

08

09

10

11

12

13

14

11. Save the forecast.

Control the Start and End Times The Forecast tab offers many additional controls over how your forecasts begin and end. Many controls are self-explanatory and can be quickly demonstrated in the working forecast. More comprehensive details may be found in the online Help.

64


Schlumberger


Start Time Controls Date

A specific date, in YYYYMMDD format

Months from End

A specific number of months after the last historical date in the database for ALL the completions in the project. This allows forecasts for multiple completions to begin on the same date.

Last Historical Date

Date of the last historical data point for the current completion or group, regardless if it is included in the fit.

From Historical The date of the last historical data point included in the fitted history match for the current completion or group.

Start Rate Controls Previous

Starts the forecast at the final rate from the previous schedule. (Not intended for the first schedule.)

From Historical The rate of the last historical data point included in the fitted history match for the current completion or group. Last Historical

Starts the forecast from the last historical production rate, regardless if it is included in the fit.

Value

Begins the forecast schedule at a specific value.

Change By

Changes the default starting rate by the given rate.

Change By Percent

Changes the default starting rate by the given percentage.

From Fit

Calculates the starting rate at the specified start time using the fit equation.

Last Fit Rate

Calculates the starting rate at the specified last historical date using the fit equation.


65


Schlumberger

Shut in

Indicates that production is to be stopped for the duration of the schedule. The result is a gap in the graph, indicating there will be no production during the time period. The graph line resumes at the end of shut in period.

Reserve Type By default, the list of entries available to you is consistent with SPE standard nomenclature for reserves estimation. As indicated earlier, regardless of the choice you make here, it has no effect on the analysis or the results generated. However, when the forecast is saved, this information is also saved. It can be used in the System Function module for tasks such as filtering criterion (@DCAReserveType()). Another example might be in the preparation of a report of proven, probable and possible reserves. OFM 2009 introduced the idea of allowing you to edit the list of default entries and to add your own.

Adding Default Entries To add your own default entries: 1. Select Tools > Reserve Type Setup. The Reserve Type Setup dialog displays. 2. Delete an entry by highlighting it and clicking Delete. 3. Add a new entry by clicking Add and entering the new name.

4. Click OK and the new name appears on the list. 5. Click OK to save your changes.

66


Schlumberger


6. Double-click on a forecast graph and click the Forecast tab

7. Observe the new entry.

Decline Parameters There may be occasions when, even if you have a satisfactory history match, you may wish to create a forecast with different settings. For example, your match may show it to be hyperbolic but, for a conservative forecast, you may wish the forecast to be expressed as an exponential decline.

Exercise 2

Creating a Forecast with UserDefined Parameters

To create a forecast with user-defined settings: 1. Choose completion HOGL2_1029 2. Use any technique you wish to exclude all data points prior to the spike in production in the middle of 2005. Note that the auto fit through this data seems to be hyperbolic (b = 0.586).


67


Schlumberger

Forecast HOGL2_1029 Working Forecast Parameters Phase : Oil Case Name : Case1 b : 0.585879 Di : 0.0181868 M.n. qi : 2.34274 ti : 01/31/2009 te : 01/31/2019 Final Rate : 0.574499 Cum. Prod. : 52.021 M Cum. Date : 12/31/2008 Reserves : 4.17781 M Reserves Date : 01/31/2019 EUR : 56.1988 M Forecast Ended By : Time DB Forecast Date : Not Saved Reserve Type : None

January 20 2010

30


10

5

1 0.8 1999

2000

01

02

03

04

05

06

07

08

09

10

11

12

13

14

3. A hyperbolic forecast may prove overly optimistic for this well. Other wells in the area tend to exhibit exponential declines of approximately 2% per month. To recognize this, make changes to the Forecast tab, as shown in the figure.

4. Clear the checkmark next to the Reinitialize box. Click OK and observe the effect. The future trend for this completion is a straight line exponential, despite the historical regression match being hyperbolic.

68


Schlumberger


Forecast HOGL2_1029 Working Forecast Parameters Phase : Oil Case Name : Case1 b :0 Di : 0.02 M.n. qi : 2.34274 ti : 01/31/2009 te : 01/31/2019 Final Rate : 0.212599 Cum. Prod. : 52.021 M Cum. Date : 12/31/2008 Reserves : 3.24182 M Reserves Date : 01/31/2019 EUR : 55.2628 M Forecast Ended By : Time DB Forecast Date : Not Saved Reserve Type : None

January 20 2010

30


10

5

1 0.8 1999

2000

01

02

03

04

05

06

07

08

09

10

11

12

13

14



69


Schlumberger

Lesson 3

Re-Initializing

Earlier, you modeled the effect of adding a new producing layer in the well. In terms of reservoir recovery, this action has the effect of adding incremental reserves. The new line adds a 'wedge' of new production on top of the base profile. There will be occasions, however, when the activity is aimed at returning the well to its original conditions as a way of accelerating the recovery of existing reserves.

Exercise 1

Simulating the Effect of Rate Acceleration Techniques

Now, your completion will be subjected to work designed to return the well to its original conditions to speed up the process of recovery of reserves. To simulate the effect of rate acceleration: 1. Choose completion HOGL3_1132. The default fit through this dataset is hyperbolic. This serves very well for the purposes of illustrating the feature. 2. In the Forecast tab, change the End Rate to Value = 0.5 bbl/d. 3. Save the forecast without changes. Note the estimated reserves at 2.03 Mbbl. 4. Make these changes to the Forecast tab: • Start Rate = Value = 2 bbl/d (The result of the stimulation work, performed today) • End Time = Months From Start = 240 5. Check the box marked Reinitialize and click OK to observe the effect of this change.

