A
Project Report
Submitted in partial fulfillment of the
Requirements for the award of the degree of
BACHELOR OF TECHNOLOGY
In
COMPUTER SCIENCE & ENGINEERING
By
Name
Abhishek Shukla
Anjali Yadav
Anuj Awasthi
Anurag Kanwar
Roll No.
R970213004
R970213011
R970213012
R970213013
CANDIDATES DECLARATION
I hereby certify that the project work entitled IBM PREDICTIVE PRODUCTION DATA MODELS AND
FORECASTING in partial fulfillment of the requirements for the award of the Degree of BACHELOR OF
TECHNOLOGY in COMPUTER SCIENCE AND ENGINEERING with specialization in oil and gas informatics
and submitted to the Department of Computer Science & Engineering at Center for Information
Technology, University of Petroleum & Energy Studies, Dehradun, is an authentic record of my/ our work
carried out during a period from MAY, 2016 to JULY, 2016 under the supervision of Dr.Kingshuk
Srivastava, Assistant Professor UPES.
The matter presented in this project has been submitted by me/ us for the award of any other degree of
this or any other University.
Name
Abhishek Shukla
Anjali Yadav
Anuj Awasthi
Anurag Kanwar
Roll No.
R970213004
R970213011
R970213012
R970213013
This is to certify that the above statement made by the candidate is correct to the best of my
knowledge.
Date: 6th August, 2016
Dr. M. Venkatadri
Program Head - CSE- Oil and Gas Informatics
Center for Information Technology
University of Petroleum & Energy Studies
Dehradun 248007 (Uttarakhand)
ACKNOWLEDGEMENT
We wish to express our deep gratitude to our mentor Mr. Mohammad Siddique and comentor Dr. Kingshuk Srivastava, for all advice, encouragement and constant support he has
given us throughout our project work. This work would not have been possible without his
support and valuable suggestions.
We sincerely thank to our respected Program Head of the Department, Dr. M. Venkatadri,
for his great support in doing our project.
We are also grateful to Dr. Manish Prateek, Associate Dean and Dr. Shrihari, Dean CoES,
UPES for giving us the necessary facilities to carry out our project work successfully.
We would like to thank all our friends for their help and constructive criticism during our
project work. Finally we have no words to express our sincere gratitude to our parents who
have shown us this world and for every support they have given us.
TABLE OF CONTENTS
Abstract
1. Introduction
1.1. Reservoir Prediction Techniques
2. Objective
3. Literature Review
3.1. Volumetric Method
3.2. Material Balance Equation
3.2.1. Applications of Material Balance
3.2.2. Limitations of Material Balance
3.3. Decline Curve Analysis
3.3.1. Exponential Decline
3.3.2. Hyperbolic Decline
3.3.3. Harmonic Decline
3.3.4. Advantages and Decline Curve Analysis
3.4. Reservoir Simulation
3.4.1. Mathematical Model
3.4.2. Numerical Model
3.4.3. Computer Model
4. Materials and Methodology
4.1. Case Study
5. Results and Discussion
6. Conclusions
7. References
7
8
8
9
10
10
13
16
16
17
18
19
20
20
21
22
23
25
27
34
35
36
List of Figures
1.
2.
3.
4.
5.
6.
List of Tables
1.
2.
3.
4.
5.
Abstract
1. INTRODUCTION
Reservoir Prediction techniques: Methods of forecasting reservoir can be divided into three categories: analogical
methods, experimental methods and mathematical methods. Analogical methods
use properties of mature reservoirs that are either geographically or
petrophysically similar to the target reservoir to attempt to predict reservoir
performance of the target zone. Experimental methods measure physical
properties (such as rates, pressures, or saturations) in laboratory models and
scale these results to the entire hydrocarbon accumulation. Mathematical
methods on the other hand use equations to predict reservoir performance.
Following techniques have been worked upon in our project:
Volumetric Method
Material Balance Equation
Decline Curve Analysis
Reservoir Simulation
2. Objective
The objective of this research project is to investigate fundamental properties and
behavior of crude oil production and examine some model approaches for
creating realistic outlooks for the future.
It also deals with the comparative study of the different techniques
used in the market for forecasting the well.
3. Literature Review
3.1 Volumetric Method
Knowing the volume of hydrocarbons in place in a reservoir is of fundamental
importance. Any development project on a field obviously depends on the oil
and/or gas in place in the reservoir rock.
The assessment of the accumulations in initial stages is difficult because of the
complexity of the porous medium, uncertainty as to the exact shape of the
reservoir and, little petrophysical data. Therefore, difficulty resides in calculating
all the parameters which characterize the volume of hydrocarbons in place.
