26 vues

Transféré par DanAlexandruJoita

dsafdsgfhg

dsafdsgfhg

© All Rights Reserved

- What Are Opportunity Crude Oils
- Kuwait Oil and Gas Sector_Oil & Gas Directory ME_2010
- Chapter15 - Economics
- PM1 Plant Maintenance
- Oil Refining Industry in South East Asia, 2011
- Grace Fact Sheet.
- MB0048 Operations Research
- FullModuleLectures.pdf
- CATPAPER.pdf
- Solving Difficult Multicommodity Problems (PPRN, CPLEX, EMNET, Castro, 2003)
- Crude Oil Refining
- The Significance of Reducing Setup Times_setup Costs
- Syllabus for New or Modified Courses
- 2012 - Weight Minimization of Frames With SGA and PGA
- Lingo_15_Users_Manual.pdf
- QP Null Space Method
- PetroPowerChile
- APCOM_Haulage Impact on Mine Cut-Off Grade Strategy
- 1-s2.0-S0098135411001001-main
- 01 Introduction

Vous êtes sur la page 1sur 192

Cassio R. Tamara

Optimisation

School of Engineering

Department of Process & Systems

Engineering

MSc Thesis

Cranfield UNIVERSITY

School of Engineering

Department of Process & System Engineering

MSc Thesis

Academic Year 2002-2003

Cassio R. Tamara

Optimisation

Supervisor: Dr. Zhigang Shang

September 2003

This thesis is submitted in partial fulfilment of the requirements for the

Degree of Master of Science

Cranfield University, 2003. All rights reserved. No part of this publication may be

reproduced without the written permission of the copyright owner

_____________________________________________________________________

ii

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

ABSTRACT

Nowadays, the development of global competition has been one of the main factors

that have driven the efforts toward the optimisation development. Therefore, oil

refineries have been encouraged to be restructured for competing successfully in this

new scenario with low profit margin, tighter environmental regulations and more

efficient plant operation. However, many years and a lot of human and computational

efforts have been dedicated to improve the techniques applied for the overall refinery

optimisation. Good developments have come successfully operating at the planning

level; but developing and solving rigorous overall plant optimisation models at the

production scheduling level still are at research stage and much more work must be

done to continue improving in this field through the involvement of difficult tasks due

to the mathematical complexity of the models which have the compulsory use of a

large quantity of equations and variables that hugely increase the size of the problem.

This Thesis presents a new generic mixed integer linear programming model for

optimising the scheduling of crude oil unloading, inventories, blending and feed to oil

refineries that usually unload several kinds of crude oils with different compositions.

The objective function of the model consists on minimising the operational cost

generated during the mentioned operation. Case studies are presented and compared

each other illustrating the capabilities of the model to solve operation scheduling

problems in this area and to support future expansion projects for the system as they

happen in real situations. The solution involves optimal operation of crude oil

unloading, optimal transfer rates among equipments in accordance with the pumping

capacities and tank volume limitations, optimal oscillation of crude oil blended

compositions and fulfilment of the oil charging demand per process unit.

ACKNOWLEDGEMENT

In memory of my father.

I still remember when I left my country far away from here, with my wife and my two little

children to come to Cranfield, full of hopes and wishes, with the only purpose to study and fulfil

my major dream attending and completing post grade studies abroad in a well known university

as Cranfield is known. It was a difficult decision to come to the UK with limited resources in the

middle of my professional productive career, with more than 15 years of continuous work in an

oil refinery from Ecopetrol.

I would like to give my thanks to my wife Norma, my daughter Natalia and my son Sebastian for

staying with me during the long months living and sharing together busy and happy moments,

as well as uncertain times. Moreover, I really appreciate the constant prayers of my mother and

my wife during all our stay here.

Furthermore, I would like to give my thanks to Colfuturo which gave me the financial support for

my tuition fees and part of my living expenses. On the other hand, despite the policies and the

difficult moments encountered by Ecopetrol and my country, I really appreciate the tireless and

willing support from the oil refinery manager Mr. Antonio Escalante searching for economical

approval to cover the financial support given by Colfuturo . Thanks also to Mrs Martha

Espinosa, Mr Carlos Bustillo, Mr Jorge Villalba and other important people at the high staff and

board level that were supporting the process.

Finally, I really appreciate the advice and support from Dr. Zhigang Shang for encouraging me

to improve the quality and value of my Thesis and his high motivation towards my research.

Furthermore, I would also like to give my thanks to Professor Mike Sanderson, Mrs Linda

Withfield and Mrs Janet Dare for all their kind support and help from the Process and System

Engineering Department during this very important year of my life. From Mr Ivor Rhodes, I

thank him for patiently keeping me on the waiting list to come to Cranfield for three years.

Cassio Tamara, Cranfield, 28th August 2003

_____________________________________________________________________

ii

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

TABLE OF CONTENTS

Chapter 1. INTRODUCTION.......................................................................................1

1.1 Oil Refinery Optimisation....................................................................................1

1.2 Other facts and Developments about Oil Refinery Optimisation. .......................3

1.3 Present Work........................................................................................................6

Chapter 2. OVERVIEW OF SCHEDULING OPTIMISATION ..................................8

2.1 Scheduling General Concern. ..............................................................................8

2.2 Crude Oil Inventory Scheduling Optimisation. ...................................................9

2.3 Review of Production Scheduling Developments .............................................10

Chapter 3. PRODUCTION SCHEDULING PROBLEM DEFINITION. ...................13

3.1 Problem Definition.............................................................................................13

3.2 Optimisation Model Formulation Introduction..................................................16

3.3 Model Mathematical Formulation. ....................................................................21

3.3.1 Vessel Arrival and Departure Operation Rules. .........................................22

3.3.2 Material Balance Equations for the Vessel. ................................................24

3.3.3 Material Balance Equations for the Storage Tank. .....................................25

3.3.4 Material Balance Equations for the Charging Tank. ..................................27

3.3.5 Material Balance Equations for Component k in the Storage Tank. ..........28

3.3.6 Material Balance Equations for Component k in the Charging Tank.........29

3.3.7 Operating Rules for Crude Oil Charging to Crude Distillation Units. .......30

3.3.8 Problem Solving Direction. ........................................................................30

3.4 Chapter Summary. ............................................................................................33

Chapter 4. MODEL REAL TEST................................................................................35

4.1 Introduction........................................................................................................36

4.2 Example 1 Results and Analysis........................................................................37

4.3 Example 2 Results and Analysis........................................................................42

4.4 Example 3 Results and Analysis........................................................................49

4.5 Chapter Summary ..............................................................................................55

Chapter 5. MODEL APPLICATION TO A PRACTICAL CASE..............................57

5.1 Presentation of Case 1........................................................................................58

5.2 Scheduling Alterations of Case 1.......................................................................66

5.3 Bigger Scheduling Alteration of Case 1. ...........................................................72

5.4 Effect of Not Control Oil Sulphur Content Applied to Case 1. .........................78

5.5 Effect of Not having Tank Inventory Cost for Case 1. ......................................81

5.6 Effect Caused by a High Reduction in Changeover Cost for Case 1.................82

5.8 Chapter Summary. .............................................................................................82

Chapter 6. MODEL APPLICATION TO CASES WHERE THE SYSTEM

EQUIPMENT LIMITATIONS ARE HIGHLIGHTED. .............................................84

6.1 Case 2, Effect of Increasing the Vessel Unloading Volume..............................84

6.2 Scheduling Alteration of Case 2. .......................................................................91

6.3 Case 3, Effect of a High Increase in Unloading Cost. .......................................95

6.4 Case 4, Effect of CDU Shut Down. ...................................................................98

6.5 Chapter Summary. ...........................................................................................101

Chapter 7. MODEL APPLICATION TO CASES WHERE THINKING EITHER IN

REVAMP OR CHANGE EQUIPMENT IS POSSIBLE...........................................103

7.1 Cases 5 and 6, Increasing Capacity of Charging Tanks...................................103

7.2 Case 7, Effect of Changing Pumping Rate Capacity. ......................................112

_____________________________________________________________________

iii

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

7.4 Chapter Summary. ...........................................................................................120

Chapter 8. CONCLUSIONS AND FUTURE WORKS ............................................122

8.1 Conclusions......................................................................................................122

8.2 Future Works. ..................................................................................................123

BIBLIOGRAFY.........................................................................................................125

APENDIX A ..............................................................................................................127

A.1 Example 1 Computer hardcopy output ...........................................................127

A.2 Example 2 Computer hardcopy output ...........................................................131

A.3 Example 3 Computer hardcopy output ...........................................................137

A.4 Case 1 Computer hardcopy output..................................................................143

A.5 Case 2 Computer hardcopy output..................................................................150

A.6 Case 3 Computer hardcopy output.................................................................156

A.7 Case 6 Computer hardcopy output..................................................................163

APENDIX B ..............................................................................................................170

_____________________________________________________________________

iv

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

LIST OF FIGURES

Figure 1.1- The cycle of refining operations. ................................................................3

Figure 1.2 - Optimisation Efforts for improving profitability .....................................5

Figure 3.1 - Problem Representation. ..........................................................................13

Figure 4.1 Oil flow network for example 1. .............................................................38

Figure 4.2 Example 1. Storage tank optimal volume variation. ...............................40

Figure 4.3 Example 1. Charging tank optimal volume variation..............................40

Figure 4.4 Example 1. Optimal feeding to the crude distillation unit. Solid bars:

Blended oil X ; Blank bar: Blended oil Y............................................................41

Figure 4.5 Example 1. Optimal component concentration variation in charging

tanks. ....................................................................................................................41

Figure 4.6 Example 1. Optimal results from the existing model (right side). ..........42

Figure 4.7 Oil flow network for example 2. .............................................................43

Figure 4.8 Example 2. Storage tank optimal volume variation. ...............................46

Figure 4.9 Example 2. Charging tank optimal volume variation.............................46

Figure 4.10 Example 2. Optimal feeding to the CDU 1. Dashed bars: blended oil 2;

solid bars: blended oil 3. ......................................................................................47

Figure 4.11 Example 2. Optimal feeding to the CDU 2. Blank bars: blended oi1 1;

Dashed bars: blended oil 2; solid bars: blended oil 3. .........................................47

Figure 4.12 Example 2. Optimal concentration variation of component 1 in charging

tanks. ....................................................................................................................48

Figure 4.13 Example 2. Optimal concentration variation of component 2 in charging

tanks. ....................................................................................................................48

Figure 4.14 Example 2. Optimal results from the existing model. ..........................49

Figure 4.15 Oil flow network for example 3. ...........................................................50

Figure 4.16 - Example 3. Storage tank optimal volume variation. ..............................53

Figure 4.17 - Example 3. Charging tank optimal volume variation. ..........................53

Figure 4.18 - Example 3. Optimal feeding to the CDU 1. Dashed bars: blended oil 2;

solid bars: blended oil 3. ......................................................................................54

Figure 4.19 - Example 3. Optimal feeding to the CDU 2. Blank bars: blended oil 1;

solid bars: blended oil 3. ......................................................................................54

Figure 4.20 Example 3. Optimal results from the existing model. ...........................55

Figure 5.1 Oil Flow Next Work for the Oil Refinery. ..............................................57

Figure 5.2 - Case1. Storage tank volume variation results. .........................................62

Figure 5.3 - Case1. Charging tank volume variation results........................................62

Figure 5.4 - Case 1. Feeding to the crude distillation unit results. Solid bars: blended

oil 1; blank bars: blended oil 2.............................................................................63

Figure 5.5 - Case 1. Crude oil sulphur content variation results in charging tanks. ....65

Figure 5.6 - Case1a. Storage tank volume variation results. .......................................68

Figure 5.7 - Case1a. Charging tank volume variation results......................................68

Figure 5.8 - Case1a .Crude oil sulphur content variation results in charging tanks. ...69

Figure 5.9 - Case1b. Storage tank volume variation results. .......................................71

Figure 5.10 - Case1b. Charging tank volume variation results....................................71

Figure 5.11 - Case1b. Crude oil sulphur content variation results in charging tanks. .72

Figure 5.12 - Case1c. Storage tank volume variation results. .....................................74

Figure 5.13 - Case1c. Charging tank volume variation results....................................75

_____________________________________________________________________

v

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Figure 5.14 - Case 1c. Feeding to the crude distillation unit results. Solid bars:

blended oil 1; blank bars: blended oil 2. ..............................................................75

Figure 5.15 - Case1c. Crude oil sulphur content variation results in charging tanks. .76

Figure 5.16 - Case 1d. Storage tank volume variation results. ....................................79

Figure 5.17 - Case1d. Charging tank volume variation results....................................80

Figure 6.1 - Case 2. Storage tank volume variation results. ........................................87

Figure 6.2 - Case 2. Charging tank volume variation results.......................................88

Figure 6.3 - Case 2. Feeding to the crude distillation unit results. Solid bars: blended

oil 1; blank bars: blended oil 2.............................................................................89

Figure 6.4 - Case 2. Crude oil sulphur content variation results in charging tanks. ....91

Figure 6.5 - Case 2a. Storage tank volume variation results. ......................................92

Figure 6.6 - Case 2a. Charging tank volume variation results.....................................93

Figure 6.7 - Case 2a. Feeding to the crude distillation unit. Solid bars: blended oil 1;

blank bars: blended oil 2. .....................................................................................94

Figure 6.8 - Case 2a. Crude oil sulphur content variation results in charging tanks. ..95

Figure 6.9 - Case 3. Storage tank volume variation results. ........................................96

Figure 6.10 - Case 3. Charging tank volume variation. ...............................................97

Figure 6.11 - Case 3. Feeding to the crude distillation unit. Solid bars: blended oil 1;

blank bars: blended oil 2. .....................................................................................97

Figure 6.12 - Case 4. Storage tank volume variation results. ......................................99

Figure 6.13 - Case 4. Charging tank volume variation results...................................100

Figure 6.14 - Case 4. Feeding to the crude distillation unit. Solid bars: blended oil 1;

blank bars: blended oil 2. ...................................................................................100

Figure 7.1 - Case 5. Storage tank volume variation results. ......................................104

Figure 7.2 - Case 6. Storage tank volume variation results. ......................................105

Figure 7.3 - Case 5. Charging tank volume variation results.....................................105

Figure 7.4 - Case 6. Charging tank volume variation results.....................................106

Figure 7.5 - Case 5. Feeding to the crude distillation unit results. Solid bars: blended

oil 1; blank bars: blended oil 2...........................................................................107

Figure 7.6 - Case 6. Feeding to the crude distillation unit results. Solid bars: blended

oil 1; blank bars: blended oil 2...........................................................................107

Figure 7.7 - Case 5. Crude oil sulphur content variation in charging tanks. .............110

Figure 7.8 - Case 6. Crude oil sulphur content variation in charging tanks. .............110

Figure 7.9 - Case 7c. Storage tank volume variation results. ....................................114

Figure 7.10 - Case 7c. Charging tank volume variation. ...........................................114

Figure 7.11 - Case 8. Storage tank volume variation results. ....................................117

Figure 7.12 - Case 8. Charging tank volume variation results...................................118

Figure 7.13 - Case 8. Feeding to the crude distillation unit results. Solid bars: blended

oil 1; blank bars: blended oil 2...........................................................................118

_____________________________________________________________________

vi

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

LIST OF TABLES

Table 4.1 - Comparisons of optimal results with an existing model. ..........................36

Table 4.2 - System information for Example 1............................................................39

Table 4.3 Example 1. Optimal unloading starting date for vessels. .........................39

Table 4.4 - System information for Example 2............................................................44

Table 4.5 Example 2. Optimal unloading starting date for vessels. .........................45

Table 4.6 - System information for Example 3............................................................51

Table 4.7 Example 3. Optimal unloading starting date for vessels. .........................51

Table 5.1 - Case 1. Input data summary. ....................................................................60

Table 5.2 - Case 1. Unloading start and finish results. ................................................61

Table 5.3 - Case 1.Vessel volume variation (bbl x 1,000) results. ..............................61

Table 5.4 - Case 1. Results for volumetric flow rate (bbl x 1,000) per day from vessels

to storage tanks. ...................................................................................................64

Table 5.5 - Case 1. Results for volumetric flow rate (bbl x 1,000) per day from storage

tanks to charging tanks.........................................................................................64

Table 5.6 - Case 1a. First change of arriving schedule. ...............................................66

Table 5.7 - Case 1a. Unloading start and finish results. ..............................................67

Table 5.8 - Case 1b. Second change of the arriving schedule. ....................................69

Table 5.9 - Case1b. Unloading start and finish results. ...............................................70

Table 5.10 - Case 1c. Bigger change of the arriving schedule. ...................................73

Table 5.11 - Case 1c. Unloading start and finish results. ............................................74

Table 5.12 - Case 1c. Vessel volume variation (bbl x 1000) results. ..........................74

Table 5.13 - Case 1c. Results for Volumetric flow rate (bbl x 1,000) per day from

storage to charging tanks. ....................................................................................76

Table 5.14 - Case 1c. Results & comparisons among different crude oil volumes for

storage tank 1. ......................................................................................................77

Table 5.15 - Case 1d. Unloading start up and finish results. .......................................79

Table 5.15 - Case 1d. Results for volumetric flow rate in bbl x 1000 per day from

storage to charging tanks. ....................................................................................80

Table 5.16 - Case 1e. Results for volumetric flow rate (bbl x 1,000) per day from

vessels to storage tanks. .......................................................................................81

Table 5.17 Resource summary used to solves cases.................................................82

Table 6.1 - Case 2. Input data summary. ....................................................................85

Table 6.2 - Case 2.Unloading start and finish results. .................................................86

Table 6.3 - Case 2. Vessel volume variation (bbl x 1000) results. ..............................86

Table 6.4 - Case 2. Results for volumetric flow rate (bbl x 1,000) per day from

vessels to storage tanks. .......................................................................................89

Table 6.5a - Case 2. Results for volumetric flow rate (bbl x 1,000) per day from

storage to charging tanks. ....................................................................................90

Table 6.5b - Case 2. Results for volumetric flow rate (bbl x 1,000) per day from

storage to charging tanks. ....................................................................................90

Table 6.6 - Case 2a. First change of arriving schedule. ...............................................91

Table 6.7 - Case 2a. Unloading start and finish results. ..............................................92

Table 6.8 - Case 2a. Results for volumetric flow rate (bbl x 1000) per day from

vessels to storage tanks. .......................................................................................94

Table 6.9 - Case 3. Unloading start and finish results. ................................................96

_____________________________________________________________________

vii

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Table 6.10 - Case 3. Results for volumetric flow rate (bbl x 1000) per day from

vessels to storage tanks. .......................................................................................98

Table 6.11 - Case 4. Unloading start and finish results. ..............................................99

Table 6.12 Resource summary used to solves cases...............................................101

Table 7.1 - Cases 5 and 6. Unloading start and finish results....................................104

Table 7.2 - Case 5 and 6. Volumetric flow rate in bbl x 1,000 per day from vessels to

storage tanks.......................................................................................................108

Table 7.3a - Case 5. Results for volumetric flow rate (bbl x 1,000) per day from

storage to charging tanks. ..................................................................................108

Table 7.3b - Case 5. Results for volumetric flow rate (bbl x 1,000) per day from

storage tanks to charging tanks ..........................................................................109

Table 7.4a - Case 6. Results for volumetric flow rate (bbl x 1000) per day from

storage to charging tanks. ..................................................................................109

Table 7.4b - Case 6. Results for volumetric flow rate (bbl x 1,000) per day from

storage to charging tanks. ..................................................................................109

Table 7.5 - Case 6. Comparisons / different charging tank capacities.......................112

Table 7.6 - Case 7. Input data summary exception....................................................112

Table 7.7a - Case 7c. Results for volumetric flow rate in bbl x 1,000 per day from

storage to charging tanks. ..................................................................................115

Table 7.7b - Case 7c. Results for volumetric flow rate in bbl x 1,000 per day from

storage to charging tanks. ..................................................................................115

Table 7.8 - Case 8. Unloading start and finish results. ..............................................117

Table 7.9 - Case 8. Results for volumetric flow rate (bbl x 1,000) per day from vessels

to storage tanks. .................................................................................................119

Table 7.10a - Case 8. Results for volumetric flow rate (bbl x 1,000) per day from

storage to charging tanks. ..................................................................................119

Table 7.10b - Case 8. Results for volumetric flow rate (bbl x 1,000) per day from

storage to charging tanks. ..................................................................................119

Table 7.11 - Case 8. Comparisons / different storage tank capacities. ......................120

Table 7.12 Resource summary used to solves cases...............................................121

_____________________________________________________________________

viii

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

NOTATION

CDU

CDUs

LP

Linear Programming

MILP

MINLP

NLP

LPG

bbl

Barrels

vessel number

j,y

VE

ST

CT

CDU

COMP

SCH

VSi

max

VSimin

VBjmax

VBjmin

CUv

CSEAv

CSINVi

_____________________________________________________________________

ix

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

CBINVj

CCHANGl

TARRv

TLEAv

PUMPCAPv

PUMPCAPi

EVv,k

ESi,k

DMCDUt,l

DMBCOj

XFv,t

XLv,t

XWv,t

F i,j,t

D j,l,t

Z j,y,l,t

TFv

TLv

VVv,t

VSi,t

VBj,t

FVS v,i,t

FSB i,j,t

continuous variable for flow rate from st. tank i to ch. tank j

FBC j,l,t

FKVS v,i,k,t cont. variable for flow rate of component k from vessel v to st. tank i

FKSB

i,j,k,t

cont. variable for flow rate of component k from st. tank i to ch. tank j

FKBC j,l,k,t cont. variable for flow rate of component k from st. tank j to CDU l

VKS i,k,t

VKB j,k,t

ES i,k,t

EB j,k,t

_____________________________________________________________________

x

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

COST

FVS v,i,t

max

FVS v,i,t

min

FSB i,j,t max maximum flow rate from storage tank i to one charging tank j

FSB i,j,t min minimum flow rate from storage tank i to one charging tank j

FBC j,l,t max maximum flow rate from charging tank j to one CDU l

FBC j,l,t max minimum flow rate from charging tank j to one CDU l

ESi,kmax

ESi,kmin

EBj,kmax

EBj,kmin

_____________________________________________________________________

xi

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Chapter 1. INTRODUCTION.