70


Schlumberger


6. Save your new forecast. Forecast HOGL3_1132 100

50


10

January 20 2010

Working Forecast Parameters Phase : Oil Case Name : Case1 b : 0.264346 Di : 0.0228463 M.n. qi :2 ti : 01/31/2009 te : 03/31/2015 Final Rate : 0.495025 Cum. Prod. : 44.4553 M Cum. Date : 01/31/2009 Reserves : 2.32529 M Reserves Date : 03/31/2015 EUR : 46.7806 M Forecast Ended By : Rate DB Forecast Date : 01/20/2010 Reserve Type : None

5

1

0.5

0.1 1990 91 92 93 94 95 96 97 98 99 2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16

Note that the forecast is not an extension of the fit; it is a copy of the production history, repositioned to start at 2 bbl/d. The shape may look a little different, but that is the effect of the logarithmic scale. The reserves are recovered over a shorter time period, approximately 1.5 years sooner. Working to the same rate limit of 0.5 bbl/d, you see it may be possible to also improve the recovered volume, from 2.03 Mbbl to 2.32 Mbbl.


71


Schlumberger

This is the effect of the Reinitialize feature - you are simulating the return to initial conditions by replicating the original history. WARNING: OFM is not a reservoir simulator. The results generated by this feature should be considered no more than engineering estimates.

Lesson 4

Additional History Matching Data Reduction Features

In this section, we will explore two additional options for controlling the data used in the history match process - digitizing points and setting a variable limit.

Exercise 1

Digitizing Points

There may be occasions when none of the data points are reliable, or when there are too few to perform a meaningful statistical match. In this case, OFM allows you to build a forecast using your own data points instead. To discard ALL historical data points and perform a fit using an alternative set: 1. Select completion HOGL3_0832. 2. For various reasons, it has been decided that this entire data set is unreliable. To replace it, right-click on the graph and click Digitize Points. 3. Using the left mouse button, click (digitize) a series of points on the graph which represent your opinion of the true performance of this completion. Use the figure as a guide.

72


Schlumberger


4. After creating the last point, right-click and select Done. OFM removes all the historical data, and performs a conventional fit through only your points.


73


Schlumberger

5. Save the forecast. Forecast HOGL3_0832 100


10

January 20 2010

Working Forecast Parameters Phase : Oil Case Name : Case1 b :0 Di : 0.0204607 M.n. qi : 0.411647 ti : 01/31/2009 te : 11/30/2014 Final Rate : 0.0983969 Cum. Prod. : 41.5544 M Cum. Date : 06/30/2007 Reserves : 0.465993 M Reserves Date : 11/30/2014 EUR : 42.0204 M Forecast Ended By : Rate DB Forecast Date : Not Saved Reserve Type : None

1

0.1

0.01 1989 90 91 92 93 94 95 96 97 98 99 2000 01 02 03 04 05 06 07 08 09 10 11 12 13 14

74


Schlumberger

Exercise 2


Using a Calculated Variable to Exclude Data Points

As evidenced by many completions in this filter, there are considerable numbers of abnormally high and low data points. They could all be removed using the various techniques already described, but OFM offers one more feature - excluding data based on the result of a calculation expression. To set a variable limit, select a variable which returns a True or False (1 or 0). A data point will be excluded from the analysis if the value of the variable is False (or zero) on the same date. Conversely, any data point which maps to a variable value of True (or one) on the same date will be included in the analysis. To illustrate this, you need a variable. In principle, any variable which returns a True or False result is valid but, here, you will build one which is based on the data you wish to affect. For example, you may decide if the value of oil rate on any date is within 10% of the previous monthly value, the point is considered to be valid. If not, the point is considered bad. (This is not very scientific, but it illustrates the principle.) Here is the corresponding Calculated Variable definition which matches that statement: Test = @if(@abs((prd.pdoil@previous(prd.pdoil))/@previous(prd.pdoil))< 0.1, 1,0) To set a limit using the results of a variable: 1. Chose completion HOGL3_1429. 2. Right-click on the graph and select Limits > Variable.


75


Schlumberger

3. Check the box Use Variable and choose the Calculated Variable Test.

4. Click OK and observe the result. A large number of abnormal data points, based on the criterion applied by the Calculated Variable, have been very quickly eliminated. Forecast HOGL3_1429 Working Forecast Parameters Phase : Oil Case Name : Case1 b : 1.29165 Di : 0.00377155 M.n. qi : 2.495 ti : 01/31/2009 te : 01/31/2019 Final Rate : 1.74708 Cum. Prod. : 42.0195 M Cum. Date : 01/31/2009 Reserves : 7.56122 M Reserves Date : 01/31/2019 EUR : 49.5807 M Forecast Ended By : Time DB Forecast Date : Not Saved Reserve Type : None

January 20 2010

200

100


50

10

5

1 1992

93

94

95

96

97

98

99

2000

01

02

03

04

05

06

07

08

09


76


Schlumberger

Exercise 3


Task 1: Forecasting Oil for HighVolume Producers

Now that you have seen the range of capabilities OFM has to offer in this analysis module, you are in a position to complete Task 1 of the Provo project workflow. Forecast the future performance of the following completions: •

Oil rate for all oil completions which historically have produced more than 40,000 bbl

•

Gas rate for all gas completions which have produced more than 25 MMcf

As you may remember, there are eight oil completions and eight gas completions. Use your best judgment to determine the most appropriate fit through the data, and apply a variety of forecast settings and limits. The purpose of this exercise is practice, and your instructor is available if you have questions.

Lesson 5

Group Forecasting

Up to now, you have been concerned with forecasting individual completions. Soon, you will want to forecast entities at higher levels in the project, such as fault compartments, reservoirs, and fields. These are all examples of group forecasts. In OFM, the concept is no more complex than what you have already learned, but there are one or two new aspects you should understand. To explore the concept of group forecasts, return to the Best Oil collection of completions.


77


Schlumberger

Exercise 1

Task 2: Forecasting Group Oil

This exercise is equivalent to completing Task 2 of the Provo project forecasting workflow. To forecast group oil: 1. Clear the current filter. 2. Load the Best Oil filter from the OFM Query section of the Filter pane.

This filter contains eight completions. To create a group representing the aggregated (summed) performance of all eight, click Group

on the Filter pane.