Volumetric method states that we can determine the volume in reservoir
conditions using following mathematical operation:
Volumetric method is the most used technique which is almost always used and
plays the pivotal role in early stages of a field, after incorporating various factors,
the final equation for Original oil in place becomes:
Also, the oil produced by various primary mechanism drives can also be calculated
using volumetric method because the oil recovered will be equal to oil originally
in place, minus the oil remaining in the reservoir abandonment. Although these
calculations are primitive and may not be highly accurate but they play an
important role as the future data generated has a reference point, on which more
accurate predictions can be made.
Oil recovered by primary producing means in a solution gas drive is given by:
Np = 7758Ah
)---------------(3)
Note that the pressure at which the reservoir will be abandoned will depend on
economic factors, the value of oil, the operating costs, and the oil producing rate.
Np=7758 Ah[
] ------------(4)
So far we have discussed equations concerning oil only, the same can be
extended for gas after making some changes in the basic equations, out of which
the primary one is the change in mathematical factor, it changes from 7758 for oil
to 43560 for gases. This leads to the following volumetric equation for gases:
G=43,560 Ah [
] --------(5)
From the Tank Model, we can comprehend that: Pore Volume occupied by the gas at initial pressure, pi
+
Pore Volume occupied by the oil at initial pressure, pi
=
Pore Volume occupied by the remaining oil at pressure, p
+
Pore volume occupied by the gas in the gas cap at p
+
Pore volume occupied by the evolved solution gas at p
+
Pore volume occupied by the net water influx at p
+
Change in pore volume due to connate water expansion and pore
volume reduction due to rock expansion
+
Pore volume occupied by the injected gas at p
+
Pore volume occupied by the injected water at p
From these Nine Equations and assuming that no Gas Cap is formed we can
develop a general formula: -
The general form of the equation can be described as net withdrawal (withdraw injection) = expansion of the hydrocarbon fluids in the system + cumulative water
influx.
3.3.1 Exponential Decline: As mentioned above, in the exponential decline, the wells production data plots
as a straight line on a semi-log paper. The equation of the straight line on the
semi-log paper is given by:
q=qie
Dt
Where:
q = wells production rate at time t, STB/day
qi = wells production rate at time 0, STB/day
D = nominal exponential decline rate, 1/day
t = time, day
Equation
Rate
q=qie-Dt
Np=(qi q)/D
D=-ln(1-De)
De=(qi-q)/qi
De=1-e-D
Life
t={ln(qi/q)}/D
Table 1
3.3.2 Hyperbolic Decline: Alternatively, if the wells production data plotted on a semi-log paper concaves
upward, and then it is modeled with a hyperbolic decline. The equation of the
hyperbolic decline is given by:
1/b
q= qi (1 + bDit )
Where:
q = wells production rate at time t, STB/day
qi = wells production rate at time 0, STB/day
Di = initial nominal exponential decline rate (t = 0), 1/day
b = hyperbolic exponent
t = time, day
The following table summarizes the equations used in hyperbolic decline:
Hyperbolic Decline b>0, b1
Description
Equation
1/b
Rate
q= qi (1 + bDit )
Np=qib/Di(1-b) x (qi1-b-q1-b)
Dei=(qi-q)/qi
De=1-e-D
Life
t={(qi/q)b-1}/bDi
Table 2
3.3.3 Harmonic Decline: A special case of the hyperbolic decline is known as harmonic decline, where b
is taken to be equal to 1.
The following table summarizes the equations used in harmonic decline:
Harmonic Decline b=1
Description
Equation
Rate
q=qi/(1+bDit)
Np=qi/Di x ln(qi/q)
Di=Dei/(1-Dei)
Dei=(qi-q)/qi
Life
t={(qi/q)-1}/Di
Table 3
Finite Difference
One way to solve a Partial differential equation is to convert the PDE to finitedifference form. The finite-difference form is obtained by replacing the
derivatives in the PDE with differences that are obtained from Taylors series.
To illustrate the procedure, let us suppose that we know the function f(x) at two
discrete points x = xi and x = xi + x, where x is an increment along the x-axis.
We can approximate the derivative df(x)/dx at x = xi by solving the Taylors series,
.........(3)
for df(x)/dx.
The result is
....................(4)
where ET is the term
....................(5)
If we neglect ET, we obtain the finite-difference approximation of the first
derivative.
....................(6)
Eq. 6 is an approximation because we are neglecting all of the terms in ET, which is
called the truncation error. In the limit as the increment x approaches zero, the
truncation error approaches zero, and the finite difference approaches the
definition of the derivative.