1.1 Oil Refinery Optimisation.

The encouragement to optimise the production planning in oil refineries comes from

the mid fifties when the first applications of linear programming appeared for crude

oil blending and product pooling. Since then, the potential benefits of optimisation for

process operations have come continuously improving. This phenomenon has been

accentuated by the development of global competition and the emergence of

international markets generated from the eighties. Therefore, the competence has

become hard and many oil refineries and petrochemical industries are being

restructured for competing successfully in this new scenario with requirements of low

profit margin, tighter environmental regulations, and more efficient plant operation.

In addition, unit optimizers have been introduced with the implementation of

advanced control systems, generating significant gains in the productivity of the

plants. These successful results have increased the demand for more complex

automation systems that take into account the production objectives {Zhang 2000}.

However, unit optimizers determine optimal values of the process variables but

simply considering current operational conditions {Pinto, Joly, et al. 2000} within a

plant subsystem.

On the other hand, the optimisation of subsystems in the plant does not assure the

global economic optimisation of the plant. The objectives of individual subsystems in

the plant are usually conflicting among them and as consequence; they contribute to

suboptimal and many times infeasible operations. Furthermore, the lack of

computational technology for production scheduling has been the main obstacle for

the integration of production planning objectives into process operations {Barton,

Allgor, et al. 1998}. This integration would help to foresee and solve all the possible

operation infeasibilities on time.

Oil refinery managers are increasingly concerned with improving the planning and

scheduling of their operations for achieving better results and increasing the profit

margin. The major factor that makes this labour difficult among others is the dynamic

nature of the economic environment assisted by the continuous changing of markets.

Then, oil refineries must face the potential impact of demand variations for final

product specifications, volumes requested and prices or even be able to explore

immediate market opportunities {Joly, Moro, et al. 2002}.

Furthermore, the most successful refineries are those which closely monitor their

performance, adjust properly their operations and identify the main weakness for

promptly correcting them {Zhang 2000}. There are many decisions involved to

achieve the optimal operation of an oil refinery. From the managerial level, managers

need to decide which crude oils to purchase, which oil blended compositions to

process, which products to produce, which operating rules to follow, which catalyst to

use, which operating mode to use for each process and so on. And from the process

level, operators have to determine and control the detailed operation conditions for

each equipment and subsystem in the plant. Finally, all decisions should interact the

best among each other {Zhang & Zhu 2000}.

Nevertheless, an operation cycle (See Fig. 1.1), has been proposed by {Pelham &

Pharris 1996} for oil refineries to help to integrate the main functions and achieve

profitable manufacturing by producing quality products under a safe operating

margin. The cycle starts with central planning to determine long term and mid-term

operations. Then, scheduling deals with the short term day-to-day operations.

Advanced control and on line optimisation should translate the goals set by planning

and scheduling to real time process targets, which should be executed by regulatory

control. Monitoring and analysing the results will provide feedbacks to the initial

decision-making procedure. The cycles should be completed with overall refinery

optimisation. The main task consists of finding the best combination of those

decisions to maximise the overall profit. Since overall refinery optimisation almost

covers all the aspects relating to the profit making refinery operations, this is still

considered one of the most difficult and challenging optimisation tasks.

_____________________________________________________________________

2

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Planning

Scheduling

Analysis

Advanced control

and optimisation

Monitoring

Regulatory

Control

Figure 1.1- The cycle of refining operations.

1.2 Other facts and Developments about Oil Refinery Optimisation.

As it was mentioned above since the fifties when linear programming technology

started to produce industrial applications mainly for planning purposes, considerable

research efforts have been applied for oil refinery optimisation. As a result, problems

have been formulated as linear models only because the algebraic applied between

variables are linear or can be closely approximated by linear equations. However, for

process operations which have highly non linear formulations, regarding the kinetics,

thermodynamics, hydromechanics, etc, usually the operation still is manually

controlled based on the operators experience.

The availability of LP-based commercial software for refinery production planning,

such as RPMS (Bonner and Moore, 1979) and PIMS (Process Industry Modelling

System-Bechtel Corp., 1993) have allowed the development of general production

plans for the whole refinery, which can be interpreted as general trends. Then, it is

possible to consider that planning technology is well developed, fairly standard and

widely understood and major changes could not be seen, but evolutionary changes

_____________________________________________________________________

3

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

could occur {Pelham & Pharris 1996 }. In fact, nowadays it already has started to

appear interesting particular development for overall plant and oil refinery production

planning using non linear programming (NLP) techniques {Moro, Zanin, et al. 1998};

{Zhang & Zhu 2000} and {Pinto & Moro 2000}. But, there are still few commercial

tools for production scheduling and these have not allowed a rigorous presentation of

plant particularities {Moro, Zanin, et al. 1998}. For that reason, refineries are

developing in-house tools strongly based on simulation in order to obtain essential

information for a given system {Magalhaes, Moro, et al. 1998 } very particular for

each refinery. Then, nowadays the efforts are strongly guided toward the development

of production scheduling packages that represent the particularities of each plant or

system. In addition, the major work expected in the future whether this development

is successful should be the integration in one integrated information system of the

production planning and the scheduling functions, which allow planners and

schedulers to operate from a single workstation using applications that have access to

the same databases.

The current situation is similar as fig. 1.2 shows. Fewer efforts have been required to

achieve the biggest benefits by preparing the strategic planning and even the

production planning using linear programming in a horizon of one month up to one

year or more. Then, linear programming has come addressing the long term planning

optimisation models achieving these objectives with no major problems. If it moves

down in the fig. 1.2; much more efforts have to be applied to increase the profit

margin in much less proportion. These are the cases related with the production

scheduling and the process operation level which should handle period of time from

hourly base to horizon up to 15 days.

On the other hand, there are some books and papers that mention specific applications

based on mathematical programming, which compare continuous LP and formulations

using MILP, MINLP or NLP and point out the low applicability of models based only

on continuous variables and discuss the lack of rigorous models for refinery

scheduling using other formulations. A mixed integer linear program (MILP) is the

extension of the LP model that involves discrete variables that help to greatly expands

_____________________________________________________________________

4

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

the ability to formulate and solve real-world problems, because logical decisions

(those with variables 0 or 1) can be included. Moreover, in reason of the

combinatorial nature introduced by discrete variables in MILP problems, these have

been very hard to solve. However, MILP and MINLP techniques are essential for

solving optimisation problems in the production scheduling field that mostly relies on

discrete variables.

In summary, the integration of new technologies into process operations is an

essential profitability factor and this can be achieved by through appropriate planning

and scheduling {Joly, Moro, et al. 2002}. Therefore, the importance of on line

integration of planning, scheduling and control has come increasing. Overall plant and

oil refinery production planning presently continue relying on linear programming,

while process optimisation mainly has used either mixed integer linear (MILP), non

linear (NLP) or mixed integer non linear (MINLP) programming techniques. For

individual process or subsystem optimisations, rigorous models have been used to

optimise detailed operating conditions, such as temperatures, detailed process flow

rates, tank volume fluctuations, pressures, blending range compositions, etc. The

results of this kind of optimisation are much closer to the reality. However, there is

still no proper link so far between an overall plant linear programming (LP)

optimisation and different process optimisations in MILP, NLP or MINLP.

_____________________________________________________________________

5

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

The main purpose of this Thesis out of briefly facilitating the understanding of the

huge work that still is demanding the oil refinery optimisation at the scheduling level,

is to develop for oil refineries a new generic scheduling MILP model for optimising

the total operational cost for all the operation that involves the crude oil unloading,

inventory, blending and feeding to crude distillations units in oil refineries, that is one

of the most critical scheduling work in oil refineries due to the operational and

economic impact that it represents. Then, the model development includes the

scheduling of vessel unloading, storage and inventory control, blending and crude oil

component composition control and feeding to crude distillation units. The model

emphasises along the scheduling horizon to strictly follow all the system restrictions

i.e. crude oil component composition range; maximum pumping rate; continuous

feeding rate to guarantee regular feeding to CDUs and so on.

In addition, chapter 2 presents an overview of scheduling optimisation and briefly

reviews all the quite few developments made for production scheduling in oil

refineries pointing out their main particularities. Chapter 3 describes the problem

definition and presents the mathematical formulation of the new model; also it points

out through the chapter content the main conceptual and mathematical differences

between this new model and the model made by {Lee, Pinto, et al. 1996}.

Particularly, in chapter 4, the new model is tested applying the same conditions given

by the examples indicated by {Lee, Pinto, et al. 1996} to show the advantages of the

new model and the shortcomings of the existing model made by {Lee, Pinto, et al.

1996} comparing both results. The model presented by {Lee, Pinto, et al. 1996} is

the oldest one better illustrated found in the literature for this similar purpose and

have encouraged subsequent developments in this field.

On the other hand, some practical cases are presented along chapters 5, 6 and 7

following the request of one particular oil refinery from ECOPETROL, Colombia;

they have been solved using the model to show its advantages and strengths for

solving other examples with different conditions. However, most of the data and

_____________________________________________________________________

6

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

conditions used for the cases are close to the reality for whichever oil refinery. For

better understanding of these cases it is recommendable to start reviewing them from

chapter 5. Some examples and comparisons among them are shown to explain the

benefits of the model not only for production scheduling optimisation, also for helping

to discover and solve equipment limitations and to analyse the impact caused by

equipment expansion applications logically inside the production scheduling scheme,

keeping the target of the operational cost optimisation. Some of the example features

ever have been shown so far in any paper. Thus, it is demonstrated through the

examination of the cases that the new model is flexible and could be adapted to solve

similar cases in oil refineries.

Chapter 8 presents the conclusion of the Thesis and future works. Appendix A shows

partial hardcopy outputs from GAMS {Brooke, Kendrick, et al. 1998 } indicating

only the solve summary and the data used in the Thesis for all the three examples

analysed in chapter 4 and for some practical cases analysed in chapters 5,6 and 7.

Normally, one output per exercise involves more than 100 pages showing all the

features of the solving steps to get the optimal solution as well as the solution

summary and data mentioned. On the other hand, appendix B shows as an example,

the model programmed using the software GAMS, specifically for solving case 1.

Other examples and cases were run changing the input data according to the new

conditions and adding or taking away some equations or constraints according to the

model mathematical formulation explained in chapter 3.

Finally, due the handling of a huge amount of equations and variables by the model

for solving these examples and cases; definitely, the documentation was found to be

indispensable for the software GAMS{Brooke, Kendrick, et al. 1998 } used to

program

the model

constraints for obtaining the optimal solutions for all the problem conditions.

_____________________________________________________________________

7

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

2.1 Scheduling General Concern.

The scheduling is strongly related to the planning at other levels, and it affects many

types of decisions in the oil refinery. The ability to efficiently construct high quality,

feasible and low cost schedules is therefore crucial for the refinery in order to be

competitive.

The scheduling should be concerned about which mode of operation to use in either

each process plant or each plant subsystem at each point of time, in order to satisfy the

demand for a given set of products. A mode of operation for a process plant is

specified by the combination of products consumed and produced in the process, and

by the yield levels for each of the products. Changeovers between modes of operation

cause disturbances and extra costs to the refinery. Hence, long sequences of the same

mode of operation applying few changeovers are preferred. However, long sequences

imply larger inventory volumes for some products and an increased need for storage

capacity with associate larger holding and capital costs. Thus, there must be a tradeoff between the negatives effects of frequents changeovers (feed switch and start up

costs), and the cost of keeping large inventory volumes {Goethe-Lundgren, Lundgren,

et al. 2002}.

On the other hand, scheduling models are intended to determine the timing of the

actions to execute the plan, taking in account i.e. available storage, arrivals of

feedstock (by vessels or oil pipes) and handling of blending compositions. They

should focus on the when to do. They investigate the limitations imposed by space

and time. Then, an accurate description of the initial inventory in all the tanks

(compositions and quantities) and a prediction of all the streams required or presented

in the boundary of the system during the coming days are needed. They should draw

attention to problems encountered in storage (overflow or shortage) and timing

(demurrage). Solving these problems may require rescheduling, reallocation of

storage facilities or even reformulation of another plan {Hartmann 1998} .

_____________________________________________________________________

8

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

In addition, regarding the scope of the new model developed in this Thesis; the model

is designed to help to solve all the concerns mentioned above related to the scheduling

problem. Moreover, the activities of crude oil unloading, storage, blending and

feeding operations, are considered one of the major bottlenecks in the production

chain in oil refineries and therefore, it is interesting to start assisting with the model to

solve the scheduling concerns from this area where delays imply loss of time and lack

of resources and deliveries ahead of the deadlines may cause excess of inventory. At

the end of the day, every refinery must proceed with efficient schedules regarding

theses activities within their operational planning scope.

2.2 Crude Oil Inventory Scheduling Optimisation.

Typically, an oil refinery receives its crude oil through a pipeline, which is linked to a

docking station where oil vessels or tankers unload. The unloading schedule of these

tankers is usually defined at the corporate level and can not be changed easily {Joly,

Moro, et al. 2002}. Thus, for a given scheduling horizon, the number, type and

arriving and departure times of the oil tankers are known a priori.

This thesis is focused on the production scheduling optimisation of operation modes

concerning crude oil vessel unloading, storage, blending and feed to crude distillation

units (CDUs). The new optimisation model proposes the strategic operation for the

system in accordance with the given conditions and the optimal operational cost

calculated. The strategic operation if it is feasible must follow the proposed suitable

unloading days for vessels and the proposed flow rates among vessels and tanks,

among storage and charging tanks and among tanks and plants for keeping the optimal

production scheduling given by the model results. Moreover, this model can be used

as a viable tool not only for supporting the shipment planning, also for discovering

system infeasibilities and for strategic decisions concerning investments in storage

and pumping systems.

The purpose is to satisfy the demand while the inventory volumes and the pumping

system capacities are not violated. The shipment plan requires to be realizable in

_____________________________________________________________________

9

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

terms of production, and a cross checking between the shipment plan and the

corresponding required feeding to plants according to the production schedule should

be done to be sure that the feeding can be done in due time {Goethe-Lundgren,

Lundgren, et al. 2002 }. Moreover, the new optimisation model will help to do this

cross checking and answer whether a shipment plan definitely can match in terms of

production or not.

On the other hand, the need for some supporting optimisation tools for the scheduling

is also accentuated by the frequent change in the planning situation. For example, the

shipment plan may change since it is based on forecasts of demand and vessels may

arrive delayed or interchanged. Whenever a change occurs in the shipment plan, a re

scheduling of the involved system needs to be carried out in order to not lose money

stopping the possible operational cost increase by this effect. Moreover, there would

be other non planned situations for instance that could affect the production

scheduling and therefore, a rescheduling also is needed to check if the conditions are

still feasible and the optimal operational cost has not been negatively impacted i.e.

crude oil vessel unloading volume increase; cost alteration in vessel operations and so

on.

In summary, the new optimisation model can be used to evaluate whether a suggested

shipment plan is feasible or not, to evaluate the optimal operational cost incurred by a

particular shipment plant, and to support decisions on whether a shipment plan

should be changed or not. Of great importance to mention, it is also the possibility of

analysing many scenarios and to be able to make fast evaluations of consequences of

sudden changes in shipment plans, proposed changes in some tanks or pumps

capacities, etc.

2.3 Review of Production Scheduling Developments

A good mention of important contributions to the production planning and scheduling

in oil refineries already has been done through chapters 1 and 2. However, regarding

the concern of this Thesis related to the optimal operational scheduling for crude oil

_____________________________________________________________________

10

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

unloading, inventory , blending and feeding to CDUs, the first MILP works were

reported by {Shah 1996 }and {Lee, Pinto, et al. 1996 }.{Shah 1996} complementarily

mentions that his model could be extended to cover the design of tanks not giving any

practical example about this matter and {Lee, Pinto, et al. 1996 } particularly presents

a better illustrated model than {Shah 1996}, focused on using a continuously feed

following a demand of blended crude oils for charging the CDUs . Other

developments in this area have appeared later by {Pinto, Joly, et al. 2000}; {Joly,

Moro, et al. 2002}; {Goethe-Lundgren, Lundgren, et al. 2002} ; {Mas & Pinto 2003 }

and {Jia, Ierapetritou, et al. 2003 }. {Pinto, Joly, et al. 2000} and {Joly, Moro, et al.

2002} present a MILP problem unloading the crude oil by scheduled batches of

different composition through a single oil pipeline instead unloading from vessels;

moreover, they use only one set of tanks for managing the storage and blending

operations just before feeding the CDUs. {Goethe-Lundgren, Lundgren, et al. 2002}

emphasises the operation modes and formulates a MILP model considering an

example changing storage tank capacities for analysing future investments. {Mas &

Pinto 2003} presents a decomposition strategy applying several MILP models for

each subsystem of a large system analysed which has events too difficult to solve

simultaneously and {Jia, Ierapetritou, et al. 2003 } propose a solution which leads to a

MILP model with fewer binary and continuous variables and fewer constraints saving

computer time; moreover, their model was tested using the same condition data of the

examples of {Lee, Pinto, et al. 1996 }; nevertheless, the results from {Jia, Ierapetritou,

et al. 2003} show that the optimal operational costs have increased for three out of

four examples of {Lee, Pinto, et al. 1996 }; specifically, 8.76 % more for example 1,

10.82% more for example 2 and 1.98 % more for example 4; example 3 presents a

reduction of 3.68 %; but, they do not present any explanation with respect to the cost

increase.

On the other hand, considering another section of the oil refinery, few developments

for the production scheduling optimisation have been done for particular process

plants in oil refineries. These applications should involve models developed using

MILP or MINLP techniques. {Joly, Moro, et al. 2002} develops a MINLP model for

managing the production scheduling problem in a fuel oil and asphalt plant including

_____________________________________________________________________

11

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

its inventory control and distribution; but the MINLP is transformed to a MILP

because no global solution was guaranteed by the convectional MINLP solution

algorithms due to the bilinear terms in the viscosity constraints. {Magalhaes, Moro,

et al. 1998} describes the development of an integrated production scheduling system

(SIPP) for an oil refinery that at the date of the paper publication was in

commissioning. The system integrates other refinery applications and databases and

the MILP optimisation techniques; but they do not refer to any interface done between

the LP-tool at the planning level and the tools used at the scheduling level. Even at the

scheduling level is not mentioned any interface among the applications that works at

this level and the MILP optimisation techniques used. The SIPP user has to apply an

iterative process until the scheduling programme is feasible. Moreover, special care

should be applied to identify is the feasible solution is the optimal. On the other hand,

process units models are presented as steady-state based in the data input from the

LP-planning and instead, tanks and pipelines have transient models. Separately, they

mention an application using MILP techniques to assist the scheduling production of

the LPG area and their future intention to be managed by the SIPP.

Finally, regarding the production scheduling optimisation of finished product

blending , storage and distributions, it has been other developments by {Breiner Avi

& Maman 2001 } and {Jia & Ierapetritou 2003}. {Breiner Avi & Maman 2001}

introduces a MINLP developed system (MPMP) successfully installed in an oil

refinery for managing and optimising the final product blending and storage

production scheduling. The MINLP is extremely difficult to solve satisfactorily and

hence, they use a four step algorithm and separate the MILP from the NLP operation

for solving it. On the other hand, {Jia & Ierapetritou 2003} introduces a MILP model

to manage the operation of gasoline finished products in an oil refinery, specifically

blending, storage and distribution pipelines.

_____________________________________________________________________

12

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

DEFINITION.