The entry in the main entity selector toolbar shows that OFM is now working with an 'object' that is a group titled Completions Selected (8).

You can proceed to forecast this new object. 3. If a forecast window is not already open, open one (Analysis > Forecast). 4. The graph displays a default fit and forecast for the new object (group).

78


Schlumberger


From this point forward, you use exactly the same tools and techniques as we did for the individual completion forecasts. There is absolutely no difference in the analysis procedure. You are already fully familiar with these techniques, so they will not be repeated here. Forecast Completions Selected (8) 1000

January 27 2010

500


100

Working Forecast Phase Case Name b Di qi ti te Final Rate Cum. Prod. Cum. Date Reserves Reserves Date EUR Forecast Ended By DB Forecast Date Reserve Type


50

10

5

1 198081 8283 84 8586 8788 89 9091 92 9394 9596 97 989920000102 0304 05 0607 08 0910 1112 13 1415 16 1718 19

5. Apply a single data reduction filter and set the following range limit on Date: • Minimum = 19940101 • Maximum = 20021231


79


Schlumberger

6. Observe the revised fit and forecast.

7. Set a forecast End Rate limit of 3 bbl/d. Observe that the forecast ends in mid-2017, and a reserve estimate of 18.1 Mbbl is calculated.

80


Schlumberger


8. OFM offers you the option of changing the name of this new entity. Rename it Best Oil Group and save the new forecast by selecting File > Save.

9. Click OK. 10. Inspect the Forecasts pane, as a new entity is displayed. OFM recognizes this entity has a saved oil forecast under Case1.

Exercise 2

Task 3: Forecasting Group Gas

Task 3 is similar to Task 2, but it focuses on gas wells instead of oil wells. To forecast group gas: 1. Clear the current filter. 2. Apply the Best Gas filter. (Gas completions which have produced more than 25 MMcf.) 3. Group the resulting eight completions. 4. Choose the Gas phase for analysis.


81


Schlumberger

5. The aggregated performance is rather erratic, with no single candidate to draw your fit through. This may be an instance for a manual fit but, in the end, it is up to you. 6. Apply your historical fit, and create a forecast. 7. Save the forecast as Best Gas Group. 8. Confirm that the new group forecast entity is captured in the Forecasts pane.

Forecasting Categories You created group forecasts by aggregating completions from various filters. Categories also represent groups of completions. Category data forecast follows the same procedure as you have seen for groups. The only difference is in the way you load the objects you wish to work with. To forecast a category: 1. Clear the current filter. 2. From the Step pane, choose Flow_Station.

3. From the resulting list, choose Flow Station TB0532. Observe from the basemap that these completions are in the Provo east field, so they will be oil producers. 4. Open a new forecast window. 5. Ensure that Phase is set to Oil.

82


Schlumberger


6. Set Case to Base Case. Observe from the title of the graph and the legend that the object being worked on represents the total oil rate of all completions which have been mapped to the TB0532 Flow Station. 7. At this point, you may use all of the tools for fitting and forecasting you have used so far. For the purposes of training, however, the unedited fit looks acceptable, so select File > Save. 8. Observe the behavior of the Forecasts pane. You now have a saved base case oil forecast for Flow Station TB0532.

Exercise 3

Task 4: Forecasting the Performance of the VK 1 Pool

Using the technique for categories forecast, create and save a base case gas forecast for the VK 1 Pool. 1. Clear the current filter. 2. Select Step = Pool. 3. Select Pool = VK 1. 4. Set Phase = Gas 5. Set Case = Base Case


83


Schlumberger

6. Proceed as normal with your own fit and forecast settings. 7. Save the forecast. 8. Confirm details in the Forecasts pane.

Lesson 6

Cases

From time to time, saving a forecast under a case name has been mentioned. In OFM forecasting, cases have a special role to play. For example, you may wish to create and save for future comparison several forecasts for one well. The first may represent your base case - do nothing, but allow the well to decline as normal. •

A second forecast for the same well may assume a more optimistic future than the original projection.

•

A third one may assume the opposite - that the future profile is more pessimistic than the original projection.

•

A fourth situation may assume that a workover will be performed in this well and that an increase in production rate is anticipated.

•

A fifth possibility might assume that the workover will be delayed by a year.

Each of these forecasts is a valid analysis, but if you overwrite the old analysis each time, all the old work is lost and you cannot perform the comparison. To navigate through this problem, OFM uses the concept of cases. All of these scenarios can be captured independently in OFM, by saving them under different cases. TIP: Think of a case as a folder containing a collection of forecasts which all follow the same set of assumptions, for example 'workover delayed by 1 year'. It makes logical sense to keep all the forecasts that adopt the same rules in the same place. This is only a suggestion, but cases are a good method of organizing your forecasts in a way that makes sense to you.

84


Schlumberger


With this additional 'layer' of information, working with forecast results introduces another dimension to your ability to differentiate exactly which results set you wish to work with. When identifying the forecast you wish to analyze, it is possible to recall not only the entity and phase, but also the specific assumptions, such as the case name. In Figure 7, take a look at how OFM embeds case information in your work. Observe your Forecast window.

Figure 7

Your Forecasts window

Note the arrowed regions. The active (current) case name information is confirmed in two locations - the Current Case entry on the Forecasts pane and the results legend. Additionally, a 'tree' structure is prominent on the Forecasts pane that identifies the status of each case in the project.


85


Schlumberger

The active case name also appears in the main title bar at the top of the OFM application window. There is no excuse for not knowing which is the current Case!

Exercise 1

Setting Up a Default Case

Your first OFM forecast session defaults to Case1. All the work you have done so far has been under Case1. Assume that all your existing work is a 'base case.' You may wish to rename the default case. To set up the default case: 1. Right-click on the Case1 node on the tree structure. 2. Select Rename. 3. Rename it Base Case. 4. Press Enter to save the changes. Observe now that all the entries with arrows (Figure 7) have been renamed.