....................(7)
....................(8)
Eqs. 6 through 8 are all derived from Taylors series.
Black-oil
Used for basic fluid flow, it treats oil and gas as single component
Compositional
It treats oil and gas as different components, but it computationally expensive
Thermal
It is used for hot water, steam flooding, and insitu combustion
Chemical
Used for modeling operations such as polymer injection
Miscible
It can model solvent injection and similar operations
Single-porosity or dual-porosity
Used for different porosity type reservoirs, such as single porosity system and
fractured reservoir
and more.
As per the requirements and objectives, the simulator is chosen and then
different options are simulated, and the best technique is chosen for the
operation.
It is desired to estimate Initial oil in place in order to determine whether this field
should be developed and if it will be economically feasible to produce this field.
To calculate this, we use volumetric method to estimate ioip using porosity, initial
water saturation and initial oil formation volume factor.
As the asset seems economically lucrative, ONGC goes ahead to develop the field
and drills a number of wells across the field. 6 years down the line, more amount
of data has been gathered which is more accurate, to revise the production
objectives reservoir engineers decide to recalculate initial oil in place using
material balance equation. Following is the data which is present at hand.
Petrophysical Properties
Current reservoir
condition
P, psi
3000
2500
Bo, bbl/STB
1.35
1.33
Rs, scf/STB
600
500
Np, MMSTB
7170
361
Gp, MMMSCF
5.5
Bw, bbl/STB
1.00
We, MMbbl
Wp, MMbbl
0.2
Bg, bbl/scf
0.0011
0.0015
q, STB/d
Cw, Cf
Table 4
Additional information:
Volume of Bulk oil zone: 100,000 ac-ft
Volume of bulk gas zone: 20,000 ac-ft
To calculate initial oil in place, we require m which is ratio intial gas cap volume
to intial oil volume bbl/bbl. As, the porosity and water saturation for both are
equal, m will be given as:
m = Volume of bulk gas zone/ Volume of bulk oil zone
= 20,000/100,000
=0.2
Cumulative gas oil ratio is also required for the calculations, which is given as the
ratio of gas produced and oil produced in duration of 6 years.
Rp= 5.5 x 109 / 5 x 106
=1100 scf/STB
Therefore, using material balance equation, initial oil in place is found to be 31.14
MMSTB which is 6.93 MMSTB (18.2%) less than that calculated using volumetric
method.
Since the production is decreasing, therefore it is necessary to have an idea of
future performance of field so that suitable measures can be planned and taken
accordingly.
To do that, Decline Curve Analysis is being employed which will tell the
production rate for last five years as well as predict that of the future.
As we extrapolate the data, we see a steady decline in flow rate per day which
becomes uneconomical at certain point in time and estimating uptil 13.5 years,
the daily flow rate comes down to as low as 9STB/d and in these 7.5 years the
production has increased from 5004938 to 5265485 STB which is only 5.20%.
This cumulative production is very low and gives a recovery factor of 16.9%, that
too some of it uneconomically.
So, depending upon the operating expenses, the drive mechanism of reservoir
and various other characteristics of reservoir, suitable steps such as enhanced oil
recovery, usage of artificial lifts and other methods must be adopted when age of
field has reached around 7 years.
To obtain better understanding of future predictions, team decides to take help of
Mbal PETEX software to predict cumulative production in future.
As it can be seen from the above plot, it shows that the cumulative production
will be 5.790951 MMSTB after 6 years from the current scenario.
Now, the same timeline cumulative production results obtained from Decline
curve are a little different i.e. 5.265485 MMSTB.
Technique
Results
Volumetric Method
Material Balance
Mbal PETEX
8. References
1. Mbal handbook.
2. Petroobject.com.
3. Permian Basin Oil and Gas Recovery Conference, Midland, Texas, 16-18
March 1994. SPE-27684-MS. http://dx.doi.org/10.2118/27684-MS
4. Predictive analytics the rise and value of predictive analytics in enterprise
decision making. CGI group INC., 2013
5. Predictive Analytics 101: Next-Generation Big Data Intelligence, Intel IT
center, March 2013.
6. Tarek Ahmad. Reservoir Engineering Handbook.
7. Walsh, M.P. 1995. A Generalized Approach to Reservoir Material Balance
Calculations.
J
Can
Pet
Technol
34
(1).
PETSOC-95-0107. http://dx.doi.org/10.2118/95-01-07
8. Walsh, M.P., Ansah, J., and Raghavan, R. 1994. The New, Generalized
Material Balance as an Equation of a Straight Line: Part 1 - Applications to
Undersaturated, Volumetric Reservoirs.