This chapter presents the problem definition and the mathematical formulation of the

production scheduling optimisation problem for the unloading, storage, blending and

feeding of an oil refinery. The new model will be formulated using general notation

showing the possibility of using the model for a general problem of this matter.

3.1 Problem Definition

The system configuration for this production scheduling optimisation model

corresponds to a multistage system consisting of vessels, storage tanks, charging

tanks, and CDUs as is illustrated in fig.3.1.

For a given scheduling horizon, crude oil vessels arrive to the refinery docking station

which only allows one vessel for unloading. In accordance with the planning level

each vessel will have a reasonable date for leaving that should be at most the date that

the next vessel arrives. The day one vessel arrives could start to unload depending of

_____________________________________________________________________

13

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

the optimal results recommended by the model in accordance with all the current

conditions of the system analysed. The day one vessel finishes unloading should be up

to one day before its maximum departure date fixed at the planning level; the reason is

because the maximum departure date allowed for the preceding vessel is the same

planned arriving date of the next vessel and potentially this next vessel could start

unloading on the arriving date. Instead {Lee, Pinto, et al. 1996} is not clear in their

model description about the established limit for the vessel departure date after it has

completed unloading. Each time that a vessel rescheduling is done, the maximum

departure dates for vessels are reformulated in accordance with the new arriving dates

for each vessel and both will be new input conditions to the optimisation model for

avoiding basic interferences.

The crude oil is unloaded into storage tanks at the docking station and the problem

considers one pre selected storage tank per vessel that manages the same crude oil

composition of the vessel. Then, the crude oil is transferred from storage tanks to

charging tanks. Each crude oil inside the charging tank must carry out to be within a

range of blended crude oil composition determined at the planning level for the

scheduling horizon. The fulfilment of this blended composition range is made per tank

having in account the material balance of the remaining volume and the different

flows coming in or coming out the tank with their different compositions per time

interval. Then, blended crude oils from charging tanks are charged into the CDUs and

whenever that it is optimally required feed switches are done from one kind of

blended crude oil to another for each CDU. Finally, it is important to mention that

whether a charging tank is feeding a CDU, it must not be fed by any storage tank or

vice versa.

In addition, there will be probably special problems that could consider more storage

tanks than crude oil vessels and therefore, vessels have to unload to more than one

storage tank. This situation could be managed by the optimisation model carrying out

a pre established crude oil blended composition range in each storage tank like

charging tanks. Then, storage tanks have to carry out blending material balance

conditions and limitations the same presented in charging tanks i.e. they must not be

_____________________________________________________________________

14

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

fed by vessels if they are feeding charging tanks or vice versa. Through the model

mathematical explanation, the equations that should be used to solve this particular

problem will be pointed out. This situation makes more difficult the solving solution

as it will be illustrated solving the example 3 of {Lee, Pinto, et al. 1996} in chapter 4.

On the other hand, the new model for the production scheduling optimisation

developed in this Thesis is concerned to meet a feed demand per each crude

distillation unit (CDU) and not concerned instead to meet a fixed demand of a specific

blended crude oil demand for charging the CDUs along the scheduling horizon,

disregarding the allowed charging rate variation per each CDU as {Lee, Pinto, et al.

1996} presents. However, the new model easily can be configured selecting the

properly required solving option depending on the desirable kind of demand for the

problem. Furthermore, to meet a feed demand to each plant does not mean that the

control of charging different blended crude oils will be lost; although, there is not a

fixed amount of blended oil to consume along the scheduling horizon, the manager

could know what kind of blended oil will optimally charge each CDU following its

feed changeovers i.e. either high or low sulphur content crude oil. Moreover, to meet

a feed demand guarantees the stable running of the plant along the scheduling horizon

and a clear understanding of the production planning fulfilment.

Although {Lee, Pinto, et al. 1996} shows in their existing model the operating

constraints for the flow rate variation, it is not clear how they manage the restriction

to total flow rates variation among vessels and tanks and among storage and charging

tanks. This new model totally takes care about this situation that is very important in

real situations in accordance with the pumping system capacities and the real flow

limitations with respect to pumping in parallel to different tanks from one single

source that could be either a tank or a vessel.

On the other hand, {Lee, Pinto, et al. 1996} compares the results from their example 1

and

Regarding this matter, it is understood that ruled based or heuristic schedules are not

the best and therefore, this Thesis is interested not only to find the optimal schedules

_____________________________________________________________________

15

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

and the optimal operational cost for each situation, also in exploring and analysing

how the optimal operational cost of the system could be improved, for instance

attending one manager concern like this: If there is some money to invest in the

feeding system to plants, what we should do to reduce the optimal operational cost

and simultaneously getting at each time interval (day) the feeding rates to plants

closer or equal to the maximum capacity as well. Finally, it is possible to observe

other own particularities of this new model along the model presentation and

explanation in this chapter.

3.2 Optimisation Model Formulation Introduction

Given the configuration of this multistage system and the arrival time of the vessels,

the equipment capacity limitations and the key component concentration ranges for

crude oils, the problem will focus on determine the following operating variables to

minimise costs:

a. Waiting time for each vessel in the sea after arriving.

b. Unloading duration time for each vessel.

c. Crude oil unloading rate from vessels to storage tanks.

d. Crude oil transfer and blending rates from storage tank to charging tanks.

e. Inventory volumes of storage and charging tanks.

f. Crude distillation unit charging rates fulfilling the demand per each CDU.

Instead {Lee, Pinto, et al. 1996} sets to fulfil a demand of blended crude oils.

g. Sequence of type of blending crude oil to be charged in each CDU in

accordance with the optimal mode changeovers.

The following are the operating rules that have to be obeyed:

a. In the scheduling horizon each vessel for unloading should arrive and leave

the docking station.

b. If a vessel does not arrive at the docking station, it can not unload the crude

oil.

_____________________________________________________________________

16

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

c. If a vessel leaves the docking station, it can not continue unloading the crude

oil.

d. The vessel can start unloading the same arriving date.

e. The vessel can not unload on its maximum departure date. Except for the last

vessel.

f.

While the charging tank is charging one CDU, crude oil can not be fed into

the charging tank and vice versa.

g.

Each charging tank can feed at most one CDU at one time interval.

h. Each CDU is charged by only one blended crude oil at one time interval.

On the other hand these are the following operating constrains that must be met:

a. Equipment capacity limitations: Tank capacity and pumping rate.

b. Quality limitations of each blended crude oil: Range of component

concentrations in each blended crude oil.

c. Demand per interval of time (day) of each CDU. Instead {Lee, Pinto, et al.

1996}sets up to follow a demand of each blended crude oil for the scheduling

horizon.

The model minimises the operation cost for the total system shown. The model is

formulated using general notation and MILP formulation, showing the possibility of

using the model for a general problem of this matter. In order to develop the multiperiod MILP model, continuous and binary variables are associated with the system

network. The following are the assumptions for the proposed model:

a. Only one vessel docking station for crude oil unloading is considered.

b. The time applied for the changeover are neglected and also the transient flows

generated during either start up or shut down when a changeover is done.

c. Perfect blending is assume for each charging tank while it is being fed by

different crude oils, and additional blending time inside the tank is not

required before it charges the CDU.

_____________________________________________________________________

17

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

d. The composition of the crude oil is decided by the amount of key components

presented in the crude oil or in the blended crude oil. In general, sulphur is at

least one of the key components for differentiating between crude oils.

A uniformed discretisation is chosen in the given scheduling horizon for the proposed

scheduling model. The selection of the length of each discretised time span involves a

trade off between accurate operation and computational effort. The reviewed cases in

this thesis involve 15 time intervals during the scheduling horizon. Instead, the 3

examples of {Lee, Pinto, et al. 1996}reviewed in this Thesis involve 8, 10 and 12

time intervals respectively.

The optimisation problem involves the following sets, parameters and variables:

Sets

VE = {v = 1, 2... V / crude oil vessel or tankers}

ST = {i = 1, 2... I / storage tanks}

CT = {j,y = 1, 2... J / charging tanks}

COMP = { k = 1, 2... K / crude oil components}

CDU = {l = 1, 2... L / crude distillation units}

SCH = {t = 1, 2... T / time intervals along the scheduling horizon}

Param eters

VSimax : storage tank maximum capacity.

VSimin : storage tank minimum capacity.

VBjmax : charging tank maximum capacity.

VBjmin : charging tank minimum capacity.

CUv :

CSINVi : inventory cost of storage tank i per unit time interval.

CBINVj : inventory cost of charging tank j per unit time interval.

CCHANGl : changeover cost of CDU l.

_____________________________________________________________________

18

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

TLEAv : crude oil vessel maximum departure date from the docking station.

PUMPCAPv: maximum pumping system capacity or total flow capacity from each

vessel v to storage tanks.

PUMPCAPi : maximum pumping system capacity or total flow capacity from each

storage tank i to charging tanks.

EVv,k :

ESi,k :

tank i.

ESi,kmax : maximum concentration of component k in the blended crude oil of storage

tank i.

EBj,kmin : minimum concentration of component k in the blended crude oil of charging

tank j.

EBj,kmax

charging tank j.

DMCDUt,l : demand of each CDU l per time interval.

TOTDMCDUl : total demand of each CDU l along the scheduling horizon.

DMBCOt,j : demand of each blended crude oil j along the scheduling horizon. It will

be used to solve the examples of {Lee, Pinto, et al. 1996}.

FVS v,i,tmax : maximum crude oil rate from vessel v to one storage tank i.

FVS v,i,tmin : minimum crude oil rate from vessel v to one storage tank i. This variable

is not mandatory to have a value. It could be either 0 or a positive value depending of

the minimum flow restriction for the vessel pumping system that normally is assisted

working in series with the refinery pumping system to storage tanks.

FSB i,j,t max : maximum crude oil rate from storage tank i to one charging tank j.

FSB

i,j,t

min

: minimum crude oil rate from storage tank i to one charging tank j. By

default this parameter has a value of 0 as the minimum value. A small optimal flow

rate if it is optimally required could be managed with centrifugal pumps and process

control instrumentation.

_____________________________________________________________________

19

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

FBC j,l,t max : maximum crude oil rate from charging tank j to one CDU l . In general

this value corresponds to the maximum CDU l feed capacity or it could be adjusted to

a lower value.

FBC

j,l,t

min

: minimum crude oil rate from charging tank j to one CDU l .This value

could be adjusted to the same value DMCDUt,l , depending on the problem conditions.

Binary variables

XFv,t : variable to denote if vessel v starts unloading at time t.

XLv,t : variable to denote if vessel v finishes unloading at time t.

XWv,t : variable to denote if vessel v is unloading its crude oil at time t.

F i,j,t : variable to denote if the crude oil blended in storage tank i is feeding charging

tanks at time t; otherwise storage tank i could be being fed by vessel v.

D j,l,t : variable to denote if the crude oil blended in charging tank j charges CDU l at

time t; otherwise charging tank j could be being fed by storage tanks.

Z

j,y,l,t

: variable to denote switch of the blended crude oil fed to CDU l from the

Integer variables

TFv : vessel v unloading initiation time.

TLv : vessel v unloading completion time.

Continuous variables

VVv,t : volume of crude oil in vessel v at time t.

VSi,t : volume of crude oil in storage tank i at time t.

VBj,t : volume of crude oil in charging tank j at time t.

FVS v,i,t : volumetric flow rate from vessel v to storage tank i at time t.

FSB i,j,t : volumetric flow rate from storage tank i to charging tank j at time t.

FBC j,l,t : volumetric flow rate from charging tank j to CDU l at time t.

FKVS v,i,t : volumetric flow rate of component k from vessel v to storage tank i at time

t.

FKSB i,j,k,t : volumetric flow rate of component k from storage tank i to charging tank j

at time t.

_____________________________________________________________________

20

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

FKBC j,l,k,t : volumetric flow rate of component k from storage tank j to CDU l at time

t.

VKS i,k,t : volume of component k in storage tank i at time t.

VKB j,k,t : volume of component k in charging tank j at time t.

ES

i,k,t

time t.

EB

j,k,t

time t.

COST : total optimal operational cost.

Initial conditions

VVv, TARRv : initial volume of crude oil vessel at time equal TARRv .

VSi,1 : initial volume of storage tank i at time equal 1 in the start up of the scheduling

horizon.

VBj,1 : initial volume of charging tank j at time equal 1 in the start up of the

scheduling horizon.

ES

i,k,1

EB j,k,1 : concentration of component k in the blended crude oil of charging tank j at

time t equal 1 in the start up of the scheduling horizon.

VKS i,k,1 : initial volume of component k in storage tank i at time t equal 1 in the start

up of the scheduling horizon .

VKB

j,k,1

3.3 Model Mathematical Formulation.

The model focus on minimising the following operation cost of the system for the

operations of crude oil vessel unloading, storage, blending and feeding to crude

distillation units in an oil refinery. Then, this is the main objective equation that

represents the total operation cost of the system:

_____________________________________________________________________

21

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

COST =

v=1

I

-TFv + 1) CUv

+

v=1

[ (TFv - TARRv

) CSEAv

i=1 t=1

J

[ (TLv

j=1

t=1

(CCHANGl

Z j,y,l,t)

(3.1)

3.3.1 Vessel Arrival and Departure Operation Rules.

Each vessel must arrive to the docking station for unloading only once through the

scheduling horizon:

T

t=1

XFv,t = 1 , v

VE

(3.2)

Each vessel leaves the docking station only once through the scheduling horizon:

T

t=1

XLv,t = 1 , v

VE

(3.3)

T

TFv =

tXFv,t

, v

VE

(3.4)

t=1

T

TLv =

tXLv,t

, v

VE

(3.5)

t=1

_____________________________________________________________________

22

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Each vessel must start unloading either after or on the arrival time established at the

planning level:

TFv

TARRv , v

VE

(3.6)

Each vessel must finish unloading up to one interval of time before the maximum

departure time established at the planning level:

TLv < TLEAv , v

VE , v

(3.7)

TLv

TLEAv

v= V

(3.8)

TLv - TFv

1 , v

VE

(3.9)

The preceding vessel must finish unloading one time interval before the next vessel in

the sea arrives and starts to unload:

TFv+1

TLv + 1 , v

VE

(3.10)

Unloading of vessel v only will be possible between time TFv and TLv:

T

XWv,t

m

XFv,t , t

m , v

VE,

t=1

{ m = TARRv , TLEAv }

(3.11)

XWv,t

XLv,t , t > m , v

VE,

t=1

{ m = TARRv , TLEAv }

(3.12)

_____________________________________________________________________

23

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

The crude oil in vessel v at time t+1 must be equal to the crude oil in vessel v at time t

taking away the crude oil transfer from vessel v to storage tank i at time t:

VVv,t = VVv, TARRv , t = TARRv

(3.13)

VV v,t+1 = VVv,t -

i=1

FVS v,i,t , v

VE , t

SCH

(3.14)

The crude oil of each vessel v has different composition. Then, if there is one storage

tank assigned to each vessel, the solution must guarantee that the crude oil from each

vessel v is transferred to the corresponded storage tank i which will handle the same

vessel crude oil composition. Then, for this general case, i is equal to v to identify the

respective storage tank for the vessel:

I

i=1

t=1

FVS v,i,t

i=1 t=1

FVS v,i,t , v

VE ,

(3.15)

If there is more storage tanks than vessels, the problem could consider managing

crude oil blending composition ranges per storage tank and then, each vessel could

optimally unload to several tanks meeting with the pre established blending range for

each storage tank as it was explained above. Whether this is the case, equation 3.15 is

not useful and must not be used for this situation. However, instead the following

equation should be used to control the total flow from each vessel to storage tanks

within a reasonable margin:

I

i=1

FVS v,i,t

PUMPCAPv , v

VE

SCH

(3.16)

_____________________________________________________________________

24

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

The volume of the total crude oil transferred from vessel v to storage tanks during the

schedule horizon must be equal to the initial crude oil volume of vessel v:

i=1 t=1

FVS v,i,t

VVv, TARRv, v

VE

(3.17)

Equations 3.18 and 3.19 are the operating constraints on crude oil transfer rate from

vessel v to storage tank i at time t:

v

VE , i

FVS v,i,t

ST, t

ST, j

FVS v,i,t

CT , v

SCH

(3.18)

VE , t

SCH

VE ,

(3.19)

Equation 3.19 will be applicable only if storage tanks are managing crude oil blending

ranges. The term (1 - F

i,j,t

storage tank i , there is no oil transfer from storage tank i to any charging tank j at

time t.

The crude oil in storage tank i at time t+1 must be equal to the crude oil in storage

tank i at time t plus the crude oil transferred from vessel v to storage tank i at time t

taking away the crude oil transferred to charging tanks j at time t:

VSi,t

= VSi, 1 ,

t =1

(3.20)

_____________________________________________________________________

25

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

VS i,t+1 = VSi,t +

i

v=1

ST,

FVS v,i,t -

j=1

FSB i,j,t ,

SCH

(3.21)

VSimin

VSimax ,

VSi,t

ST,

SCH

(3.22)

Equations 3.23 and 3.24 indicate the operating constraints on crude oil transfer rate

from storage tank i to charging tank j at time t:

FSB i,j,t

0

i

ST,

CT, l

CDU , t

SCH

(3.23)

The term (1- D j,l,t ) denotes that if charging tank j is charging any CDU l , there is no

oil transfer from any storage tank i to charging tank j.

FSB i,j,t

i

ST, j

CT, t

SCH

(3.24)

Equation 3.24 will be applicable only if storage tanks are managing crude oil blending

ranges. The term F

i,j,t

charging tanks, it must not have oil transfer from a vessel v to storage tank i at time t.

There is no minimum set rate different from 0, because the problem considers the use

of centrifugal pumps commonly used for this kind of service and therefore, the

optimal flow FSB

i,j,t

optimal range including the optimal lower flows that would be required for one or

more time intervals along the scheduling horizon if the transfer between the storage

tank and one charging tank is not null at any interval of time t.

_____________________________________________________________________

26

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

However, the total maximum flow allowed by the capacity of the pump or the

pumping system from one storage tank i to several charging tanks j at time t is

controlled by the following constraint:

FSB i,j,t

j=1

PUMPCAPi , i

ST, t

SCH

(3.25)

Generally speaking, the flow rate from one storage tank to one charging tank(FSB i,j,t )

almost always is a little smaller than the total flow rate from the same tank to several

charging tanks (

FSB

i,j,t

pumping system taking from one storage tank i is discharging to one charging tank j

circuit than when is discharging to several charging tanks. Furthermore, FSBi,j,tmax

depends on the system characteristics from each storage tank i to each charging tank

j. However, for the cases analysed in this Thesis; a symmetrical system is considered

and therefore, FSBi,j,t

max

will have the same value for each circuit from one storage

3.3.4 Material Balance Equations for the Charging Tank.

The crude oil blended in the charging tank j at time t+1 must be equal to the crude oil

in the charging tank j at time t plus the crude oil transferred from the storage tanks

taking away the crude oil transferred to the CDU l at time t:

VBj,t

= VBj, 1

VB j,t+1 = VBj,t

j

CT, t

, t =1

(3.26)

L

FBC j,l,t ,

l=1

SCH

(3.27)

_____________________________________________________________________

27

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

VBjmin

VBjmax , j

VB j,t

CT, t

SCH

(3.28)

The volume of component k in storage tank i at time t+1 is equal to the volume of

component k in storage tank i at time t plus the volume of component k in crude oil

transferred from vessel v to the storage tank i taking away the volume of component k

in the blended crude oil i transferred to charging tanks at time t: (Theses equations

must be considered only if storage tanks have blending functions).

t = 1

(3.29)

VKS i,k,t+1

VKS i,k,t

CT , k

FKVS v,i,k,t

v=1

COMP ,

FKSB i,j,k,t

j=1

SCH

(3.30)

The following is the operating constraint on volumetric flow rate of component k from

vessel v to storage tank i:

FKVS v,i,k,t = FVS v,i,t EVv,k , v

i

ST , k

COMP ,

VE ,

SCH

(3.31)

The following are the operating constraints on volumetric flow rate of component k

from storage tank i to charging tank j:

FKSB i,j,k,t

CT , k

ST,

COMP ,

SCH

(3.32)

The limitations in the volume capacity for component k in storage tank i at time t are

given by:

_____________________________________________________________________

28

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

VS i,t ESi,kmin

i

VS i,t ESi,kmax

VKS i,k,t

ST , k

COMP , t

SCH

(3.33)

The volume of component k in charging tank j at time t+1 is equal to the volume of

component k in charging tank j at time t plus the volume of component k in crude oil

transferred from storage tanks to the charging tank j taking away the volume of

component k in the blended crude oil j transferred to CDU l at time t:

VKB j,k,t = VKB j,k, 1 ,

t = 1

I

VKB j,k,t+1

j

VKB j,k,t

CT , k

FKSB i,j,k,t

i=1

COMP , t

FKBC j,l,k,t ,

l=1

SCH

(3.34)

The following is the operating constraint on volumetric flow rate of component k from

storage tank i to charging tank j:

FKSB i,j,k,t = FSB i,j,t ESi,k ,

i

ST ,

CT , k

COMP ,

SCH

(3.35)

Constraint 3.35 is not needed when storage tanks are operating as blending tanks.