Working with Cases Think back to the five possible situations in which you might wish to forecast for any well. Your first task is to create new nodes or folders in your tree, representing the new cases. To create a new case: 1. Click Add Case

on the Forecasts pane.

2. Enter the name Optimistic. 3. Optional: Enter a Userid and Comment. (The Comment field allows you to describe in detail the reason for creating this case.) 4. Click OK. Observe that the new case node appears on the tree, and that it is automatically assigned the new active (current) case.

86


Schlumberger


Exercise 2

Adding New Cases

Create these new cases: •

Pessimistic

•

Workover

•

Workover Delay 1 Yr

Current Case At any time, the current case (also termed the active case) may be assigned using the Current Case dialog at the top of the Forecasts pane.

Figure 8

Forecasts pane

To re-assign the current case, open the drop-down list and choose the relevant case. TIP: Forecasts are normally saved to the current case.


87


Schlumberger

Exercise 3

Changing the Current Case

Make Case1 the active case.

Changing the Order of the Display The Forecasts pane shows, in a tree format, information regarding the forecasting work you have saved in this project. Observe how the tree is organized.

Figure 9

Organization of the tree format

While this may be a useful way of organizing the data, it becomes difficult to use when attempting to answer a question such as, ‘Which cases contain saved gas forecasts for well X?’ OFM 2009 introduced the means to answer the question by allowing you to re-organize the tree. To re-arrange the Forecasts pane tree: 1. On the Forecasts pane toolbar, click the drop-down arrow on the

icon.

2. Select Entity > Phase > Case option. 3. Observe the new behavior of the pane tree.

88


Schlumberger


Among other things, it shows that, for the Best Gas group of completions, a gas forecast has been saved to only the base case.

Compare Cases One of the primary roles of a petroleum engineer is to evaluate the effect of alternative courses of action for a well or group of wells. A vital tool in this analysis is the comparison of the projected future of the well under each of the possible alternatives. This requires creating forecasts, and comparing them for each possibility.

Creating New Futures To illustrate OFM’s capability in this respect, create new futures for a well. To create new futures: 1. Clear any existing filter. 2. Choose completion HOGL1032. 3. Open a new forecast window. 4. Set the Phase to Oil. 5. Set the Current Case to Base Case. 6. Confirm that the graph displays an existing saved forecast (light blue line).


89


Schlumberger

7. Set the Current Case to Optimistic. OFM should display an unedited dataset, representing the fact that no saved forecast exists for this well in this case. You could proceed with a conventional data reduction exercise, but that is not the objective here. All you need to do is save the forecast under this new case name. 8. Select File > Save. The Forecasts pane should now indicate a saved oil forecast.

9. Set the Current Case to Pessimistic. Because you made no changes for the Optimistic case, the new case will initially follow the same profile. 10. Double-click on the graph to edit the forecast settings. 11. Set Start Rate = Last Historical Rate. 12. Click OK. 13. Save the new forecast by selecting File > Save. Again, the tree on the Forecasts pane should update to reflect that there are now three saved oil forecasts for this completion.

90


Schlumberger


Comparing Cases To compare all three forecasts on the same graph: 1. Select View > Cases. The Display Multiple Cases dialog displays.

2. Select (check) the first three items on the list and click OK.


91


Schlumberger

NOTE: At any time, there can only be one active (current) case. This is evident on the graph in two places - the active case displays a working forecast (red dotted line), and the new legend shows the attributes of the non-active cases. TIP: Even though you are in a comparison 'mode' of display, you may continue working as normal. You may choose to switch the active case, and return to an earlier forecast to make adjustments.

92


Schlumberger


Exercise 4

Task 5: Forecasting Multiple Cases

This task calls for the creation of multiple cases to conduct the comparison. Completion HOGL2_1129 has declined since it began production in 1992, and an unedited fit through the historical data shows harmonic decline. In 2008, however, the performance of this completion began to deteriorate rapidly, as shown in Figure 10.

Figure 10

Harmonic decline graph


93


Schlumberger

To assess the feasibility of a potential workover, three forecasts are required: •

A 'do nothing' forecast, assuming the completion will continue as per the current forecast. This begins at the last historical data point.

•

Assuming that a short workover on this completion will be successful in June, 2010, and that production will return to its historical decline curve. NOTE: This requires a second scenario.

•

Assuming the work will be delayed an additional six months.

Exercise 5

Task 6 Comparing all Three Forecasts Graphically

Save all three forecasts to individual cases and use the Case Comparison feature to graphically display all three forecasts.

Extract Forecast Results When you choose to save an OFM forecast, OFM saves a large volume of information, including the future profile and all 'single value' results such as b value, Di, qi, reserves, and more. Also, a single forecast could be for: •

any one of several entities (completions or groups)

•

any one of several phases (oil, gas, or water)

•

any one of several cases

Clearly, there are numerous combinations of these three elements, and each combination can be a separate forecast. Extraction of the result of any one combination requires that all three elements must be specified.

Forecast System Functions It will come as no surprise that System Functions play a pivotal role in this process. A search of the online Help system (Help > OFM Help > Search = Forecast System Functions) returns a list of eight functions.

94


Schlumberger


Each function is designed to extract specific results from the collection you saved. They fall into two groups - functions which return profiles (time- and cumulative phase-based), and functions which return single values, such as Di, qi.

Profile-Based Forecast Results To illustrate how forecast profiles may be extracted, proceed through a common workflow. NOTE: Because you have not yet created Ratio forecasts, you are unable at this time to illustrate cumulative phasebased profiles.

Exercise 6

Task 7: Comparing Group and Sum Forecasts

A common workflow is the comparison of a single group forecast (for example, the full field as a single entity) with the sum of the forecasts of the individual completions or wells which make up that entity. By conducting both, you can achieve a degree of certainty that your work is valid. To compare a single group forecast with the sum of the forecasts of each of the completions which make up the group, you will create calculated variables to represent both results. To create calculated variables: 1. Create a calculated variable to retrieve the forecast oil rate for a single group of completions. FCST.Group_Oil=@Forecast(Date,“oil”,“Base Case”,“group”) Observe the arguments necessary for this function: • Date • Phase (in quotes) • Case Name (in quotes) • Option (sum or group) 2. Assign units to the variable: Unit = bbl and Multiplier = 1.


95


Schlumberger

3. Create a calculated variable which sums the individual saved oil rate forecasts for all completions belonging to a particular group. FCST.Sum_Oil = @Forecast(Date,“oil”,“Base Case”,“sum”) 4. Assign units to the variable: Unit = bbl and Multiplier = 1. To make use of these new tools, you need to recall a group forecast. The only place you can do that is in the Forecasts pane; simply regrouping the completions from the Filter pane is insufficient. The group 'object' – the Best Oil group - exists only as a forecast group and the forecast tables contain the information about which completions make up the group. Without this knowledge, your workflow will fail. 5. In the Forecasts pane, choose the entity Best Oil Group. 6. Click

to recreate the group.

7. The comparison will be made in a plot window. A comparison can be done in a forecast window, but it is likely you will want to capture the final comparison in a presentation style graphic, such as a plot. The forecast window is where you do only the basic work. 8. Open a new Plot window. 9. Plot the following variables against time (Date): • Prd.PDOil (The historical data) • FCST.Group_Oil • FCST.Sum_Oil 10. Observe the results. Note that the two profiles are different, which is not surprising. The criteria you used for building the individual completion forecasts (such as, data reduction, fit type, forecast constraints, and more) are not the same as the criteria you adopted for the single group forecast. Nevertheless, this is a very useful tool when creating fieldwide production forecasts. By proceeding down two independent avenues, you can satisfy yourself that your assumptions are defensible.

96


Schlumberger


Single Value Forecast Results In addition to extracting profile based results, you can use special system functions to recover single value results, such as the Arps ‘b’ exponent, recovered reserves, ultimate recovery, and others. Illustrate the concept with an example. To display forecast reserves and ultimate recovery: 1. Create a calculated variable for the volume recovered by the forecast. This is termed Remaining Reserves: FCST.RemRes = @DcaResults("oil RES","Base Case",1)


97


Schlumberger

NOTE: The @DcaResults system function requires three arguments: - Phase and Result Parameter (both inside same quotes) - Case Name - (Optional) Schedule Number (default = 1) See the online Help for details. 2. Assign units and a multiplier to the new variable: Unit = bbl and Multiplier = M. 3. Create a calculated variable to represent the total volume expected to be produced by this entity. Expected ultimate recovery (EUR) is equal to the historical production plus the forecasted reserves. FCST.EUR = @DcaResults("oil EUR","Base Case",1) 4. Assign units and a multiplier to the variable: Unit = bbl and Multiplier = M. 5. Use the Filter pane to choose the completions with saved Base Case Oil forecasts.

98


Schlumberger


Displaying Results in a Report A useful form of displaying the results is in a report. Follow these instructions to display the results. To display results in a report: 1. Select Analysis > Report to open a Report window. 2. Select Edit > Report Parameters. The Edit Report dialog displays. 3. In the Select section of the Edit Report dialog, enter: Date, Prd.PDOil, FCST.Group_Oil, FCST.Sum_Oil, FCST.RemRes, FCST.EUR 4. Click OK. 5. Choose completion HOGL1032 from the filter list. The report displays.


99


Schlumberger

6. Use the Next and Previous buttons to navigate through all the completions in the filter. As you can see, for individual completions the Group and Sum values are the same. NOTE: You are comparing a group of 1 completion with the sum of 1 completion. 7. In the Forecasts pane, double-click on the Best Oil group node to load the group entity. 8. Observe the new results. OFM has successfully recovered the forecast profile, and you can once again see the difference between the Group and Sum options, now that you are displaying for a group. The remaining reserves and EUR display for the group as single values that are not time-dependent. The display is a time-dependent report that is simply a means of displaying these single values.

Lesson 7

Automated Forecast Calculated Variables

Although the process of creating a forecast is highly interactive using rich graphical techniques, in versions of OFM before 2009, the next step of extracting the results of a forecast work can be complex. This complexity is difficult for anyone other than expert users to apply system functions and calculated variables. The value of the work is compromised if the results cannot be reached easily and effectively. OFM 2009 introduced two new features aimed at getting around this obstacle - Automated Calculated Variables and Automated Forecast Reports. This lesson discusses the former, while the next lesson presents the latter.

100


Schlumberger


Using the Automated Calculated Variables Feature To use the Automated Calculated Variables feature: 1. Choose a node in the Forecasts pane which represents a case and a phase, as shown in the figure.

2. Click . The Create Forecast Calculated Variables dialog displays. NOTE: When you first begin working in the Forecast module, the Automated Calculated Variable option is disabled. Only when a forecast has been saved does the option become available.


101


Schlumberger

Observe several features in this window: • A total of nine calculated variables have been prepared for you, each representing the most commonly used results form an OFM forecast. • The name of each forecast calculated variable has the prefix 'DCA,' followed by the case name. In this instance, each calculated variable carries the prefix 'DCABase_Case.' • You can edit the prefix by overtyping the entry in the Prefix text box. Try replacing the default entry with Base. Notice that all of the forecast calculated variables automatically rename. • Columns labeled Description and Variable are not editable in this dialog. TIP: These variables are controlled by a file ForecastCalcVars.xml, located in your OFM installation folder. You may edit the contents of the file in the install folder, including the default descriptions and variable names. • The column labeled Name may be edited. Any changes here are immediately reflected in the tree on the left and in the Variable column. • Initially, all forecast calculated variables are selected. You may choose to accept or reject one or more variables. 3. Click OK to accept your changes, and commit the forecast calculated variables to your project. WARNING: Future changes to parameters which are included in the saved forecast calculated variables, such as the case name, are not automatically reflected in your saved forecast calculated variables. If you make any such change, you must manually edit each of the calculated variables are affected by the change. 4. Confirm the changes by reviewing your project calculated variables (Database > Calculated Variables). The nine new variables are now in your project, and each variable is fully and accurately defined, including units and multipliers.