The following are the operating constraints on volumetric flow rate of component k

from charging tank j to CDU l:

EBj,kmin

FBC j,l,t

j

CT , l

FKBC j,l,k,t

CDU, k

COMP

SCH

(3.36)

The limitations in the volume capacity for component k in charging tank j at time t are

given by:

_____________________________________________________________________

29

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

VBj,k,t EBj,kmin

j

VBj,k,t x EBj,kmax

VKB j,k,t

CT , k

COMP

SCH

(3.37)

3.3.7 Operating Rules for Crude Oil Charging to Crude Distillation Units.

As it was stated above each CDU l only can be charged by one charging tank j at time

t:

J

j=1

D j,l,t = 1 ,

CDU , t

SCH

(3.38)

On the other hand, each charging tank j can charge at most one CDU l at time t:

L

l=1

D j,l,t

1 ,

CT , t

SCH

(3.39)

Equation 3.39 is useful when the system analysed considers more than one CDU.

If the CDU l is charged by crude oil blended j at time t and after is charged by crude

oil blended y at time t+1 then, changeover cost must be involved. The following is the

condition that confirms that changeover cost shall be charged:

Z j,y,l,t

l

j,l,t

CDU,

+ D y,l,t+1 -1

t

, j,y (j

SCH

y)

CT ,

(3.40)

Only one of the following groups of constraints must be added to the model

depending of the production planning demand target for the crude distillation units

along the scheduling horizon.

_____________________________________________________________________

30

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

a. If it is requested a fixed demand of blended crude oils from charging tanks along

the scheduling horizon to charging all CDUs and a feed rate variation range to

feed the crude distillation units is allowed, the following constraints are

applicable:

FBC j,l,t min D j,l,t

j

CT , l

L

FBC j,l,t

CDU,

SCH

(3.41)

CT

(3.42)

l=1 t=1

This group of equations will be used in the new model to solve the examples of

{Lee, Pinto, et al. 1996 }. Particularly, this option a with these two constraints

could have infeasibility problems depending on the problems conditions. {Lee,

Pinto, et al. 1996} uses a higher charging rate variation range for feeding the

CDUs and this condition may have helped to solve the problems easier. Then, for

their examples, the charging rate variation normally goes from 25% to 100% of

the maximum value as they will be observed in their graphs of results in Chapter

4.

b. If it is requested a charging rate demand per time interval for each CDU and it is

allowed a feed rate variation up to either the maximum plant capacity per time

interval or any other maximum allowed feed per time interval, the following

constraint is applicable:

DMCDUt,l D j,l,t

j

CT , l

FBC j,l,t

CDU,

t

SCH

,

(3.43)

charging variation range allowed between the feed demands from plants and their

maximum charging feeds previously fixed. The cases analysed in this Thesis in

chapters 5, 6 and 7 apply this constraint in the model to solve practical cases in an

_____________________________________________________________________

31

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

oil refinery where the manager has requested to review the potential use of the

model in the oil refinery; specifically in its crude oil feeding system to the existing

crude distillation unit. It is not desirable to be under the CDU demand charging

rate per day adjusted in 75,000 barrels per day; furthermore, the production

planning allows an overproduction up to 10 % above the CDU charging demand

per day along the scheduling horizon. Particularly, the charging feed rate per day

for the CDU could go up to 82,000 barrels per day (9.33 % more of the demand

rate per day) only if the conditions of the problem optimally allow it.

Finally, solving cases using this option b will allow an easier understanding of

the model not only for solving scheduling problems also for review the impact of

future development in the system. Particularly, a case with the CDU shutting

down one day n in the middle of the schedule horizon was analyzed using the

additional constraint:

FBC j,l,t = 0 , t = n , n

SCH

(3.44)

Logically, the feed demand per day is not fulfilled the day n; however an analysis

of the overall schedule problem is analysed with this option giving some

recommendations.

c. If it is requested a fixed feed rate demand per time interval for each CDU, the

following constraint is applicable:

FBC j,l,t

j

CT , l

DMCDUt,l D j,l,t

CDU,

SCH

(3.45)

d. If it is requested a fixed total demand per CDU along the scheduling horizon and a

charging rate variation range to feed the crude distillation units is allowed, the

following constraints are applicable.

_____________________________________________________________________

32

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

CT , l

j

J

FBC j,l,t

CDU,

SCH

(3.46)

j=1 t=1

CDU

(3.47)

The term D j,l,t denotes that if there is any oil transfer from storage tank i to charging

tank j , there is no oil transfer from charging tank j to CDU l.

It could have other options using for instance a combination of the above groups; but

this situation hardly could get feasible solutions unless some data needed for the

constraints are relaxed and therefore, new combination easily could come out with

similar solutions to the solutions given by one of the options presented above.

3.4 Chapter Summary.

This chapter deeply described the problem definition and the mathematical

formulation in an algebraic form of the new generic optimisation production

scheduling model for managing and optimising the operational cost in an oil refinery

regarding the scheduling functions of unloading, storage, blending and feeding of

crude distillation units. The main objective function to minimise equation 3.1, was

described including among other the changeover cost, unloading cost and inventory

cost. Moreover, all equations, constraints and variables required for the model were

well described explaining in some cases their particularities. Also, the different

directions for solving scheduling problems with this new model were indicated

pointing out the complementary different constraints that must be used in the model to

solve each one of the different options.

_____________________________________________________________________

33

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Some of the differences between this new model and the existing model developed

by {Lee, Pinto, et al. 1996} have been pointed out through this chapter except along

the model mathematical formulation due the existing quite amount of differences

detected. The main mathematical formulation differences are:

The new model presents the main objective function plus 31 equations and

constraints used to solve option b and c and 32 for options a and d

respectively. 8 equations and constraints more must be added if the problem

involves blending in storage tanks to whichever of the options presented.

Instead the existing model for solving option b presents the main objective

function plus 25 equations and constraints; the equations and constraints to

solve problems like example 3 including blending in storage tanks are missed;

{Lee, Pinto, et al. 1996}only mention that this example needs additional

constraints including a material balance equation for components in storage

tanks and the operation rule related to vessel unloading to storage tanks when

they are not feeding charging tanks or vice versa; but, they did not include any

equations and constraints in the existing model related to this matter.

The basic equations and constraints missed by {Lee, Pinto, et al. 1996}are

related with the total flow rate restriction from vessels to storage tanks;

specifically equation 3.15 and constraint 3.16. Moreover, constraint 3.25

related to the total flow rate from one storage tank to several charging tanks

also is missed. On the other hand, the existing model missed the constraints

3.6 and 3.7 related to the vessel maximum departure date.

In general, the formulation of the new equations and constraints are different

compared with the existing model; there are some differences even in

equations or constraints with similar purposes i.e. the main objective function

of the existing model is missing out to charge one day to the total unloading

cots per vessel: they wrote the total unloading cost per vessel v as (TLv -TFv )

CUv and not (TLv -TFv + 1) CUv as the new model presents. On the other

hand the outline of the constraints 3.11 and 3.12 regarding the only possibility

of unloading for vessel v between TFv and TLv is different in both models.

_____________________________________________________________________

34

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Finally, in accordance with the example and cases analysed, the model was

programmed using the software GAMS{Brooke, Kendrick, et al. 1998}, to help to get

all the possible optimal solutions since all the presented equations and constraints

tremendously multiply and grow themselves in quantity according to the different

problem conditions set up . Definitely, this has been a useful tool for assisting to get

the results otherwise by hand had been very difficult to solve problems with more

than 1,600 single equations (equations and constraints), more than 850 single

variables a more than 250 discrete variables per each example or case analysed in this

Thesis.

_____________________________________________________________________

35

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

4.1 Introduction.

The purpose of this chapter is to test the new model solving three out of four

examples used by {Lee, Pinto, et al. 1996} related with real industrial cases in order

to optimise the production scheduling of vessel unloading, storage, blending and feed

to crude distillation units process plants. The optimal results generated by the new

model with no omission of any data given by {Lee, Pinto, et al. 1996}are shown and

compared in advance with the optimal results of the existing model in table 4.1:

Examples

Items

1996} Model

New Model

Optimal Value

217.667

206.95

(US $ x 1,000)

Equations &

331

552

constraints

Single variables

192

337

Example 1

Discrete variables

36

116

Iterations

1,695

4,393

Solving

73.4

5.21

time(seconds)

Optimal Value

352.55

331.74

(US $ x 1,000)

Equations &

825

1807

constraints

Single variables

456

1138

Example 2

Discrete variables

70

396

Iterations

331,493

1,598,388

Solving time

69.3

104.7

(minutes)

Optimal Value

296.56

248.64

(US $ x 1,000)

Equations &

1222

2991

constraints

Single variables

581

1399

Example 3

Discrete variables

84

582

Iterations

> 515,541

4,236,122

Solving time

Not indicated

301.9

(minutes)

Table 4.1 - Comparisons of optimal results with an existing model.

_____________________________________________________________________

36

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

As it can be observed in table 4.1 all the optimal objective values given by the new

model as results, are lower than the existing model values. In accordance with the

mathematical explanation of the new model in chapter 3, the equations and constraints

handle by the new model really are more than the existing model s ones; but the

precision and the optimal value results of the new model are better than the existing

model because they have achieved more money savings solving the same operational

scheduling problem. Furthermore, the computational advance has avoided that the

machine time running of the model for each example has been much bigger

proportional to the new number of equations and constraints.

For all the three examples explained in detail forward, the given data for quantifying

volumes and flow rates are given in barrels x 10,000 and barrels per time interval x

10,000 respectively; changeover costs are given in US$ x 1,000; sea waiting costs and

unloading costs are given in US$ x 1,000 per time interval (day) and tank inventory

unit cost are given in US$ x 0.1 per oil barrel. Therefore, optimal value results will be

in US$ x 1,000.

4.2 Example 1 Results and Analysis.

For the example 1, fig 4.1 sketches the flow network and Table 4.2 shows the input

data given by {Lee, Pinto, et al. 1996} . The given data miss some information that is

necessary for solving the problem in accordance with the option a pointed out in

chapter 3. Then, for solve the problem was used a maximum and a minimum charging

rates to the CDU of values 50 and 10 respectively. This feed rate range can be

deduced in their graphs of results corresponding to the example 1 in fig 4.6, from the

right side indicated by them with the title Proposed Schedule. Furthermore, there is

not any information also required as input data with respect to the flow rate

restrictions among vessels and tanks and among tanks; however suitable values

indicated by (*) in table 4.2, have been chosen in accordance with the two days

observed for unloading the total crude oil volume per vessel and the size of the flow

network among tanks for example 1.

_____________________________________________________________________

37

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

CDU

Vessels

Storage

Tanks

Charging

Tanks

Scheduling Horizon (# of unit times: days)

Number of Vessels Arrivals

Arrival Time

Crude oil

Concentration

Vessel 1

1

100

0.01

Vessel 2

5

100

0.06

Number of Storage tanks

2

Storage Tanks

Capacity

Initial Oil

Key Component

Amount

Concentration

Tank1

100

25

0.01

Tank2

100

75

0.06

Number of Charging tanks

2

Charging Tanks

Capacity

Initial Oil

Initial ( min-max )

Amount

component

concentration

Tank1

100

50

0.02 (0.015- 0.025)

Tank2

100

50

0.05(0.045 - 0.055)

Number of CDUs

1

Unit costs involved in vessel operation Unloading cost: 8, Sea Waiting cost: 5

Tank inventory unit cost

Storage Tank: 0.08 , Charging Tank:0.05

Unit changeover cost for charged oil switch 50(independent of sequence and

CDU)

_____________________________________________________________________

38

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Blended oils demand from charging tanks to CDUs : Blended oil 1: 100,

blended oil 2: 100, blended oil 3: 100

*Maximum flow rate from vessel to one storage tank

50

* Maximum flow from one storage tank to one charging tank 40

* Maximum total flow rate from one storage tank to charging tanks at any time

interval

40

Table 4.2 - System information for Example 1.

Table 4.3 points out the comparisons of the optimal unloading starting up date results

from both models for example 1.Vessel 1 starts to unload on day 3 for the existing

model; that is one day more than new model which considers this operation on day 2.

Instead, for vessel 2 both coincide starting unloading on day 7.

Vessels

1996} Model

3

2

1

7

7

2

Table 4.3 Example 1. Optimal unloading starting date for vessels.

Figures 4.2, 4.3, 4.4, 4.5 show other optimal results from the new model for the

example 1 conditions. These results can be compared with the optimal results from

the existing model in fig. 4.6. Both results show two changeovers or feed switches to

the CDU. The new model presents the changeovers at days 2 and 5 and the existing

model at days 2 and 4 as it can be observed in fig. 4.4 and 4.6 respectively. Moreover,

despite the tank inventory costs per volume (input data in table 4.2) are higher for

storage tanks than for charging tanks there is not any explanation why the existing

model optimal behaviour for this example 1, shows higher crude oil volumes in

storage tanks than in charging tanks; optimally the effect should be in appositive

direction as show the new model results in fig.4.2 and 4.3. Then, this situation may be

causing a big part of the difference between both optimal values since the optimal

operational schedule of the existing model should be paying more money in inventory

total cost. On the other hand, the existing model results is paying one day more of sea

waiting cost (US$ 5,000) than new model results as is indicated in table 4.3.

_____________________________________________________________________

39

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

In summary, the optimal results for example 1 are better for the new model indicating

a total operational cost saving for the problem conditions of US$ 10,717 (4.92 % cost

reduction with respect to the existing model) during the 8 day scheduling horizon in

accordance with the results showed in table 4.1

Example 1

80

70

60

50

40

30

20

10

0

1

Storage Tank A

Storage Tank B

Example 1

120

100

80

60

40

20

0

1

Charging Tank X

Charging Tank Y

_____________________________________________________________________

40

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Example 1

60

50

40

30

20

10

0

1

Figure 4.4 Example 1. Optimal feeding to the crude distillation unit. Solid bars:

Blended oil X ; Blank bar: Blended oil Y.

Example 1

0.05

0.04

tanks

0.06

0.03

0.02

0.01

0

1

Charging Tank X

Charging Tank Y

tanks.

_____________________________________________________________________

41

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Figure 4.6 Example 1. Optimal results from the existing model (right side).

4.3 Example 2 Results and Analysis.

For the example 2, fig 4.7 sketches the flow network and Table 4.4 shows the input

data given by {Lee, Pinto, et al. 1996} . The given data also miss some information

_____________________________________________________________________

42

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

that is necessary for solving the problem in accordance with the option a pointed out

in chapter 3. Then, for solve the problem was used a maximum and a minimum

charging rates to the CDUs of values 20 and 4 respectively. This feed rate range can

be deduced from their graphs of results for the example 2 in fig.4.14 for both CDUs.

Furthermore, with respect to the flow rate constraints among vessels and tanks and

among tanks there is not any information also required as input data; however suitable

values indicated by (*) in table 4.4, have been chosen in accordance with the two days

observed for unloading the total crude oil volume per vessel and the size of the flow

network among tanks for example 2.

Example 2

CDU 1

Flow network

CDU 2

Vessels

Storage

Tanks

Charging

Tanks

Table 4.5 points out the comparisons of the optimal unloading starting date

recommended by both models for example 2. According to the existing model optimal

results, all vessels start to unload up to three days after the arriving day. Instead, the

new model optimal results consider starting unloading the same arriving days for the

first two vessels and allow only a delay of one day for vessel 3 before starting to

unload.

_____________________________________________________________________

43

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

10

Arrival

Amount of

Time

Crude oil

Vessel 1

Vessel 2

Vessel 3

3

Component 1

Component 2

100

0.01

0.04

100

0.03

0.02

100

0.05

0.01

Storage

Capacity

Tanks

Initial Oil

Component 1

Component 2

Amount

Tank 1

100

20

0.01

0.04

Tank 2

100

50

0.03

0.02

Tank 3

100

70

0.05

0.01

Charging

Capacity

Tanks

Initial Oil

Initial (min-max)

Initial(min-max)

Amount

comp.1

comp.2

concentration

concentration

Tank 1

100

30

0.0333(0.03 - 0.038)

Tank 2

100

50

Tank 3

100

30

Number of CDUs

Blended oil 1: 100,

blended oil 2: 100, blended oil 3: 100

*Maximum flow rate from vessel to one storage tank

50

* Maximum flow from one storage tank to one charging tank 60

* Maximum total flow rate from one storage tank to charging tanks at any time

interval

70

Table 4.4 - System information for Example 2.

_____________________________________________________________________

44

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Vessels

1996} Model

3

1

1

7

4

2

9

8

3

Table 4.5 Example 2. Optimal unloading starting date for vessels.

Figures 4.8, 4.9, 4.10, 4.11, 4.12 and 4.13 show other optimal results from the new

model for example 2. These results can be compared with the optimal results from

the existing model in fig. 4.14.

The results of both models show three changeovers in total for both CDUs: The new

model presents one at day 4 for CDU 1 and two at days 2 and 7 for CDU 2.Instead the

existing model presents two at days 3 and 7 for CDU 1 and one at day 2 for CDU 2.

Changeovers can be observed in fig. 4.10 and 4.11 for the new model and 4.14 for the

existing model. In general, the volume variation behaviour in both storage and

charging tanks are similar; however there are some differences in magnitude of

volume managed. On the other hand, the existing model schedule results are paying

six days of sea waiting costs that means five days more (US$ 25,000) than the new

model schedule results.

In summary, the optimal results for example 2 are better for the new model indicating

a total operational cost saving for the problem conditions of US$ 20,810 (5.90% cost

reduction with respect to the existing model) along the 10 days scheduling horizon in

accordance with the results showed in table 4.1. Regarding the new model results, it is

possible that part of the US$25,000 saved in sea waiting cost have been spent in a

higher total tank inventory cost paid due to the early unloading.

_____________________________________________________________________

45

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Example 2

120

100

80

60

40

20

0

1

10

Storage Tank 1

Storage Tank 2

Storage Tank 3

Example 2

120

100

80

60

40

20

0

1

10

Charging Tank 1

Charging Tank 2

Charging Tank 3

_____________________________________________________________________

46

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Example 2

25

20

15

10

0

1

10

Figure 4.10 Example 2. Optimal feeding to the CDU 1. Dashed bars: blended

oil 2; solid bars: blended oil 3.

Example 2

25

20

15

10

0

1

10

Figure 4.11 Example 2. Optimal feeding to the CDU 2. Blank bars: blended oi1

1; Dashed bars: blended oil 2; solid bars: blended oil 3.

_____________________________________________________________________

47

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Example 2

0.05

0.04

tanks

0.06

0.03

0.02

0.01

0

1

10

Charging Tank 1

Charging Tank 2

Charging Tank 3

charging tanks.

Example 2

0.03

0.025

tanks

0.035

0.02

0.015

0.01

0.005

0

1

10

Charging Tank 1

Charging Tank 2

Charging Tank 3

charging tanks.

_____________________________________________________________________

48

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

4.4 Example 3 Results and Analysis.

For the example 3 fig 4.15 sketches the flow network and Table 4.6 shows the input

data given by {Lee, Pinto, et al. 1996}. The given data also miss some information

_____________________________________________________________________

49

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

that is necessary for solving the problem in accordance with the option a pointed out

in chapter 3. Then, for solve the problem was used a maximum and a minimum

charging rates to the CDUs of values 10 and 2 respectively. This feed rate range can

be deduced from their graphs of results for the example 2 in fig.4.20 for both CDUs.

Furthermore, with respect to the flow rate constraints among vessels and tanks and

among tanks there is not any information also required as input data; however suitable

values indicated by (*) in table 4.6, have been chosen in accordance with the two days

observed for unloading the total crude oil volume per vessel and the size of the flow

network among tanks for example 3.

Flow network

Example 3

CDU 1

CDU 2

Vessels

Storage

Tanks

Charging

Tanks

Table 4.7 points out the comparisons of the optimal unloading starting date

recommended by both models for example 3. Both models coincide with the same

optimal unloading day for each vessel recommended for the same arriving day of each

one.