102


Schlumberger


Saving New Gas Forecast Calculated Variables To save new gas forecast calculated variables: 1. Choose a gas forecast node from the Forecasts pane. If possible, choose one under a case other than the base case. 2. Use the Forecast Calculated Variable utility to save all nine calculated variables, but using a new prefix of your choice. 3. Confirm your action by reviewing the list of project calculated variables.

Lesson 8

Automated Forecast Reports

Reports are not the most exotic of OFM’s visualization techniques, but reporting is a universal form of display and data transfer. Just as new automated calculated variables become available in the new Forecasts pane toolbar so, too, does the second new feature, the Automated Forecast Report, also introduced in OFM 2009. As with the Automated Forecast Calculated Variables feature, the Automated Forecast Report feature is not available until a forecast has been saved.

Using the Automated Forecast Report Feature To use the Automated Forecast Report feature: 1. Choose a node on the Forecasts pane which represents a case and a phase, as you did for the calculated variables. 2. Click

in the Forecasts pane toolbar.

3. The drop-down arrow indicates you must make a choice. Forecast Parameters Report Case: OR Forecast Report Case: where and reflect the choices you made from the tree.


103


Schlumberger

Forecast Parameters Report Choosing this option returns a report with one line of results for the entity. If you choose completion HOGL1032, you should see the report in Figure 11.

Figure 11

Sample report using completion HOGL 1032

Each entry in the report is a single result because there is no time dependency in this report. Notice that, upon opening the report, a new item appears in your Analysis pane. The report can be recalled from here at any time in the future. TIP: This report will benefit enormously from a View > By Item summary format.

Forecast Report The second choice from the drop-down list provides a timedependent overview of results.

Figure 12

104

Time-dependent Forecast report


Schlumberger


The columns in this report are self-explanatory, but the following are worth noting: •

The Forecast Oil Rate is based on the parameter used in the original forecast. This may have been a producing day rate, a calendar day rate, or any other user-defined rate. OFM has no way of knowing. However, OFM can always calculate a calendar day rate, because a forecast is based on volumes. This is why a calendar day rate is provided.

•

Forecast Cumulative Volume includes historical produced volumes.

•

Forecast Remaining Reserves is a 'countdown' number that reaches zero at the end of the forecast.

As with the automated forecast calculated variables, all the information contained in these two reports is contained in the file ForecastReports.xml, located in the OFM installation folder. The report details are easy to locate and edit. If you edit the file, to incorporate additional reports, for example, the changes show every time you clicked the Automated Forecast Reports icon on the Forecasts pane toolbar.


How would you set up OFM to forecast a new well, knowing that its early production rates were highly unstable?

•

Your field contains 9 normal wells and one well which must be produced intermittently. How do you obtain an accurate field forecast without needing to create and sum 10 individual well forecasts?

•

How would you ensure that a history match was obtained through only those data points captured when the production choke was fully open?


105


Schlumberger

Summary In this module, you learned about:

106

•

work with phases

•

apply additional history matching data reduction features

•

perform group forecasting

•

work with cases

•

extract forecast results

•

automate forecast calculated variables and reports.


Schlumberger


NOTES


107


Schlumberger

NOTES

108


Schlumberger

Ratio Forecasts

Module 4 Ratio Forecasts All the forecasting work you have done to this point has been time-based. There are other, equally valid methods of forecasting and the most commonly used alternative method is the Ratio forecast. Ratio forecasts operate on the basis that there is a well understood, and moderately predictable, relationship between the parameter under investigation – a water/oil ratio, for example and the recovered volumes (production) of oil or gas. This type of forecasting uses cumulative production.


work with two ratio forecasting models - Fluid Cut versus Cumulative Oil, and P/Z versus Cumulative Gas

•

combine the results of a Ratio forecast with a Primary Phase forecast to create a new phase forecast, the Calculated Forecast.

Lesson 1

Fluid Cut Forecast Analysis

Here, you perform a Water Cut versus Cumulative Oil analysis as an introduction to the concept of Ratio forecasting in OFM.

Exercise 1

Creating a Water Cut Forecast

To create a water cut forecast: 1. Open the current scenario dialog and, in the Flow Model dialog, choose the settings shown in the figure.


109

Ratio Forecasts

Schlumberger

TIP: Note the unit multipliers.

2. Click OK to accept the changes. 3. Choose completion HOGL1132 and the forecast displays. Forecast HOGL1132 125

100 Case Name : Base Case Slope : 1.1082 Intercept : 35.5072 Start WCT : 99.5055 % End WCT : 95 % Cum. Prod. : 62.8225 Mbbl Reserves : 0 Mbbl EUR : 62.8225 Mbbl

Water.Cut, %

75

50

25

0 0

15

30

45

60

75

Cumulative Oil, Mbbl

110


Schlumberger

Ratio Forecasts

Classical water cut development typically follows an ‘S’ shape. As cuts increase to higher values, the line tends to roll over, causing the subsequent projection to be much flatter. In this completion, you can see that the 'roll over' is well established. This means that your historical match should be restricted to the near-end region of the data. 4. Using the Limits > Range option (or any option you wish), set the following: • Min = 60 Mbbl • Max = 70 Mbbl By default, an OFM water cut forecast ends at 95%. In your completion, the current (last) data point is already 99.5%, so there is currently no forecast. To create a forecast, you must reset the end value. For now, assume it is possible to produce this completion to 99.9% water cut. 5. On the Properties pane, select Forecast > Start WCUT Option > From Fit. 6. On the Properties pane, select Forecast > End WCUT Option > Value. 7. Set the End WCUT to 99.9. 8. Zoom into the area of interest to see the effect.

9. Observe that a forecast has been created. The legend shows that if the completion produced to a 99.9% water cut, the completion would recover an additional 2.44 Mbbl of oil. 10. Save the forecast.


111

Ratio Forecasts

Schlumberger

Other Fluid Cuts The previous exercise can be used for any of these fluid cut analyses in OFM: •

GOR - (Gas Oil Ratio)

•

WOR - (Water Oil Ratio) NOTE: For WOR analysis, the default END value is 10 times the START value.

•

OCUT - (Oil Cut) TIP: While the definitions of water cut and water oil ratio are co-dependent, their forecast limits may be independent, for differing reasons. Typically, wellbore hydraulics (lift capacity) govern the water cut limit, while surface process design may provide a water oil ratio limit.

Accessing Automated Forecast Calculated Variables for Ratio Analysis Just as a family of forecast calculated variables are available for Phase analysis (oil, gas, and water), there is also a set of forecast calculated variables which pertain to Ratio analysis. To access the forecast calculated variables for Ratio analysis: 1. On the Forecasts pane, highlight a saved Ratio analysis. TIP: You may choose any entity for this.

112


Schlumberger

Ratio Forecasts

2. Click to open the Create Forecast Calculated Variables dialog.

3. Observe the new family of calculated variables. • Water Cut -The time-based water cut profile • Oil.EUR.FromWCut - Oil EUR estimate, from water cut forecast • Oil.Res.FromWCut - Oil Reserves estimate, from water cut forecast • Slope and Intercept - Properties of the regression fit line As before, you are free to modify the prefix and the name of each variable before saving it to the project. NOTE: The time based water cut profile requires a previously saved oil forecast. The saved oil forecast will be the source of time-based forecast cumulative oil, which then subsequently becomes the basis for the timebased forecast water cut.


113

Ratio Forecasts

Schlumberger

4. When done, click OK. The saved calculated variables are now in your project, correctly defined and attributed. You should take some time to review how they are constructed to better understand how to build your own.

Accessing Automated Forecast Reports from Ratio Analysis Just as a family of forecast reports is available for Phase analysis, there is also a set of forecast reports pertaining to Ratio analysis. To access the forecast calculated variables for Ratio analysis: 1. On the Forecasts pane, highlight a saved Ratio analysis, such as Water Cut. TIP: You may choose any entity for this. 2. Click the arrow next to

to display a list of reports.

There are two report formats. One format lists the ratio as a function of cumulative oil, while the other format lists the single value results of the analysis. Examples of both are shown in the figures.

114


Schlumberger

Lesson 2

Ratio Forecasts

Calculated Forecast

A very common question among OFM users has been: “I have saved an oil v time forecast, and I have saved a water cut v cum oil forecast. I need to generate a fluid forecast for this well. How do I combine these mutually independent profiles to obtain a water v time forecast?” In this scenario, oil and water cut data are readily available, but water data is not. For example, oil rates are measured directly, while water cut is the result of wellhead sampling. Another scenario may be one in which water data is available, but it is considered to be unreliable. In these situations, to obtain a water forecast might require a series of external manual calculations involving lookup tables, interpolations, and more. An expert user may know how to achieve the same result in OFM, but it would not be intuitive for a casual user. TIP: In fact, this method of water forecasting requires a very obscure form of a System Function.


115

Ratio Forecasts

Schlumberger

Exercise 1

Creating a Calculated Forecast

OFM 2009 introduced a new feature to allow all OFM users to overcome this obstacle. It is called the Calculated Forecast, and it is designed to directly answer the question posed in the first scenario. This feature is being introduced in the training at this time because the most common way of using it would not be possible until a Ratio analysis is completed. To create a calculated forecast: 1. Ensure completion HOGL1132 is selected. 2. Verify in the Forecasts pane that you have saved Oil and Water Cut forecasts for this completion. 3. Right-click on the graph and select Scenario to display the Edit Scenario dialog. 4. To generate a Water forecast, choose Phase = Water. 5. Ensure that the following variables are set: • Time (Date) = Date • Cum Water, Mbbl = CumWater * • Water Rate, bbl/d = Prd.PDWater* TIP: * You may need to create these variables first. It is best to use the corresponding oil variables CumOil and Prd.PDOil as guides for the definitions.

116


Schlumberger

Ratio Forecasts

6. Click the Calculated tab.

7. Because your selected phase is water, this dialog provides several options for calculating a Water forecast: The WATER phase is not Calculated from other phases • This is the default option. The forecast is done using water production data. The WATER phase is calculated from… • Forecast Phase _________ and Ratio Forecast _________ •

Oil is the only useable primary phase; Ratio forecasts are based on cum oil.

•

For Calculated Water, the available ratio options are WCUT, WOR and OCUT.

•

OFM calculates the water rate versus time from these two sources.


117

Ratio Forecasts

Schlumberger

• Forecast Phase ___________ multiplied by constant _________ •

You must supply a constant value. This is used to directly convert Oil forecast values to Water forecast values.

• Forecast Phase _________ multiplied by variable _________ •

A variable is chosen from the standard project variable list.

•

Note that the variable must evaluate at future dates; the calculated result is in the future.

• Calculate from variable __________ •

Choose a project variable to represent the Water forecast.

8. For your example, choose Forecast Phase and Ratio Forecast. The phase will be OIL, and the Ratio will be WCUT. 9. Click OK to accept the choices and perform the calculation.

118


Schlumberger

Ratio Forecasts

Make a note of several features: • The Calculated forecast is independent of the historical regression. • Decline results, such as b and qi, are not displayed on the legend, as they are not involved in the process. • The Forecast Ended By setting in the legend reads Calculated, confirmation of the process. 10. Save the forecast.

Lesson 3

P/Z Analysis

P/Z is the classical method for analyzing the performance and potential future of one or more gas wells. It is useful for several reasons: •

The mathematics behind the technique are very simple. P/Z analysis is a simplified form of material balance calculation. It states that for a single gas reservoir, unaffected by any aquifer, the parameter P/Z declines linearly as a function of gas volume recovered. This simple relationship allows for the calculation of future pressures, as well as the estimation of gas volume in place. Finally, operational constraints on pressure can be used in the analysis to yield recoverable volumes. In computational form:

•

The analysis is performed on a conventional Cartesian plot.