_____________________________________________________________________

50

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Number of Vessels Arrivals

Arrival Time

12

3

Amount

Crude oil

50

50

50

of Key

Component

Concentration

Vessel 1

1

0.01

Vessel 2

5

0.085

Vessel 3

9

0.06

Number of Storage tanks

3

Storage Tanks

Capacity

Initial Oil

Initial ( min-max )

Amount

component

concentration

Tank1

100

20

0.02 (0.01- 0.03)

Tank2

100

20

0.05(0.04 - 0.06)

Tank3

100

20

0.08(0.07 0.09)

Number of Charging tanks

3

Charging Tanks

Capacity

Initial Oil

Initial ( min-max )

Amount

component

concentration

Tank1

100

30

0.03 (0.025- 0.035)

Tank2

100

50

0.05(0.045 - 0.065)

Tank3

100

30

0.08(0.075 0.085)

Number of CDUs

2

Unit costs involved in vessel operation Unloading cost: 10, Sea Waiting cost: 5

Tank inventory unit cost

Storage Tank: 0.04 , Charging Tank:0.08

Unit changeover cost for charged oil switch 50(independent of sequence and CDU)

Blended oils demand from charging tanks to CDUs :

Blended oil 1: 50,

blended oil 2: 50, blended oil 3: 50

*Maximum flow rate from vessel to one storage tank

25

* Maximum flow from one storage tank to one charging tank 40

* Maximum total flow rate from one storage tank to charging tanks at any time

interval

40

* Maximum total flow rate from vessel to storage tanks at any time interval 25

Table 4.6 - System information for Example 3.

Vessels

1996} Model

1

1

1

5

5

2

9

9

3

Table 4.7 Example 3. Optimal unloading starting date for vessels.

_____________________________________________________________________

51

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Figures 4.16, 4.17, 4.18 and 4.19 show other optimal results from the new model for

example 3. These results can be compared with the optimal results from the existing

model in fig. 4.20.

The optimal results of the new model present only two changeovers, one at day 4 for

CDU 2 and another at day 9 for CDU 1. Instead, the existing model optimal results

present three changeovers; one more (US$ 50,000) than the new model ; two at days 4

and 8 for CDU 1 and one at day 3 for CDU 2. Changeovers can be observed in fig.

4.18 and 4.19 for the new model and 4.20 for the existing model. In general, the

volume variation behaviour in both storage and charging tanks are similar; however

there are some differences in magnitude of volume managed along the scheduling

horizon. On the other hand, both results have not considered to pay sea waiting costs

as is indicated in table 4.7.

In summary, the optimal results for example 3 are better for the new model indicating

a total operational cost saving for the problem conditions of US$ 47,920 for the 12

days scheduling horizon in accordance with the results showed in table 4.1. It is

possible, regarding the new model results that part of the US$50,000 saved in one

changeover less, has been spent in a little higher total tank inventory cost paid for

achieving this optimal objective.

_____________________________________________________________________

52

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Example 3

90

80

70

60

50

40

30

20

10

0

1

10

11

12

Storage Tank 1

Storage Tank 2

Storage Tank 3

Example 3

60

50

40

30

20

10

0

1

10

11

12

Charging Tank 1

Charging Tank 2

Charging Tank 3

_____________________________________________________________________

53

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Example 3

12

10

8

6

4

2

0

1

10

11

12

Figure 4.18 - Example 3. Optimal feeding to the CDU 1. Dashed bars: blended oil

2; solid bars: blended oil 3.

Example 3

12

10

8

6

4

2

0

1

10

11

12

Figure 4.19 - Example 3. Optimal feeding to the CDU 2. Blank bars: blended oil

1; solid bars: blended oil 3.

_____________________________________________________________________

54

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

4.5 Chapter Summary

This chapter showed the test of the new model in detail solving the same examples

used by {Lee, Pinto, et al. 1996} when they presented their model; also called in this

Thesis, existing model for distinguishing from the new model. Both model results

have been compared in detail pointing out the differences in the results. Although the

new model manages more equations and constraints, its optimal values with respect to

minimise the total operational cost have been lower than the optimal values presented

_____________________________________________________________________

55

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

by the existing model; moreover, for the three examples together the optimal results

from the new model achieved a total difference of US$ 79,447 less costly compared

with the existing model optimal results for the same operational conditions.

Furthermore, it has been demonstrated that the features of the new model design allow

that it can be adapted to different demand conditions included

the condition

established by {Lee, Pinto, et al. 1996}for their examples. Instead, the existing model

was developed to work in only one solving direction and although, most of its results

shown in this Thesis apparently go in the same optimal path, it is less precise than the

new model solving these kind of scheduling problems.

One of the main reasons of the existing model shortcomings is that it has missed some

important equations and constraints and moreover, the formulation of some of their

existing equations and constraints are different compared with the new model as it

was well explained in chapter 3. For instance, regarding the results of example 2, it is

clear the lack of the constraints 3.7 and 3.8 for control the maximum departure date

of previous vessels in accordance with the arriving date of next vessels for avoiding

unnecessary long sea waiting for vessels. On the other hand, with respect to the total

flow constraint, {Lee, Pinto, et al. 1996} do not show any flow rate data from their

optimal results for comparing; moreover, it is clear in its model formulation that they

did not use any constraint to control total flows; however, it is difficult to evaluate if

the lack of these constraints also could have caused some additional mistakes in their

optimal results. Finally, in accordance with the mistake found in the objective

function s formulation of the existing model, with respect to the unloading cost

summation within the total cost; they could have left out to charge three unloading

day cost for each one of their examples: US$ 24,000 for examples 1 and 2 and US$

30,000 for example 3; unless they could have taken in account them in another form

not clear in their model formulation because these differences are appreciable

quantities of money that would hugely impact their final optimal values shown in

table 4.1.

_____________________________________________________________________

56

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

The new model will be applied to several cases related with one oil refinery in this

chapter, including also chapter 6 and 7; each case analysed has its own features and

conditions that facilitate more the understanding of the advantages of the application

of this new optimisation model to similar cases in the industry.

Particularly the oil refinery for what the cases are developed has the flow network

configuration indicated in fig.5.1 to manage all its operation of crude oil unloading,

storage, blending and feeding to its crude distillation unit. Furthermore, all the cases

will operate in a scheduling horizon of 15 days and the solving option used is b

according to the classification explained in chapter 3.

Flow network

CDU

Charging

Tanks

Vessels

Storage

Tanks

Figure 5.1 Oil Flow Next Work for the Oil Refinery.

_____________________________________________________________________

57

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

At the planning stage three crude oil vessels will arrive at days 1, 6 and 10

respectively and their unloading should be completed by the day 15.Each vessel

contain 400,000 bbl of crude oil Alpha, Beta and Ro, respectively. The refinery has

one CDU which has a design capacity of 82,000 bbl per day of crude oil. The two

charging tanks have blended crude oils X and Y that should be used to charging the

CDU no matter whether it is either crude oils X or Y; but however the refinery is

concerned about what kind of crude oil is charging with higher or lower sulphur

content to determine the proper sulphur neutralization control downstream. The range

of the blended composition of both charging tanks has been determined at the

planning level. The main concern is to optimise the operational cost keeping a feed

demand of 75,000 bbl per day to the crude distillation unit. Feeding less than this

amount is not allowed for fulfilling the planning targets and avoiding plant shutdown

or malfunction of its equipment operating under its normal working capacity;

however, an increase of the feed is allowed up to 82,000 bbl per day only if the

optimal results indicate it. This is considered because the oil refinery allows a total

overproduction of maximum 10% with respect to the total feed demand of 1,125,000

bbl along the scheduling horizon for all the cases analysed.

The weight fractions of sulphur (component k) which determine the quality of crude

oil are 0.01 for crude oil Alpha, 0.03 for crude oil Beta and 0.05 for crude oil Ro.

These three crude oils are blended to make two types of blending: blended crude oils

X and Y. The sulphur concentration range of X should be between 0.01 and 0.02 and

Y between 0.035 and 0.045.The storage tanks have a maximum capacity of 350.000

bbl each and the initial crude oil volumes of these tanks for crude oil Alpha, Beta and

Ro are 30,000 bbl, 200,000 bbl and 140,000 bbl, respectively. On the other hand,

charging tanks have a maximum capacity of 250,000 bbl each and the initial crude oil

volumes of these tanks for crude oils X and Y are 70,000 bbl and 190,000 bbl

respectively.

_____________________________________________________________________

58

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

The operation cost involved in this problem are the inventory cost, the vessel

harbouring cost, the vessel sea waiting cost and the CDU changeover cost charged

each time the feed is switched from crude oils X to Y or vice versa.

A sketch of the process was shown in fig.5.1. The arrows indicate the flow transfer

among equipments per interval of time. The flow transfer could be null between two

particular equipments in any particular time period in accordance with the optimal

results.

Vessel 1 arrives at the docking station at day 1 and from this day it could start to

unload crude oil Alpha into storage tank 1; it should complete unloading before day 6

and leave the docking station at most the day 6. At each time interval, crude oil can be

transferred from storage tanks and blended into charging tank X and charging tank Y.

Then, blended crude oil either X or Y charges the CDU. Vessel 2 arrives at day 6 and

from this day it could start unloading crude oil Beta into storage tank 2; it should

complete unloading before day 10 and leave the docking station at most the day 10.

Finally, Vessel 3 arrives at day 10 and from this day it could start unloading crude oil

Ro into storage tank 3; it should complete unloading either before or on day 15 and

leave the docking station at most the day 15.

Unit inventory cost for each storage tank and each charging tank are 8 x 10-3 and 5 x

10-3 US$/(day x bbl), respectively. The cost causes by the feed switch to the CDU

(changeover cost) is US$50,000 each time it occurs. Costs involving vessels are

caused by waiting in the sea corresponding to US$ 8,000 per day and by harbouring

for unloading the crude oil corresponding to US$ 10,000 per day. The unloading

incurs in higher cost.

Table 5.1 summarises all the above mentioned input data plus some other important

information that also will be taken in account for solve the case 1 problem using the

model.

_____________________________________________________________________

59

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Number of Vessels Arrivals

Arrival Time

Vessel 1

Vessel 2

Vessel 2

Number of Storage tanks

Storage Tanks

Tank1

Tank2

Tank3

Number of Charging tanks

Charging Tanks

Tank1

Tank2

Day 1

Day 6

Day 10

Maximum/

Minimum

Capacity

350,000 bbl /

5,000 bbl

350,000 bbl /

5,000 bbl

350,000 bbl /

5,000 bbl

Maximum/

Minimum

Capacity

250,000 bbl /

15,000 bbl

250,000 bbl /

15,000 bbl

15 days

3

of Sulphur Content

Amount

Crude oil

400,000 bbl

400,000 bbl

400,000 bbl

3

Initial Oil

Amount

0.01

0.03

0.05

Sulphur Content

30,000 bbl

0.01

200,000 bbl

0.03

140,000 bbl

0.05

2

Initial Oil

Amount

70,000 bbl

190,000 bbl

Range of Sulphur

Content & Initial

value

0.01-0.02

0.0158

0.035 0.045

0.0389

Number of CDU

1

Maximum flow from vessel to storage tank.

250,000 bbl/day

Maximum flow from one storage tank to charging

70,000 bbl/day

tanks simultaneously during any time period.

Maximum flow from one storage tank to one

65,000 bbl/day

charging tank.

Maximum flow from one charging tank to one CDU.

82,000 bbl/day

Cost involved in vessel operation

Unloading: US$ 10,000/day

Sea Waiting: US$ 8,000/day

Tank inventory cost

Storage tank: US$ 0.008 /(day x

bbl)

Charging tank:US$0.005/(day x

bbl)

Changeover cost for feed switch

US$50,000

CDU crude oil demand

75,000 bbl/day

Table 5.1 - Case 1. Input data summary.

After applying the model for the above case the following results are highlighted:

_____________________________________________________________________

60

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

The optimal operation cost for the 15-day scheduling horizon of case 1 is

US$338,964.3.

Table 5.2 indicates the results regarding the suitable days for starting and finishing

unloading each vessel along the scheduling horizon:

Vessel

Starts unloading

Finish unloading

Day 1

Day 2

Day 6

Day 7

Day 11

Day 12

Table 5.3 indicates the crude oil volume variation results for each vessel after starting

unloading during the scheduling horizon. Flow rates from vessels to storage tanks

shown in table 5.4 help to understand table 5.3.

Vessel

1

2

3 4 5 6

400

250

0

400

8 9 10

250

11

12

13

400

250

14

15

Fig.5.2 indicates the optimal volume variation for each storage tank at each time

interval (day). It accordance with the arriving and the unloading of each vessel, it is

possible to see the influence of vessel 1 feeding the storage tank 1, vessel 2 feeding

the storage tank 2 and vessel 3 feeding the storage tank 3.

_____________________________________________________________________

61

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

400

Crude bbl x 1000

350

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

Fig 5.3 indicates the optimal volume variation at each time interval (day) for each

charging tank. It clearly indicates that when one tank is feeding the CDU, the other

tank is being fed or refilled by the crude oil from the storage tanks. Moreover, it can

be observed at the top of the graphic that in several cases the charging tank that is

_____________________________________________________________________

62

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

being refilled gets its maximum volume before the other tank has been consumed to

make the feed switch. This is a clear demonstration that charging tanks have

restriction in its volume size that it is part of the conditions of case 1. The minimum

crude oil volume for each charging tank was fixed as input condition in 15,000 bbl to

guarantee 4 hours minimum of feeding operation to the CDU if a minor emergency

happens in the system analysed upstream the charging tanks. Some refineries work

having a crude oil reserve up to 25% of the tank maximum capacity.

Fig 5.4 indicates the optimal feeding to the CDU per each time interval during the

scheduling horizon. The 75,000 bbl feed demand should be given at days 5, 8, 11 and

14. On the other hand, the 82,000 bbl maximum feed could be reached at days

1,2,3,6,9,12 and 15. The optimal total amount of crude oil proposed to feed the CDU

for case 1 during the 15 day scheduling horizon is 1,186,000 bbl (5.42 % over the

total feed demand). The feed switches (mode changeovers) are four (4) given at days

3, 6, 9 and 12 respectively and figs 5.2 and 5.3 clearly show this condition.

84

82

80

78

76

74

72

70

1

10 11 12 13 14 15

Figure 5.4 - Case 1. Feeding to the crude distillation unit results. Solid bars:

blended oil 1; blank bars: blended oil 2.

Table 5.4 indicates for case 1 the optimal volumetric flow rate at each time interval

(day) from vessels to storage tanks.

_____________________________________________________________________

63

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Vessel 1

1

3 4 5 6

8 9 10 11

12

13 14 15

150 250

150 250

150 250

Table 5.4 - Case 1. Results for volumetric flow rate (bbl x 1,000) per day from

vessels to storage tanks.

Table 5.5 also respectively indicates the optimal flow rate results from the storage

tanks to the charging tanks. The left side column of this table shows the sense of the

flow from determined storage tank to the indicated charging tank. In some cases is

shown that one charging tank is being fed by two or three storage tanks in the same

time interval. These amounts of flows are optimally calculated to be blended keeping

the sulphur content range for each charging tank and the maximum flow rate allowed

from one storage tank to several charging tanks or from one storage tank to one

charging tank. Specifically, this system (fig.4.1) never will present simultaneous flow

towards both charging tanks at the same time interval (day) because one of them

should be feeding the CDU at the same time interval.

St.T-

65

48.5

5 6

8 9

10

11 12 13 14 15

Ch.T

1-1

1-2

2-1

25.5

3-2

65 50

18.75

65

2-2

3-1

65 65

65 40

65

65

65 50

65

36.25

1.5

5

65

3.5

65

50

Table 5.5 - Case 1. Results for volumetric flow rate (bbl x 1,000) per day from

storage tanks to charging tanks.

Moreover, table 5.5 shows the optimal total amount of crude oil that should optimally

be transferred during the scheduling horizon, from the storage to the charging tanks to

_____________________________________________________________________

64

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

balance the CDU crude oil demand which for this case 1 is 1,109,000 bbl. This

quantity is lower than the total crude oil optimally fed to the CDU during the

scheduling horizon by 77,000 bbl. Despite the charging tanks starts at day 1 with a

volume bigger than their minimum volume, it could have expected that the flow

transferred were much bigger than the presented in the results helped by the lower

inventory cost applied to the charging tanks compared with the storage tank inventory

cost. However, the main restriction for this case 1 that may have not helped at all for

the expected optimal solution may be the limit of capacity of each charging tank that

it is reached each time the tank is refilled during the scheduling horizon.

Fig. 5.5 indicates the optimal fluctuation for the sulphur content in the crude oil of

each charging tank at any time interval (day) a long the scheduling horizon. It could

be identified that the sulphur content in both charging tanks is kept inside the range

established at the planning level. Some special findings could come out studying the

characteristics of the behaviour above shown, i.e. for charging tank 2, the sulphur

content is kept at the range lowest level during days 5 to 10 until the vessel 3 with the

charging tanks

0.05

0.045

0.04

0.035

0.03

0.025

0.02

0.015

0.01

0.005

0

1

10 11 12 13 14 15

Charging Tank 1 Sulphur content

Figure 5.5 - Case 1. Crude oil sulphur content variation results in charging

tanks.

_____________________________________________________________________

65

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

The following question could arise: What could happen if it is informed that the

vessel arriving schedule has changed just after the monthly plan already have been

done and have got the approval at the managerial level? It is possible to do something

at the scheduling level in order to perform again the optimization techniques as they

were applied for the first arriving schedule presented in case 1?. The answers to these

questions are very important if the problem is relating to keeping the minimum

operational cost and hence, a good profit margin. On the other hand, several

infeasibilities could present depending of the dimension of the scheduling

modifications according to the system current conditions i.e. it could be not possible

to feed the CDU at least the demand rate per day because one vessel is delayed too

much; the sulphur content for the determined charging tank could not match to be

inside the range; it could not have either enough capacity or enough start up crude oil

volume in one of the storage tanks and so on. These infeasibilities have some

difficulties to detect in an early step at the planning level.

Table 5.6 shows the new proposed arriving schedule for case 1a (same case 1

including the modifications of the arriving schedule at the late stage):

Vessel Arrival Time Amount of Crude oil Sulphur Content

1

Day 1

400,000 bbl

0.01

Day 4

400,000 bbl

0.03

Day 11

400,000 bbl

0.05

As it is can be deducted from this new arriving schedule the second vessel will arrive

two days earlier and the third vessel will arrive one day later with respect to the

original arriving schedule approved at the planning level for case 1.

_____________________________________________________________________

66

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

After applying the model for the above mentioned case 1a, the following results are

highlighted:

The optimal operational cost for the 15 day scheduling horizon of the new schedule,

case 1a is US$337,432.5. It is US$ 1,531.75 less costly than the original case.

Fortunately, the total operational cost has been reduced but it could have not been if

actions had not taken on time to correct and optimise the operational actions having in

account this change of schedule.

Table 5.7 indicates the suitable days for starting and finishing unloading in the

scheduling horizon of each vessel:

Vessel

Starts unloading

Finish unloading

Day 1

Day 2

Day 4

Day 5

Day 11

Day 12

Fig.5.6 indicates the optimal variation at each time interval (day) for each storage tank

level. Comparing figs. 5.2 (original case) and 5.6, it can be seen the following main

differences regarding case 1a:

Storage tank 2 does not reach the minimum level because it is started to be fed

2 days earlier.

Storage tank 1 is assisted earlier by the storage tank 2 to do the blending and

hence, this tank 1 conserves a little more its crude oil volume during the

schedule horizon.

Fig 5.7 indicates for case 1a, the optimal volume variation at each time interval (day)

for each charging tank. It is similar to case 1 shown in fig.5.3; except for charging

tank 2 which does not have any dead time in full condition.

_____________________________________________________________________

67

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

400

Crude bbl x 1000

350

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

Fig. 5.8 indicates the optimal fluctuation for the sulphur content in the crude oil of

each charging tank at any time interval (day) a long the scheduling horizon. For

charging tank 1 the sulphur content is kept more stable showing a sulphur

_____________________________________________________________________

68

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

concentration average of 0.02 (range maximum value) along the scheduling horizon

since day 3 except when the tank is consumed to the minimum volume at days 6 and

12. This situation may be presented by the early arriving effect of vessel 2 with a

crude oil sulphur concentration of 0.03 that partially is blended with crude oil

charging tanks

0.05

0.045

0.04

0.035

0.03

0.025

0.02

0.015

0.01

0.005

0

1

10 11 12 13 14 15

Charging Tank 1 Sulphur content

Figure 5.8 - Case1a .Crude oil sulphur content variation results in charging

tanks.

For this case 1a, the optimal number of feed switches, the switch schedule and the

optimal feeding for the CDU for each time interval during the scheduling horizon are

also the same shown above in fig. 5.4 for case 1.

Other changes of arriving schedule could be considered only if the optimal solution

continues being feasible for this kind of problem size. It could be considered a

different schedule change case 1b (same case 1 including the arriving schedule

modifications at the late stage) in accordance with table 5.8:

Vessel Arrival Time Amount of Crude oil Sulphur Content

Day 1

400,000 bbl

0.01

1

Day 7

400,000 bbl

0.03

2

Day 9

400,000 bbl

0.05

3

Table 5.8 - Case 1b. Second change of the arriving schedule.

_____________________________________________________________________

69

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

As it is can be deducted from this new arriving schedule vessel 2 will arrive one day

later and vessel 3 will arrive on day earlier with respect to the original arriving

schedule approved at the planning level (case1).

After applying the model for the above mentioned case 1b, the following results are

highlighted:

The optimal operational cost for the 15 day scheduling horizon of this case 1b is

US$333,921.9. It is US$ 5,042.35 less costly than the original case.

Vessel

Starts unloading

Finish unloading

Day 1

Day 2

1

Day 7

Day 8

2

Day 9

Day 10

3

Table 5.9 - Case1b. Unloading start and finish results.

Table 5.9 indicates for this new arriving schedule alteration the suitable days for

starting and finishing unloading of each vessel.

Fig.5.9 indicates the optimal variation at each time interval (day) for each storage tank

volume. If it is compared the figures 5.2 (case 1) and 5.9 it can be observed the

following main differences regarding case 1b:

Storage tank 3 gets the minimum volume faster because it may have assisted

storage tank 2 for not being consumed before day 7.

At the final of the scheduling horizon it could be noted that the storage tank 3

was more consumed than the storage tank 2.

_____________________________________________________________________

70

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

400

Crude bbl x 1000

350

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

Fig 5.10 indicates the optimal volume variation at each time interval (day) for each

charging tank l. It is more similar to the original case shown in Fig.5.3.

_____________________________________________________________________

71

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

tanks

0.05

0.045

0.04

0.035

0.03

0.025

0.02

0.015

0.01

0.005

0

1

5 6 7 8 9 10 11 12 13 14 15

Scheduling horizon in days

Figure 5.11 - Case1b. Crude oil sulphur content variation results in charging

tanks.

Fig. 5.11 indicates the optimal fluctuation for the sulphur content in the crude oil of

each charging tank at any time interval (day) a long the scheduling horizon. The

sulphur content of charging tank 2 is more unstable, but it is kept inside the range.

For this case 1b, the optimal number of feed switches, the switch schedule and the

optimal feeding for the CDU for each time interval during the scheduling horizon are

also the same shown above in fig. 5.4 for case 1.

5.3 Bigger Scheduling Alteration of Case 1.

Returning again to the case 1 conditions and the vessel arriving schedule changes at

the last day, it is possible to solve a more difficult problem of schedule change as is

indicated in Table 5.10.

As it is can be observed from this new arriving schedule now vessel 2 will arrive first

at day 1 and vessel 1 will arrive second at day 6. This is a big change with respect to

_____________________________________________________________________

72

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

the original arriving schedule approved at the planning level (case1) and it sometimes

could happen in the oil refinery.

Vessel

2

Arrival

Time

Day 1

Amount

of Sulphur

Crude oil

Content

400,000 bbl

0.03

Day 6

400,000 bbl

0.01

Day 10

400,000 bbl

0.05

After the model has been applied it is found that there is no solution for this case. The

current conditions have caused infeasibilities in the system and therefore, it is not

possible to obtain optimal operational cost solution. After reviewing the problem

making some iteration it was discovered that the start up volume of storage tank 1 is

not enough for waiting vessel 1 until day 6 with the conditions presented in the

system. Then, for solve this situation; storage tank 1 should start with an initial crude

oil volume of at least 80,000 bbl and not 30,000 bbl as it originally starts for solving

case 1.

Then, considering a volume of 80,000 bbl for storage tank 1, the optimal operational

cost for the 15 days scheduling horizon of this case 1c is US$354,181.5. It is

US$15,217.25 most costly than the original case 1. Frequent problems of this type

could cause losses of more than US$ 350,000 per year even considering that the

optimal operational costs are kept. Moreover, it is important to highlight that for

calculation of this new total optimal operational cost for this case 1c, the cost of the

operation for transfer the 50,000 bbl (0.01 sulphur) from somewhere in the oil refinery

to the storage tank 1 for completing its required start up volume has not been taken in

account.

Table 5.11 indicates the suitable days for starting and finishing unloading in the

scheduling horizon of each vessel:

_____________________________________________________________________

73

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Vessel

Starts unloading

Finish unloading

Day 4

Day 5

2

Day 6

Day 7

1

Day 10

Day 11

3

Table 5.11 - Case 1c. Unloading start and finish results.

Table 5.12 indicates each vessel crude oil volume variation after starting unloading

during the scheduling horizon.

As it could be observed from both tables 5.11 and 5.12, although vessel 2 arrives at

day 1, the optimal day for start unloading is at day 4.

Vessel 1 2 3 4

5

6

7

8 9 10 11 12 13 14 15

400 250 0

2

400 250 0

1

400 250 0

3

Table 5.12 - Case 1c. Vessel volume variation (bbl x 1000) results.

Figs 5.12 and 5.13 show the storage and the charging tank optimal volume variation

respectively for case 1c. Storage tank 1 keeps at the minimum level for four (4) times

intervals while vessel 1 arrives and starts unloading.

400

Crude bbl x 1000

350

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

_____________________________________________________________________

74

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

The charging tank volume variation (fig. 5.13) has minor alterations compared with

fig.5.4 for case 1. Also as it can be observed in fig. 5.14 regarding the feed to the

CDU and compared with fig. 5.3, case 1c has an optimal feed rate of 77,400 bbl the

day 4. It means for case 1c, that the total feed amount is depleted by 600 bbl along the

schedule horizon compared with case 1; but, it is still 5.37 % over the total feed

demand.

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

84

82

80

78

76

74

72

70

1

10 11 12 13 14 15

Figure 5.14 - Case 1c. Feeding to the crude distillation unit results. Solid bars:

blended oil 1; blank bars: blended oil 2.

_____________________________________________________________________

75

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Table 5.13 also respectively indicates the optimal flow rate from storage tanks to the

charging tanks for case 1c. If tables 5.13 and 5.5 are compared, storage tank 1 from

case 1c transfers the volume for blending that it can in the first two time intervals

(days) and its transfer re starts at day 6 when is being feeding by the vessel.

Stor.T- 1

8 9

10 11 12 13 14 15

Char.T

65 10

65 65

65 65

1-1

32

1-2

65 39.4

65 40

65 35

2-1

65 65 24

65 8

2-2

5

3-1

65 5

65 65

3-2

Table 5.13 - Case 1c. Results for Volumetric flow rate (bbl x 1,000) per day from

charging tanks

0.05

0.045

0.04

0.035

0.03

0.025

0.02

0.015

0.01

0.005

0

1

10 11 12 13 14 15

Charging Tank 1 Sulphur content

Figure 5.15 - Case1c. Crude oil sulphur content variation results in charging

tanks.

Fig. 5.15 shows the optimal sulphur content variation for both charging tanks. If

Figs.5.15 and 5.5 (case 1) are compared, it is possible to observe for case 1c that

during the first 8 days of the scheduling horizon, the sulphur content for both tanks in

_____________________________________________________________________

76

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

average is bigger that the same presented for case 1 in the same period, due to the

influence of having a vessel of mayor sulphur content unloading earlier.

Finally, it is interesting to show in table 5.14, other results considering other starting

up volumes for storage tank 1. These results were obtained while it was done the trial

and error exercise to find its suitable minimum starting up crude oil volume.

Storage tank

1 crude oil

150,000

80,000 bbl

81,000 bbl

82,000 bbl

85,000 bbl

optimal

US$

US$

US$

US$

US$

operational

354,181.5

354,235.3

354,341.3

354,659.3

361,900.3

- 600 bbl

- 7,000 bbl

0 bbl

0 bbl

0 bbl

95,934

1,368,463

942,089

126,339

135,000

volume

bbl

Total

cost

CDU total

feed

difference

with respect

to case 1

Iterations

Software

time spent

running

(20.7 min)

(28.9 min)

Table 5.14 - Case 1c. Results & comparisons among different crude oil volumes

for storage tank 1.

Results shown in table 5.14 help to decide that it is better to start up the storage tank 1

with 82,000 bbl, instead 80,000 bbl. For US$ 159.6 more costly, there are 600 bbl

more to be optimally fed during the scheduling horizon starting up with 82,000 bbl.

_____________________________________________________________________

77

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Logically, each time that the feed switch is done the plant manager knows if either a

higher or a lower sulphur content crude oil is feeding the CDU and this helps to do the

proper control to neutralise the sulphur downstream basically in accordance with the

sulphur content range in the charging crude oil. The products usually used for

neutralising and avoiding damage to the equipment and piping as a result of the

sulphur compound attack are very expensive and hence, it is strongly desirable to

optimally regulate their consumption to save money. But what could happen if

somebody at the managerial level believes that control the sulphur range during the

blending in charging tanks increase a lot the total operational costs and therefore he

wants to try not controlling anymore the sulphur content in the charging tanks and

hence, charging tanks only should concern about having enough blended crude oil to

feed the CDU not matter their sulphur content.

This is a good question to answer and the conditions of case 1 will be taken to

analysis this new case 1d, except that the control of sulphur content range in charging

tanks will not be taken in account. However, each vessel continues unloading to the

same assigned storage tank.

After applying the optimisation techniques for case 1d the following result are

highlighted:

The optimal operation cost for the 15 day scheduling horizon of this case is

US$333,566.5. It is US$ 5,397.75 less costly than case 1.

The optimal unloading start and finish and the vessel volume variations are shown in

table 5.15 for case 1d. Particularly, all vessels optimally unload the same arriving day.

_____________________________________________________________________

78

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Start

Finish

Day 1

Day 1

Day 2

400,000 bbl

1

2

Day 6

Day 6

Day 7

400,000 bbl

Day 10

Day 10

Day 11

400,000 bbl

Figs.5.16 and 5.17 indicates the optimal volume variation for storage and charging

tanks respectively at each time interval (day). If figs 5.2 and 5.3 from case 1 are

compared with figs.5.16 and 5.17, it may seem for case 1d that all storage tanks have

the same priority to be consumed according to their available crude volume at each

time interval; even the storage tank 3 is more consumed than case 1. The phenomenon

for charging tanks for both cases are similar; they mainly differ in that for case 1d,

each charging tank when has the chance to be refilled, this operation is made using the

maximum possible rate up to 65,000 bbl per day depending of the volume they need

to complete its maximum crude oil volume. This phenomenon also could be observed

in table 5.15, compared with table 5.5 for case 1.

400

Crude bbl x 1000

350

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

_____________________________________________________________________

79

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

The optimal feeding, the number of feed switches and the switch schedule are the

same presented in fig. 5.3 for case 1, except that the optimal feed is depleted the last

day of the period from 82,000 bbl to 75,000 bbl; but it still is 4.8% over the total feed

demand. Apparently removing the sulphur content constraint for charging tanks could

have allowed transferring more crude oil from storage to charging tanks but for this

case 1d the total optimal amount transferred along the schedule horizon (table 5.15) is

1,109,000 bbl just the same presented for case 1.

Stor.TChar.T

5 6

8 9

10 11 12 13 14 15

65 50

65 40

65

1-1

65

65

1-2

65

65

65

2-1

65 64

65

2-2

65

65 40

3-1

30

40 65

3-2

Table 5.15 - Case 1d. Results for volumetric flow rate in bbl x 1000 per day from

storage to charging tanks.

_____________________________________________________________________

80

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Finally, as it was observed with the optimal results presented, case 1d basically does

not present major additional advantages compared with case 1 (only a total optimal

cost reduction of 1.6 % compared with case 1). Instead, the plant would be losing the

optimal control of the sulphur neutralisation product consumption and hence, the

money spent in this product could increase due the unnecessary waste.

5.5 Effect of Not having Tank Inventory Cost for Case 1.

The tank inventory costs are related with the cost that the refinery has to pay for the

average crude oil volumes that manage each tank between interval of times (days).

The above analysis of case 1 and its different modification and discussions have

shown that the optimal results have taken in account the managing of the lowest

optimally possible total crude oil volumes in tanks along the scheduling horizon to

reasonably reduce the impact of this cost. A test was done considering no inventory

cost for case 1 and therefore, the following response is part of the optimal results after

the model was applied:

The optimal cost for this case is US$ 260,000. It is clear, that they correspond to four

changeovers (US$ 50,000 each) plus two days unloading per vessel (US$ 20,000 per

vessel).The changeovers also have the same schedule presented for case 1 on days

3,6, 9 and 12.

In addition, other data to highlight of this exercise is the optimal vessel unloading rate

along the scheduling horizon through table 5.16.

Vessel 1

1

3 4 5 6

8 9 10

11

12 13 14 15

150 250 0

2

3

250 150 0

150 250 0

Table 5.16 - Case 1e. Results for volumetric flow rate (bbl x 1,000) per day from

vessels to storage tanks.

_____________________________________________________________________

81

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

It is important to comment that despite there was not inventory cost to charge to the

optimal operational cost only vessel 2 is able to optimally unload the maximum rate

the first unloading day. It may possible that the other vessels were not able because

either their assigned storage tanks had enough crude oil volume during the first

unloading day or the optimal solution had not taken care about that due to the null

inventory cost.

5.6 Effect Caused by a High Reduction in Changeover Cost for Case 1.

A test was made to check the behaviour of the optimal feed switch number for a

similar problem like case 1.The refinery manager would like to review the high

changeover cost; he believes that after doing a good analysis of this cost, it could go

down if some good practices are implemented. However, reducing the changeover

cost the optimal total operational cost could be reduced but the current optimal

number of changeovers may not. The analysis of this situation showed that the

number of changeovers has a hugely dependency of the tank capacities. The test was

done with the same conditions of case 1 and reducing the changeover cost from US$

50,000 to US$ 15,000. Then, the optimal final solution indicated four changeovers on

the days 3, 6, 9 and 12, the same optimally scheduled for case 1.

5.8 Chapter Summary.

Table 5.17 shows a summary of the resources used to solve most of the cases

reviewed in this chapter.

Case 1

Case 1a

Case 1b

Case 1c Case 1d

1,683

1,682

1,670

1,562

Single variables

863

864

863

863

861

Discrete variables

261

261

261

261

261

Iterations

397,133

443,748

421,902

95,934

1,344,495

Solving time

31.9 min. 86.0 min. 78.7 min. 4.0 min. 228.97 min.

_____________________________________________________________________

82

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

A practical case has been reviewed in this chapter. The optimisation model developed

in this Thesis has been applied to this practical case, including the analysis of the

scheduling and operational cost impact when some practical modifications are applied

as they could happen in real situations i.e. vessel arrival scheduling change; inventory

cost alteration and so on. Solving the different alterations has shown that if the

conditions of the problem still are feasible, an optimal solution always could come out

for each real situation. Refinery managers would like to have this kind of support at

the scheduling level for avoiding waste of money controlling all the mentioned typical

changes that could happen during the operation of similar operating systems.

Finally, although this chapter smoothly mention the scheduling impact by the

equipment limitations, the next chapter 6 will deepen this topic.

_____________________________________________________________________

83

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

SYSTEM EQUIPMENT LIMITATIONS ARE HIGHLIGHTED.

This chapter starts to analyse the capacity limitation of the equipment presented in the

model. One of the limitations is the unloading rate capacity from vessels to storage

tanks per time interval. The optimisation results in Chapter 5 show for case 1 and the

modified situations (cases 1a, 1b, 1c, 1d and 1e) that the best way is to unload the

400,000 bbl vessel in two days. But what could happen if each vessel has 550,000 bbl

to be unloaded and the maximum capacity of the pumping system to storage tanks still

is 250.000 bbl per day?. It can be foreseen that at least three days will be needed to

unload each vessel.

On the other hand, it is possible to make other questions to evaluate the current

equipment capacity and process conditions to detect possibilities to save more money

after applying the optimisation techniques i.e. how it is possible to keep at least in an

optimal value the operational cost of the analysed system fulfilling the plant charge

demand if it is notified at the last day that the crude oil sea transportation company

has decided to increase the unloading cost?

6.1 Case 2, Effect of Increasing the Vessel Unloading Volume.

Table 6.1 has a summary of the main data for this case. It can be observed that the

data is quite similar to Table 5.1, except that the crude oil amount of each vessel has

changed.

Scheduling Horizon (# unit time)

Number of Vessels Arrivals

Arrival Time

15 days

3

of Sulphur Content

Amount

Crude oil

Vessel 1

Day 1

550,000 bbl 0.01

Vessel 2

Day 6

550,000 bbl 0.03

Vessel 2

Day 10

550,000 bbl 0.05

Number of Storage tanks

3

Storage Tanks

Maximum

Initial Oil

Sulphur Content

Capacity

Amount

Tank1

350,000 bbl

30,000 bbl 0.01

_____________________________________________________________________

84

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

(min.vol.5,000bbl)

Tank 2

350,000 bbl

(min.vol.5,000 bbl)

Tank 3

350,000 bbl

(min.vol.5,000 bbl)

Number of Charging tanks

Charging Tanks

Capacity

200,000 bbl

0.03

140,000 bbl

0.05

2

Initial Oil

Amount

Range of Sulphur

Content & Initial

value

0.01-0.02

0.0158

0.035 0.045

0.0389

Tank1

250,000 bbl

70,000 bbl

(min.vol.15,000bbl)

Tank2

250,000 bbl

190,000 bbl

(min.vol.15,000bbl)

Number of CDU

1

Maximum Flow from vessel to storage tank.

250,000 bbl/day

Maximum Flow from one storage tank to charging 70,000 bbl/day

tanks simultaneously during any time period.

Maximum Flow from one storage tank to one 65,000 bbl/day

charging tank.

Maximum Flow from one charging tank to one 82,000 bbl/day

CDU.

Cost involved in vessel operation

Unloading: US$ 10,000/day

Sea Waiting: US$ 8,000/day

Tank inventory cost

Storage tank: US$ 0.008 / (day x

bbl)

Charging tank: US$0.005/(day x

bbl)

Changeover cost per feed switch

US$50,000

CDU crude oil demand

75,000 bbl/day

Table 6.1 - Case 2. Input data summary.

After applying the model for the above case there is not feasible results because the

high sulphur oil optimal demand for blending is low to compensate the higher volume

of crude oil unloading from vessel 3 to tank 3 and therefore, the current storage

capacity for the high sulphur crude oil is not enough for keeping the remain non used

crude oil. Then to solve this problem, a tank of 600,000 bbl, available in the same

system, will be available to replace tank 3 keeping the same conditions of minimum

crude oil volume, start up crude oil volume and flow rate capacities indicated in table

6.1.

_____________________________________________________________________

85

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

After applying the model again with the new conditions for case 2 (storage tank 3

capacity changed), the following results are highlighted:

The optimal operation cost for the 15-day scheduling horizon of this case is

US$450 824.1.

Table 6.2 indicates the suitable days for starting and finishing unloading in the

scheduling horizon of each vessel:

Vessel

Starts unloading

Finish unloading

Day 1

Day 5

Day 6

Day 9

Day 10

Day 12

It can be observed from table 6.2 that the optimal unloading start up time for all

vessels is the same arriving date.

Table 6.3 indicates each vessel crude oil volume variation after starting unloading

during the scheduling horizon.

Day

1

2

3

4

5

550

510

510

475

250

Vessel 1

Day

5

6

7

8

9

550

540

475

250

Vessel 2

Day

10

11

12

13

14

550

500

250

0

Vessel 3

Table 6.3 - Case 2. Vessel volume variation (bbl x 1000) results.

6

0

10

0

15

7

11

It can be observed from table 6.3 that the optimal unloading time for vessel 1 is five

days, for vessel 2 is four days and for vessel 3 is three days. Then, according to this

table the optimal results suggest not to unload in the middle of the unloading period

for vessel 1 on day 2. This consideration may be consequence of the current available

_____________________________________________________________________

86

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

understood observing the flow rate table 6.4.

Fig. 6.1 indicates the optimal variation of the storage tank volume per time interval. It

is interesting to observe that storage tanks 1 and 2 keep at the minimum volume or

close to the minimum volume around two or three days since they have been started to

be fed by the vessels. This phenomenon is because almost all their incoming volume

has been simultaneously fed to the charging tanks and therefore, it does not allow that

both tanks significantly increase their crude oil volume in this short period of time.

This behaviour helps to keep a low inventory cost. The optimal solution considers to

increase the tank remain crude oil volume close to the finish of the unloading period

i.e. tank 1 and 2 got the maximum volume the day the vessel unloading is finished.

This result is indicated in fig. 6.1, one period of time (day) after the total mass balance

has been done (equation 3.21, chapter 3).

600

500

400

300

200

100

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

Fig 6.2 indicates the optimal crude oil volume variation of each charging tank per

time interval (day). It clearly indicates that when one tank is feeding the CDU, the

other tank is being fed or refilled by the crude oil from the storage tanks fluctuating

_____________________________________________________________________

87

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

between the minimum and maximum volume capacity of each tank. Particularly, for

this case 2, along the first half of the scheduling horizon, charging tanks never reach

the minimum volume.