•

The technique assumes a straight line fit.

•

Deviation from a straight line may be indicative of an unknown reservoir drive mechanism.

•

It works best in a group analysis mode. The individual pressure performance of each well is displayed in relation to the production performance of the group.

•

It yields a conservative value for gas in place.

•

Operational constraints are easily introduced.


119

Ratio Forecasts

Schlumberger

The technique requires three parameters: •

Cumulative Gas production – This is usually readily available.

•

Pressure data - This is less common, but even single annual values or estimates are useful.

•

An estimate of the gas deviation factor, Z.

Gas Deviation Factor, Z To make use of the graphical forecasting technique, pressure values must be divided by the appropriate gas deviation factor, Z. The deviation factor is a PVT property and, just like other properties such as the formation volume factor, there are numerous ways in which it can be obtained. The most common ways to obtain the deviation factor are when: •

Z is a constant, so P/Z is a simple calculated variable similar to: PZ = / 0.995

•

Z is available from lab analysis data. Z is then the result of a Lookup function, and it can be included in a calculated variable.

•

Z varies with pressure, and across the field. OFM’s PVT functionality must be activated.

120


Schlumberger

Ratio Forecasts

Setting the Z Factor To set the Z factor using OFM PVT functionality: 1. Select Database > PVT. The PVT Entity Name dialog displays.

2. You should have at least one PVT profile (named Default) in your PVT Entity Name list. Choose an entity name and click Edit. The Edit PVT Information dialog displays. 3. Click the Gas Correlations tab. Select the Brill & Beggs correlation for Z Factor.


121

Ratio Forecasts

Schlumberger

4. Click the PVT Data tab.

5. In practice, you should fill in the appropriate data values for the correlation you have chosen. For training purposes, however, leave the default values in place and click OK. TIP: For more information about this, select Help > OFM Help > References > Gas Properties > Gas Deviation Factor. 6. Close the PVT Entity Name window. For the function to work, you need pressure data. Your instructor will provide a file named PZ Pressure.dat, which contains data for a small group of completions you will work with later. 7. Load the data file using the OFM data loader. Select Database > Import > Data Loader.

122


Schlumberger

Ratio Forecasts

8. Create a new calculated variable to return the result: PZ = Prs.Datum_prs / @PvtZ(Prs.Datum_prs) You are now ready to conduct the analysis.

Performing a P/Z Analysis To perform a P/Z analysis: 1. Using Filter by Completion, filter to the following list: • ACL208 • ACL607 • ACL618 • AEL1205 • BTEL619 • EOG205 • EOG206 2. Open a new forecast window. 3. Choose Phase / Analysis = PZ. 4. Set the following parameters and click OK. • Time (Date) = Date • Cum Gas, MMcf = CumGas • P/Z = PZ 5. Click to group the seven completions in the filter. The new plot is shown in the figure.


123

Ratio Forecasts

Schlumberger

Several features are evident in the display: • Two clusters have formed, labeled on the graph as High Line and Low Line. Clearly, you are looking at more than one compartment. • The forecast currently attempts to use all the data. • Cumulative gas is the aggregated cumulative for all seven completions.

124


Schlumberger

Exercise 1

Ratio Forecasts

Task 8: Forecasting P/Z for Single Compartment

Your first action under this task is to remove the completions which do not form part of this compartment. Completion ACL607 is the best candidate. 1. Hold down the Ctrl key and click on completion ACL607 in the Step pane.

2. Release the Ctrl key. As you are in group mode, OFM automatically regroups the remaining six completions and adjusts the plot accordingly. It now seems clear you are dealing with two compartments, and you will remove the completions lying on the 'Low Line.' Judging by the legend, these completions are: • ACL208 • AEL1205 • EOG205 • EOG206 3. Repeat steps 1 and 2, and remove the four completions listed above. Your first analysis now consists of two completions. All P/Z analysis options are in the Properties pane.


125

Ratio Forecasts

Schlumberger

Before beginning the analysis, observe the results legend. Gas reserves generated by the default forecast are 189.888 MMcf, while expected ultimate recovery (EUR) is 237.492 MMcf. This figure is identical to original gas in place (OGIP) because you expect to recover all of the gas. In practice, of course, this will not happen. Operational constraints, such as economic and mechanical issues, prevent complete recovery. You can, however, introduce the constraint into your analysis as an End P/Z. 4. From the Properties pane, select Forecast > End P/Z Option > Value. 5. Enter an End P/Z value of 200 psi. It is important to note that this is a value for P/Z, and not just for P. 6. Observe the legend now. You changed what you expect to recover, but you have not changed what was originally in the reservoir. Several conditions, however, are different: • The forecast has been truncated. • Reserves have dropped to 181.6 MMcf. • EUR has dropped to 229.2 MMcf. • The OGIP value remains the same at 237.496 MMcf. 7. Save the forecast. Note that it is a group forecast and, therefore, a name is required.

Manual Fitting Manual fitting is also possible in P/Z analysis, and there are two ways to do this. You can either drag the blue control points at the ends of the fit line or drag the line itself. Regardless, your results will automatically update. When you begin using either of these options, the History Matching method in the Properties pane changes from Auto to Manual, and remains set to Manual until you reselect Auto.

126


Schlumberger

Ratio Forecasts


How is it possible to use the result of a Water Cut analysis to improve an Oil forecast?

•

What is the purpose of a P/Z forecast?


performing a Ratio analysis

•

creating a Calculated forecast

•

using automated calculated variables and reports

•

performing a P/Z analysis.

You have also successfully completed Tasks 1through 8 in the Provo field workflow, as assigned to you at the beginning of the course. You should have obtained a good impression of the power and flexibility offered by OFM. The final lesson to take away is that OFM is a blank canvas; it adapts to your needs and requirements.


127

Ratio Forecasts

Schlumberger

NOTES

128


OFM 2009 Forecast Analysis Fundamentals H A5

Recommend Documents