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

Fig 6.3 indicates the optimal feeding to the CDU per each time interval during the

scheduling horizon. The 75,000-bbl feed demand should be given at days 8, 11 and

14. On the other hand, the 82,000-bbl maximum feed should be given at days 1, 2,

3,4,5,6,9,12 and 15. The optimal total amount of crude oil proposed to feed the CDU

for this case during the 15 day scheduling horizon is 1,197,000 bbl; 6.4% above the

total feed demand. They are 11,000 bbl more to feed than the case 1 in chapter 5.

Definitely, despite there is more volume of crude oil unloading during the scheduling

horizon, it could have expected that the increase in the feed to the CDU were more,

but either the current tank volume capacities or the pumping system capacities may

have prevented to achieve this objective. Chapter 7 will help to understand what it is

the real problem in the system analysed.

_____________________________________________________________________

88

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

In addition, storage tanks finish this period with more crude oil volume. Therefore,

this is important information as input condition for the next period. The optimal

numbers of feed switches or changeovers for case 2 are five (5), given at days 2,4,6,9

and 12 respectively. Figs 6.2 and 6.3 clearly show this condition. That means that the

change of vessel oil volume unloaded from 400,000 bbls to 550,000 bbls has

increased the number of changeovers from 4 to 5 it is compared cases 1 and 2.

84

82

80

78

76

74

72

70

1

10 11 12 13 14 15

Figure 6.3 - Case 2. Feeding to the crude distillation unit results. Solid bars:

blended oil 1; blank bars: blended oil 2.

Table 6.4 indicates for case 2 the optimal volumetric flow rate at each time interval

(day) from vessels to storage tanks. Particularly, each vessel unloads crude oil with

one specific composition to only one storage tank already assigned at the planning

level (equation 3.15, chapter 3).

Day

1

2

3

4

5

6

7

40

0

35

225

250

Vessel 1

Day

5

6

7

8

9

10

11

10

65

225

250

Vessel 2

Day

10

11

12

13

14

15

50

250

250

Vessel 3

Table 6.4 - Case 2. Results for volumetric flow rate (bbl x 1,000) per day from

vessels to storage tanks.

_____________________________________________________________________

89

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Tables 6.5a and 6.5b also respectively indicate the optimal flow rate from storage to

charging tanks. They show that the optimal total amount of crude oil transferred

during the scheduling horizon, from the storage tanks to the charging tanks to balance

the CDU crude oil demand for this case is 1,120,000 bbl. This quantity is lower than

the total crude oil fed to the CDU during the scheduling horizon by 77,000 bbl.

Stor.TChar.T

1-1

1-2

2-1

2-2

3-1

3-2

Table 6.5a

65

65

9.8

3

35

42

65

65

65

9.2

23.8

65

65

65

10.2

- Case 2. Results for volumetric flow rate (bbl x 1,000) per day from

Stor.TChar.T

1-1

1-2

2-1

2-2

3-1

3-2

Table 6.5b

10

65

65

65

35

11

12

13

26.6

29.4

14

15

65

65

49

- Case 2. Results for volumetric flow rate (bbl x 1,000) per day from

Fig. 6.4 indicates the optimal fluctuation for the sulphur content in the crude oil of

each charging tank at any time interval (day) a long the scheduling horizon. Charging

tank 2 presents an oscillating sulphur concentration along the scheduling period inside

its range, with a tendency to stabilise finishing the period at the minimum range value

(0.035). Also for the conditions presented in case 2, the CDU is prepared to receive

whichever of these blended crude oils X or Y; instead the oil refinery is mainly

concerned that the system analysed is be able to keep the sulphur content range in

each charging tank while the demand of crude oil to feed the CDU is satisfied.

_____________________________________________________________________

90

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

charging tanks

0.045

0.04

0.035

0.03

0.025

0.02

0.015

0.01

0.005

0

1

10 11 12 13 14 15

Charging Tank 1 Sulphur content

Figure 6.4 - Case 2. Crude oil sulphur content variation results in charging

tanks.

6.2 Scheduling Alteration of Case 2.

Similar as it was done for case 1 in chapter 5, the following questions could come out

regarding case 2: what could happen if it is informed that the vessel arriving schedule

has changed just after the monthly plan already have been done and have got the

approval at the managerial level? It is possible to do something at the scheduling level

in order to perform again the optimization techniques as they were done for the first

arriving schedule?. The answers to these questions are also very important for this

case if the problem also is relating to keeping the minimum operational cost and

hence, a good profit margin.

Table 6.6 shows the arriving schedule for the new case 2a:

Vessel Arrival Time Amount of Crude oil Sulphur Content

Day 1

550000 bbl

0.01

1

Day 5

550000 bbl

0.03

2

Day 11

550000 bbl

0.05

3

Table 6.6 - Case 2a. First change of arriving schedule.

_____________________________________________________________________

91

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

As it is can be observed from this new arriving schedule vessel 2 will arrive one day

earlier and vessel 3 will arrive one day later with respect to the original arriving

schedule approved at the planning level for case 2.

After applying the model for this case 2a (same case 2 including the modifications of

the arriving schedule), the following result are highlighted:

The optimal operational cost for the 15 day scheduling horizon of case 2a is

US$452,892.8. It is US$ 2,068.7 more costly than the original case 2.

Table 6.7 indicates the optimal suitable days for starting and finishing unloading in

the scheduling horizon of each vessel:

Vessel

Starts unloading

Finish unloading

Day 1

Day 4

1

Day 7

Day 10

2

Day 11

Day 13

3

Table 6.7 - Case 2a. Unloading start and finish results.

It can be observed from table 6.7 that the optimal unloading time for vessel 2 is two

days after arriving and the optimal unloading time period for vessels 1 and 2 are four

500

450

400

350

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

_____________________________________________________________________

92

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

Figs. 6.5 and 6.6 indicate the optimal variation of storage and charging tank volume

per time interval respectively. Fig 6.7 indicates the optimal feeding to the CDU per

each time interval during the scheduling horizon for case 2a. The 75,000-bbl feed

demand should be given at days 5,8,11 and 14. On the other hand, the 82,000-bbl

maximum feed should be given at days 1,2,3,6,9,12 and 15. The optimal total amount

of crude oil proposed to feed the CDU for this case during the 15 day scheduling

horizon is 1,186,000; 5.42 % above the total oil demand. Then, as it can observe it is

the same of case 1, even the same feed flow rate per day for the CDU (fig.5.4, chapter

5). Moreover, with respect to the original case 2 this change of arriving schedule has

reduced the total optimal feed to CDU in 11,000 bbl along the schedule horizon.

Definitely, despite there is more volume of crude oil unloading during the scheduling

horizon, the optimal solution for the new conditions does not allow an increase of the

total feed to CDU beyond the results of case 1 in chapter 5. The numbers of

changeovers for case 2a are five (5) given at days 2,3,6,9 and 12 respectively

(different to case 2 in one day) and the figs 6.6 and 6.7 clearly show this condition.

_____________________________________________________________________

93

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

84

82

80

78

76

74

72

70

1

10 11 12 13 14 15

Figure 6.7 - Case 2a. Feeding to the crude distillation unit. Solid bars: blended oil

1; blank bars: blended oil 2.

Table 6.8 indicates for this case the optimal volumetric flow rate at each time interval

(day) from vessels to storage tanks. As it can be seen, for this case the optimal

solution does not present unloading interruption any interval of time inside the

unloading period.

Day

Vessel 1

34.15

40.85

225

250

Day

10

11

15

65

220

250

11

12

13

14

15

50

250

250

Vessel 2

Day

10

Vessel 3

Table 6.8 - Case 2a. Results for volumetric flow rate (bbl x 1000) per day from

vessels to storage tanks.

_____________________________________________________________________

94

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

charging tanks

0.05

0.045

0.04

0.035

0.03

0.025

0.02

0.015

0.01

0.005

0

1

10 11 12 13 14 15

Charging Tank 1 Sulphur content

Figure 6.8 - Case 2a. Crude oil sulphur content variation results in charging

tanks.

Fig. 6.8 indicates the optimal fluctuation for the sulphur content in the crude oil of

each charging tank at any time interval (day) a long the scheduling horizon i.e. in the

first half part of the scheduling horizon for charging tank 1 may be notable the

unloading delay of vessels 2 and 3 with higher sulphur content. However, in the last

half part of the scheduling horizon, compared with case 2, case 2a presents more

consumption of crude oil from storage tank 3 and hence, the sulphur content of both

charging tanks is in average higher than case 2.

6.3 Case 3, Effect of a High Increase in Unloading Cost.

A case 3 is presented with the same case 2 conditions but increasing the unloading

cost from US$ 10,000 to US$ 40,000 per vessel and per day since the vessel has

started up till it has finished the unloading. This case is presented to analyse the

optimal cost impact when the mentioned cost is considerably increased with respect to

the sea waiting cost.

The optimal operational cost for the 15 day scheduling horizon of this new case is

US$809,905.5.

_____________________________________________________________________

95

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

The table 6.9 indicates the optimal suitable days for starting and finishing unloading

in the scheduling horizon of each vessel:

Vessel

Starts unloading

Finish unloading

Day 1

Day 5

Day 6

Day 9

Day 10

Day 12

Comparing table 6.9 with table 6.2, case 2; it was not optimally possible to save any

unloading day for diminishing the impact of the high unloading cost. The current

capacity of some equipment may have not allowed achieving the expected response

with respect to reduce the unloading cost by using for the unloading operation the

fewer days as possible. This situation will be clarified forward in chapter 7.

Figures 6.9, 6.10 and 6.11 respectively show for this case the optimal storage tank and

charging tank volume variation and the optimal feeding to the crude distillation unit

respectively.

600

500

400

300

200

100

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

_____________________________________________________________________

96

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

84

82

80

78

76

74

72

70

1

10 11 12 13 14 15

Figure 6.11 - Case 3. Feeding to the crude distillation unit. Solid bars: blended

oil 1; blank bars: blended oil 2.

As is shown in fig 6.11, for this case the total feed to the CDU along the scheduling

horizon is 1,201,000 bbl; 6.76% above the total feed demand. This optimal result

shows that the increase of the unloading cost has increased the total operation cost but

also positively has increased the total amount of total crude oil feed to the CDU by

_____________________________________________________________________

97

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

4,000 bbl more compared with case 2 and by 15,000 bbl more compared with case 1.

But the refinery has to do a cost benefit analysis to check if this unloading cost

increase could be paid back by the increase of feed to the CDU in accordance with

case 3 optimal results. The numbers of feed switches are still five (5) for this case and

are presented at days 3,5,7,9 and 12 as is clearly shown in figures 6.10 and 6.11. The

sequence suggested for making these mode changeovers are different from case 2.

Tables 6.10 indicate the volumetric flow rate from vessels to storage tanks. Vessel 1

presents an optimal unloading interruption at the day 3. However, this situation is

quite similar to the case 2 presented in table 6.4.

Vessel 1 2 3 4

5

6 7 8

9

10 11 12 13 14 15

24 65 0 211 250

1

10 65 225 250

2

50 250 250

3

Table 6.10 - Case 3. Results for volumetric flow rate (bbl x 1000) per day from

vessels to storage tanks.

This case analyses the optimisation techniques applied if along the scheduling horizon

the crude distillation unit must stop one day to be inspected and to make some minor

repairs as an emergency event for what the planning level usually does not have the

enough tools to solve this problem in an optimal way to reduce the impact in the

operational cost of the system analysed in this Thesis. These cases sometimes happen

in oil refineries when major equipment has to shut down to do non planned activities.

Case 4 has the same conditions presented for case 2, except that the CDU on day 9

can not be fed by any crude oil from charging tanks.

The optimal operational cost for the 15 day scheduling horizon of this new case is

US$ 452,267.5. It is only US$ 1,443.4 more costly than case 2.

_____________________________________________________________________

98

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

The table 6.11 indicates the suitable days for starting and finishing unloading in the

scheduling horizon of each vessel:

Vessel

Starts unloading

Finish unloading

Day 1

Day 5

1

Day 6

Day 9

2

Day 10

Day 12

3

Table 6.11 - Case 4. Unloading start and finish results.

There is no variation in the unloading period for all vessels compared with case 2,

table 6.2.

Figures 6.12, 6.13 and 6.14 respectively show for this case the storage tank volume

variation, the charging tank volume variation and the feeding to the crude distillation

unit.

600

500

400

300

200

100

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

_____________________________________________________________________

99

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

100

80

60

40

20

0

1

9 10 11 12 13 14 15

Figure 6.14 - Case 4. Feeding to the crude distillation unit. Solid bars: blended oil

1; blank bars: blended oil 2.

It is possible to see in figure 6.13 the effect of not having feed for CDU on day 9;

charging tank 1 takes advantage of this situation to start to be refilled and charging

tank 2 already is full on day 9 and start day 10 with the same volume.

_____________________________________________________________________

100

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

As is shown in fig 6.14, for this case the total feeding to the CDU along the

scheduling horizon is 1,137,000 bbl; still 1.07 % over the total feed demand. This

optimisation result shows that the shut down of the CDU on the day 9 has allowed not

only to practically impact the optimal operation cost of the system analysed in a

minimum increase percentage (0.32%); also, it has helped to diminishes the effect of

the total lost of the day 9 feed (82,000 bbl, fig. 6.3 for case 2); only the total feed to

the CDU was depleted by 60,000 bbl considering the optimal results of this case 4.

The numbers of feed switches are still five (5) for this case and are presented at days

3,5,7, 10 and 12 as is clearly shown in figures 6.13 and 6.14. The order suggested for

making these mode changeovers are different from case 2.

However, special care should be applied solving this kind of situation using option b

because considering the optimal results for whichever case analysed, the total feeding

to CDU could be under the total feed demand expected for the scheduling horizon;

then, it will be convenient to try solving this problem using optiond previously

fixing the total feed demand required along the scheduling horizon and applying the

same constraint 3.44 for the shutdown day within this option.

6.5 Chapter Summary.

Table 6.12 shows a summary of the resources used to solve the cases reviewed in this

chapter.

Case 2

Case 2a

Case 3

Equations and constraints 1,682

1,682

1,682

Single variables

863

863

863

Discrete variables

261

261

261

Iterations

886,901

502,694

4,631,527

Solving time

117.9 min. 112.7 min. 571.8 min.

Table 6.12 Resource summary used to solves cases.

Case 4

1,683

863

261

1,151,095

208.7 min.

Some practical cases have been reviewed in this chapter. The normal operation of a

system could convert it in abnormal operation if some conditions of the system have

_____________________________________________________________________

101

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

drastically changed and no action is taken to face these changes. Optimal operational

costs and the production scheduling are directly affected. The descriptions of the

changes in conditions are guided to find the system limitations, specifically in its

equipment capacities explaining how these limitations prevent having better optimal

operational costs and operational schedules. Several comparisons have been done

even with cases from chapter 5 to help to clarify. For the analysis of the cases, the

new optimisation model is applied to the practical cases as in real situations could

happen when some scheduling or operation conditions have changed and hence, the

refinery has to prepare itself to promptly face these situations that affects the analysed

system i.e. increase of the vessel crude oil volume to unload, non planning plant shut

down in the middle of the scheduling horizon and so on. In addition, also as well as it

was mentioned in chapter 5, the different situations have shown that if the conditions

of each problem are feasible always an optimal solution could come out. However, an

infeasible situation was solved in this chapter for getting the optimal solution for one

case analysed.

In addition, the equipment limitations have come out looking for responses for solve

them and even for trying to get better optimal operational cost results. On the other

hand, there would be other concern about the impact in the optimal operational cost

and in the operational schedule if some equipments are modified; this will be the

matter of next chapter 7 that will solve questions like these:

What could happen if at the managerial level is decided to increase the volume of both

charging tanks up to 300,000 bbl. If one is ready first, is it worth it to start putting in

operation the new one operating with one of the other 250,000 old one? What could

be the optimal operational results if the refinery is managing cases with conditions

similar to the case 1 conditions?

_____________________________________________________________________

102

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

THINKING EITHER IN REVAMP OR CHANGE EQUIPMENT IS

POSSIBLE.

The refinery managers normally want to have some efficient tools that help them to

take decision to invest either in revamping or changing equipment in a process

system. The main question that could come out is: Is it possible to do it using the

new model to optimise the operational cost of the system analysed in this thesis?

Well, it is possible to start solving the questions presented in the chapter 6 summary.

7.1 Cases 5 and 6, Increasing Capacity of Charging Tanks.

The managerial level is consulting and preparing a project that involves the capacity

increase of both charging tanks up to 300,000 bbl. The project should be done in two

phases. The scope of the first phase is to complete the first charging tank before

starting with the other one. If one is ready first; is it worth to put in operation the new

one with the other 250,000 old one? What could be the optimal results if the refinery

is managing cases of vessel arriving schedules and conditions the same presented for

case 1 in chapter 5 ?.

In order to clarify all both cases 5 and 6, will handle the same conditions of case 1 in

chapter 5, except that for case 5, only one new charging tank will be operating with a

volume capacity of 300,000 bbl replacing one of the existing 250,000 bbl charging

tank, specifically the lower sulphur one; and case 6 will have both charging tanks new

with a volume of 300,000 bbl each. The tank unit inventory cost has not any increase

by the effect of the investment.

After applying the model for the above case 5 and case 6 the following results are

highlighted:

The optimal operation cost for the 15 day scheduling horizon of case 5 is

US$331,071.2 and for case 6 is US$ 329,186.0.

_____________________________________________________________________

103

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

It can be observed that case 5 is US$ 7,893.1 less costly than case 1 and also case 6 is

US$ 9,778.25 less costly than case 1; it means that a saving of around US$ 235,000

per year could be achieved. Although, case 6 could reduce more the optimal

operational cost it could be worth it starting when the first project phase has finished

with the first new charging tank while the other is finished in the next project phase.

The table 7.1 indicates the suitable days for starting and finishing unloading in the

scheduling horizon of each vessel. Both cases have the same optimal unloading start

and finish:

Vessel

Starts unloading

Finish unloading

Day 1

Day 2

Day 6

Day 7

Day 10

Day 11

Figs. 7.1 and 7.2 indicate the optimal volume variation for each storage tank at each

interval (day) for cases 5 and 6.

400

Crude bbl x 1000

350

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

_____________________________________________________________________

104

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

350

Crude bbl x 1000

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

It can be observed analyzing figs.7.1 and 7.2 that for both cases, storage tanks 1 and 2

have similar volume variation behaviour and tanks 2 and 3 present higher

consumption in case 6 than in case 5 because both charging tank capacities have

increased, including the high sulphur one. For both cases the optimal results indicate

a little higher consumption of storage tank oil volumes than case 1, chapter 5.

350

Crude bbl x 1000

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

_____________________________________________________________________

105

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

350

Crude bbl x 1000

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

Figs 7.3 and 7.4 indicate the optimal oil volume variation at each time interval (day)

for each charging tank. It can be observed at the top of each graphic that in some

cases the charging tank that is being refilled gets its maximum volume before the

other tank has been consumed to the minimum volume to make the feed switch. This

is a clear demonstration that charging tanks although their capacities have been

increased, they still have some volume restrictions. Later, after the other results for

both cases have been observed it would be possible to think in advance if the project

scope to increase both charging tank volumes would be profitable or not.

Figs 7.5 and 7.6 indicate for cases 5 and 6 the optimal feeding to the CDU per each

time interval during the scheduling horizon.

For case 5, the 75,000 bbl feed demand should be given at days 8 and 14; in another

hand, the 82,000 bbl maximum feed should be given at days 1, 2, 3, 4,5,6,9,10,11,12

and 15. Then, the optimal total amount of crude oil proposed to feed the CDU during

the 15 day scheduling horizon is 1,208,000 bbl; 7.38% above the total feed demand.

Hence, there are 22,000 bbl more optimally available to feed the CDU during the

schedule horizon than case 1, which means around 535,000 bbl more per year.

_____________________________________________________________________

106

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

84

82

80

78

76

74

72

70

1

10 11 12 13 14 15

Figure 7.5 - Case 5. Feeding to the crude distillation unit results. Solid bars:

blended oil 1; blank bars: blended oil 2.

100

80

60

40

20

0

1

9 10 11 12 13 14 15

Figure 7.6 - Case 6. Feeding to the crude distillation unit results. Solid bars:

blended oil 1; blank bars: blended oil 2.

Instead for case 6, the optimal result considers that the CDU could be feed each time

interval (day) by the daily maximum plant capacity. This will be a great result when

the project has been completed. That means, the optimal total amount of crude oil

_____________________________________________________________________

107

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

proposed to feed the CDU for case 6 during the 15 day scheduling horizon is

1,230,000 bbl; 9.33 % above the total feed demand, exactly the top allowed by the

plant capacity. Then, case 6 optimally presents 44,000 additional barrels to feed the

CDU during the schedule horizon with respect to case 1, which means around

1,070,000 additional barrels per year. Feed switches or changeovers for both cases 5

and 6 are four (4) given at days 3, 6, 9 and 12 respectively, the same feed switch

behaviour presented in case 1.

Table 7.2 indicates for both cases 5 and 6 the same optimal volumetric flow rate at

each time interval (day) from vessels to storage tanks. Particularly, as each vessel

unloads crude oil with one specific composition, for this case each vessel unloads to

one specific storage tank already assigned at the planning level (application of

equation 3.15, chapter 3).

Vessel 1

2

3 4 5 6

7

8 9 10 11 12 13 14 15

150 250

1

150 250

2

150 250

3

Table 7.2 - Case 5 and 6. Volumetric flow rate in bbl x 1,000 per day from vessels

to storage tanks.

Tables 7.3a&b and 7.4a&b also respectively indicates the optimal flow rate from the

storage tanks to the charging tanks for cases 5 and 6 .

Stor.T-Char.T

1

1-1

65

1-2

2-1

65

2-2

3-1

6.5

3-2

Table 7.3a - Case 5. Results for

2

65

28.5

36.4

3.35

65

36.5

6

65

7

61.8

65

54.2

65

17.75

volumetric flow rate (bbl x 1,000) per day from

_____________________________________________________________________

108

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Stor.T-Char.T

9

10

11

12

13

14

15

1-1

65

49

1-2

14.45

2-1

65

49

2-2

65

44.8

3-1

18

3-2

45.75

65

Table 7.3b - Case 5. Results for volumetric flow rate (bbl x 1,000) per day from

storage tanks to charging tanks

Stor.T-Char.T

1

2

3

65

65

1-1

36.4

1-2

65

32.3

2-1

65

2-2

2.7

3-1

65

3-2

Table 7.4a - Case 6. Results for volumetric flow

4

10.5

32.7

6

65

7

51

65

65

64.4

rate (bbl x 1000) per day from

Stor.T-Char.T

9

10

11

12

13

14 15

65

44.5

1-1

10.34

12.26

1-2

65

65

2-1

65

65

2-2

6.5

3-1

2.9

65

25.5

3-2

Table 7.4b - Case 6. Results for volumetric flow rate (bbl x 1,000) per day from

storage to charging tanks.

Moreover, tables 7.3a&b and 7.4a&b show that the optimal total amount of crude oil

transferred during the scheduling horizon, from the storage to charging tanks to

balance the CDU crude oil demand for cases 5 and 6 are 1,181,000 bbl and 1,242,000

bbl respectively. These quantities are lower than the total crude oil fed to the CDU

during the scheduling horizon by 27,000 bbl and 12,000 bbl respectively. As it can be

observed, the charging tank volume expansions have optimally increased the total

flow transferred from storage tanks to charging tanks compared with case 1 that was

77,000 bbl under the optimal total flow feeding the CDU.

_____________________________________________________________________

109

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

charging tanks

0.05

0.045

0.04

0.035

0.03

0.025

0.02

0.015

0.01

0.005

0

1

10 11 12 13 14 15

Charging Tank 1 Sulphur content

charging tanks

Figure 7.7 - Case 5. Crude oil sulphur content variation in charging tanks.

0.05

0.045

0.04

0.035

0.03

0.025

0.02

0.015

0.01

0.005

0

1

10 11 12 13 14 15

Charging Tank 1 Sulphur content

Figure 7.8 - Case 6. Crude oil sulphur content variation in charging tanks.

Figs. 7.7 and 7.8 indicate the optimal fluctuation for the sulphur content in the crude

oil of each charging tank at any time interval (day) a long the scheduling horizon. For

case 6 (fig. 7.8) is notable the influence of having during the scheduling horizon a

major consumption of storage tanks 2 and 3 (higher sulphur concentration); this effect

_____________________________________________________________________

110

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

may be shown by the bigger sulphur concentration average during the scheduling

horizon in charging tank 2 compared with case 5 (fig.7.7).

Definitely, after all the results have been observed, it is easier to tell in advance that it

would be convenient for the system to increase the charging tank volume as was

indicated in the project scope. An analysis cost-benefit should be done to check is the

project may be paid back in some years. Perhaps the cost of the project could be

hardly pay back in less than three years only having in account the

presented

reduction in optimal operational cost (2.88%) and the possibility to increase optimally

the amount of crude oil charges to the CDU. However, trials and errors have been

done with the model considering a lower increase in the capacity of charging tanks

which matches the maximum feed every time interval (day) and at least a similar

optimal cost reduction as was presented for case 6.

After trails and errors were made to get the minimum optimal charging tank capacities

for searching similar or better results than case 6, it is interesting to show the final

results in table 7.5. This exercise was made keeping all the same conditions for case 6,

while both charging tank capacities was randomly changed keeping the same size for

each one.

As it can be deducted from table 7.6, for the same conditions of case 1 and expanding

charging tanks capacity to 263,000 bbl each, is possible to get optimally the maximum

total feed to the CDU. That means that the refinery could incur investing less money

for achieving the same results of case 6. Even it could be possible to think in revamp

the charging tanks from 250,000 to 263,000 bbl (5.2% volume increase) and hence,

much less money could be needed. But, the effect of having a minor total optimal

operational cost in case 6 compared to the 263,000 bbl volume expansion

consideration by US$ 7,290.1 less costly per period (around US$ 175,000 per year)

should be taken in account for the final selection according to the choices or

alternatives for the project. Perhaps, the 268,000 bbl capacity expansion (10.7%

volume increase) is a better choice for the conditions of the problem.

_____________________________________________________________________

111

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Each

charging

tank

300,000

256,000

260,000

263,000

268,000

bbl

bbl

bbl

bbl

US$

US$

US$

US$

US$

337,643.0

336,818.5

336,476.1

331,152.1

329,186.0.

1,210,000

1,226,000

1,230,000

1,230,000

1,230,000

bbl

bbl

bbl

bbl

bbl

374,925

512,643

547,655

322,055

350,904

7,780.7sec

7,530.5sec

(143 min)

(129.7min) (125.5min)

bbl

(case 6)

capacity.

Total

optimal

operational

cost

CDU total

feed

Iterations

Software

time spent

running

(170min)

(173min)

7.2 Case 7, Effect of Changing Pumping Rate Capacity.

It is also possible to try to invest the money in another way maybe less convenient.

What happens if somebody thinks that the best way is to increase the capacity of each

pumping system from storage to charging tanks (PUMPCAPi)? Again, the conditions

of case 1 are used considering only the following change shown in table 7.6 with

respect to table 5.1, chapter 5, in order to work with the new case 7:

Maximum Flow from one storage tank to 100,000 bbl/day

charging tanks simultaneously during any time

period.

Maximum Flow from one storage tank to one 95,000 bbl/day

charging tank.

Table 7.6 - Case 7. Input data summary exception.

There are three storage tanks and for solving the problem, three pumping systems will

be considered, one per storage tank. The project is proposed to be made in three

_____________________________________________________________________

112

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

phases, each new pumping system will be started and finished one by one and it is

suggested to study how the optimal costs and the operational schedule of the system

are improving, first working with one new pumping system and two old ones (case

7a), second working with two new pumping systems and the last old one (case 7b)

and third working with all the new pumping systems (case 7c).The sequence for

installing the new pumping systems is from the lower sulphur (storage tank 1) to the

higher sulphur side (storage tank 3).

After applying the model for each case and getting the optimal operation cost for the

15 day scheduling horizon the following result are highlighted:

Case 7a optimal operational cost: US$ 333,692.5

Case 7b optimal operational cost: US$ 333,351.1

Case 7c optimal operational cost: US$ 333,286.3

As it could be observed in advance compared with the case 1 optimal cost (US$

338,964.25), the benefit reducing the optimal operational cost may have been less

than expected with the new storage tanks pumping systems. The maximum reduction

is US$ 5,677.95 (around US$ 138,000 per year) considering the same other

equipments and the scheduling conditions of case 1.

Fig.7.9 indicates the optimal volume variation for each storage tank at each time

interval (day) for case 7c. It is interesting to comment that the availability of the new

pumping systems may have allowed that storage tank 3 has been more consumed and

as a result, vessel 3 have optimally been able to start unloading from the arriving day

compared with fig. 5.2, case 1. Cases 7a and 7b do not show anything additional

relevant to comment regarding the storage tank volume variation.

_____________________________________________________________________

113

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

400

Crude bbl x 1000

350

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

Fig 7.10 indicates the optimal volume variation at each time interval (day) for each

charging tank for case 7c. Compared with fig. 5.3, chapter 5, it can be observed that

charging tanks are refilled faster and even in one occasion the charging tank 2 has to

_____________________________________________________________________

114

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

wait in the full condition after it already has been refilled for two continuous time

intervals(days) before it starts to be discharged at day 6. Therefore, it could be

anticipated that charging tanks continue with the restriction in their volume size for

this case 7 and the effort of this project to decrease the optimal operational cost and

also, increase the total feed to CDU is depleted by this fact. Cases 7a and 7b do not

show anything additional relevant to comment regarding the charging tank volume

variation topic.

The optimal feeding to the CDU per each time interval during the scheduling horizon

for case 7c is the same presented for case 1 in fig.5.4, chapter 5; except that it presents

a depletion of the feed on the day 15 from 82,000 to 75,000 bbl; then, the total feed to

the CDU is 1,179,000 bbl; 4.8 % above the total feed demand. For all cases, the feed

switches or changeovers are four (4) given at the same days of case 1.

Regarding the flow rates from storage to charging tanks, it is interesting to comment

that the optimal results for case 7c has indicated due to the higher flow rates, the

operation of the flow rate schedule using less time intervals (days). Table 7.7 a&b

shows this characteristic.

Stor.T-Char.T

1

2

3

4

5

6

7

8

1-1

95

95

30

1-2

43

2-1

85

95

15

2-2

95

3-1

3-2

86

Table 7.7a - Case 7c. Results for volumetric flow rate in bbl x 1,000 per day from

storage to charging tanks.

Stor.T-Char.T

9

10

11

12

13

14

15

1-1

95

30

1-2

10.4

2-1

95

2-2

95

15.45

3-1

5

10

3-2

49

65.15

Table 7.7b - Case 7c. Results for volumetric flow rate in bbl x 1,000 per day from

storage to charging tanks.

_____________________________________________________________________

115

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Moreover, this table 7.7 a&b shows that the optimal total amount of crude oil

transferred during the scheduling horizon, from the storage tanks to the charging tanks

to balance the CDU crude oil demand for this case 7c is 1,109,000 bbl, which it is the

same presented for case 1. This quantity is lower than the total crude oil fed to the

CDU during the scheduling horizon by 70,000 bbl. Although, this total crude oil

transferred value is the same in both cases; the total flow transferred from each

storage tank is different i.e. storage tank 3 for case 7c optimally transferred 215,150

bbl compared with case 1, that was 190,000 bbl (table 5.5, chapter 5).

7.3 Case 8, Increasing Capacity of Storage Tanks.

Regarding case 3, chapter 6 with the higher unloading cost, it is not satisfactory to

observe that although the optimal cost solution have increased the total feed to the

CDU by 4,000 bbl along the scheduling horizon (around 97,300 bbl per year)

compared with the original case 2 in the same chapter 5, it was not possible to save

any unloading day along the scheduling horizon for the conditions presented for this

case 3, even continuing using a temporary tank of 600,000 bbl as storage tank 3 used

to solve the system infeasibilities for the problem conditions presented since case 2

was analysed. The manager wants somebody to review this case in order to reduce the

optimal operational cost. What happen if a project is proposed to increase each

storage tank capacity from 350,000 bbl to 500,000 bbl for trying to reduce the

operational cost for the new situation of higher unloading cost (US$ 40,000 per vessel

per day)? For solve this problem a new case 8 is presented with the same conditions

of case 3, chapter 5, except that the storage tank capacity is 500,000 bbl each (the

temporary 600,000 bbl tank 3 is not longer used for solving this case 8).

After applying the optimisation techniques for the above case 8 the following result

are highlighted:

The optimal operation cost for the 15-day scheduling horizon of this case is

US$655,803.3. It is much less costly than case 3, chapter 5 by US$ 154,102.2 per

scheduling horizon or around US$ 3,750,000 per year.

_____________________________________________________________________

116

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Table 7.8 indicates the suitable days for starting and finishing unloading in the

scheduling horizon of each vessel:

Vessel

Starts unloading

Finish unloading

1

Day 1

Day 3

2

Day 6

Day 8

3

Day 10

Day 12

Table 7.8 - Case 8. Unloading start and finish results.

As it can be observed in table 7.8, the optimal solution suggests to unload each vessel

in three days that is the minimum possible according to the unloading pumping

capacity from vessels to storage tanks (see table 6.1, chapter 6). Also, it is optimally

suggested starting unloading the same arriving day for all vessels and hence, no sea

waiting cost should be charged this time.

Fig. 7.11 indicates the optimal variation of the storage tank volume per time interval.

Compared with fig. 6.9 for case 3, chapter 6; it is clear that the bigger capacity of

storage tanks(case 8) has optimally allowed to unload vessels in three days for the

same conditions.

600

500

400

300

200

100

0

1

9 10 11 12 13 14 15

Storage Tank 1

Storage Tamk 2

Storage Tank 3

_____________________________________________________________________

117

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Figs.7.12 and 7.13 indicate the optimal variation of the charging tank volume per time

interval and the optimal feeding to the CDU along the scheduling horizon. Then, it is

possible to observe compared with figs.5.10 and 5.11 for case 3, chapter 5; that the

conditions of case 8 reduces the number of feed switches needed from 5 to 4, just the

same number of case 1 .

300

250

200

150

100

50

0

1

9 10 11 12 13 14 15

Charging Tank 1

Charging Tank 2

84

82

80

78

76

74

72

70

1

10 11 12 13 14 15

Figure 7.13 - Case 8. Feeding to the crude distillation unit results. Solid bars:

blended oil 1; blank bars: blended oil 2.

_____________________________________________________________________

118

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Case 8 presents an optimal total feed to the CDU of 1,179,000 bbl along the

scheduling horizon which compared with case 3, chapter 5, means depletion in the

optimal total feed of 22,000 bbl (535,300 bb per year) and 7,000 bbl compared with

case 1, chapter 5. It may be possible that the reduction of feed switches has helped to

deplete the total feed to the CDU.

The flow rate from vessels to storage tanks is presented in table 7.9.

Vessel 1

1

2

3

4 5 6

9 10 11

12

13 14 15

50 250 250

50 250 250

50 250 250

Table 7.9 - Case 8. Results for volumetric flow rate (bbl x 1,000) per day from

vessels to storage tanks.

Stor.T-Char.T

1

2

3

4

5

6

7

8

1-1

65

50

65

65

1-2

33

2-1

65

65

40

2-2

65

50

3-1

3-2

65

11

Table 7.10a - Case 8. Results for volumetric flow rate (bbl x 1,000) per day from

storage to charging tanks.

Stor.T-Char.T

9

10

11

12

13

14

15

1-1

65

65

1-2

22.4

27.85

2-1

65

35

2-2

65

10.75

3-1

5

3-2

59

50

Table 7.10b - Case 8. Results for volumetric flow rate (bbl x 1,000) per day from

storage to charging tanks.

Tables 7.9 and 7.10a&b show the optimal flow rates from vessels to storage tanks and

from storage to charging tanks. Despite the storage tanks had available more crude oil

volume only was possible to optimally transfer 1,109,000 bbl from storage to

_____________________________________________________________________

119

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

charging tanks to balance the total CDU consumption which was 70,000 bbl less than

the total feed to the CDU. But again, the refinery has to do cost benefit analysis to

face the increase of unloading cost by reviewing the impact of the storage tank

capacity change and its effect in the total optimal operational cost and in the total

optimal feed to the CDU.

Trials and errors were done considering other storage tanks volume expansions for

comparing and observing other advantages with respect case 8. The results can be

observed in table 7.11. It was found that the minimum volume for each storage tank is

450,000 bbl for being able to get an optimal solution. Alternatives analysed

considering volumes under these values were infeasible. It seems that there may not

be any possibilities to increase the total feed to the CDU by only simultaneously

changing the storage tank capacity the same for each tank. The 450,000 bbl solution is

better and represents much less money investment than the 500,000 bbl solution.

Then, by only using a volume increase of 12.86% is possible to optimally save US$

155,930 per scheduling horizon (around 3,795,000 per year) compared with case 3,

chapter 6.

500,000

Each storage tank capacity.

450,000 bbl

bbl

600,000 bbl

(case 8)

Total optimal operational cost US$ 653,973.6 US$655,803.3 US$ 653,803.3

CDU total feed

1,179,000 bbl

1,179,000 bbl

1,179,000 bbl

Iterations

1,118,757

1,561,413

1,562,713

378.2 min

293.4 min

391.7min

Table 7.12 summarises the resources used to solve the cases reviewed in this chapter.

_____________________________________________________________________

120

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Case 5

Case 6

Case 7

Equations and constraints 1,682

1,682

1,682

Single variables

863

863

863

Discrete variables

261

261

261

Iterations

319,829

350,904

490,517

Solving time

130.1 min. 125.5 min. 152.3 min.

Table 7.12 Resource summary used to solves cases.

Case 8

1,682

863

261

1,561,413

293.4 min.

Some practical cases have been reviewed in this chapter focus on demonstrating the

model strength to support cases of equipment capacity increase for the system

analysed. Sometimes the conditions that represent abnormal operation could

temporally be calmed down in a convenient way to continue with the operation even

if the change of conditions have been managed using a temporary equipment of more

capacity. However, this is not always the case, in most cases the manager has to work

in a future plan to improve the capacities of some system equipment focused on

overall cost reduction (life cycle cost), including the optimal operational costs.

Furthermore, one of the main objectives of the project scope if it is economically

viable should be that the new system capacities could properly manage current or

future foreseeable changes of system conditions that hugely could impact cost and

production scheduling.

In addition, regarding the work done in this chapter, the new model has been applied

to search and discover the best optimal operational cost solution for each case as in

real situations could happen to become more profitable the production scheduling i.e.

analysis of charging tank capacity increase, pumping system changes, etc. The

analyses of the cases are compared with results from chapters 5 and 6, to help to

clarify. Moreover, as well as it was mentioned in chapters 5 and 6, the different

situations have shown that if the conditions of each problem are feasible always an

optimal solution could come out. However, some infeasible situations were helpful in

this chapter to discover the boundaries of optimal solutions for the different

conditions applied.

_____________________________________________________________________

121

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

8.1 Conclusions

In this work, a generic optimisation MILP new model have been developed and

proposed for optimising the production scheduling of crude oil inventories, blending

and feed to process plant in oil refineries that usually unload several kind of crude oil

with different kind of compositions. An important revision of all the works related to

the production planning and scheduling in oil refineries have been done. It was

highlighted that there is still too much work to do in the production scheduling field.

Working in this field involves much more mathematical complexity for formulating

the models and more computational efforts for solving them; furthermore, much less

profit is achieved compared with the planning level; however, if the production

scheduling does not properly work, the production planning targets could be seriously

damaged.

A complete oil refinery has not been modelled at the scheduling level so far. All the

developments for managing production scheduling have been done considering one

plant or a system of an oil refinery; even a subsystem of a complete plant has been

preferred for modelling it to reduce the number of equations, constraints and variables

in the model formulation. The intention to link the planning with the scheduling level

tools through a software interface still is alive but it has not been done. In fact, for

assisting somehow to do this link with no interface, many exercises have been done

using iteration results at the scheduling level for improving the input data at the

planning level. However, the efforts also have been guided toward continuing

working hard at the scheduling level for developing better integrate solutions that

represents an overall oil refinery or a big system inside an oil refinery.

Several practical cases have been presented using the model with the purpose to offer

a good understanding of the potential benefits of the model used at the scheduling

level for optimising the operational cost applied to the crude oil inventory operations.

Moreover, it has been possible to observed reviewing the cases the advantage of the

_____________________________________________________________________

122

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

model not only to review sudden changes of vessel arriving schedules, also for

detecting infeasibilities and for supporting decisions for future system expansions

projects through iterations that have helped to identify possible better alternatives.

On the other hand, the model was successfully tested using the same conditions used

by {Lee, Pinto, et al. 1996} to solve similar cases related with the same scheduling

problem for the operation of crude oil inventory and feeding system to an oil refinery.

The differences of this new model with the model developed by {Lee, Pinto, et al.

1996} were well explained highlighting the advantages of the new model and the

shortcoming of the existing model.

Finally, it is interesting to say that despite this work has involved long hours

programming and running the model for the scheduling problem presented in this

Thesis; it has been worth it. Better production scheduling and lower operational cost

have been achieved comparing the optimal results of this new model with the optimal

results of an existing model working with the same conditions.

8.2 Future Works.

Optimisation continues being a very active area of research in academia and industry.

Then, the scope of these optimisation techniques is increasing as a result of a

combination of theoretical and algorithmic advances, which are complemented with

the rapid progress in computer hardware and software. During the last years the

solvable optimisation problems have increased in size and it is expected that this trend

continues in parallel with the computer architecture developments.

The overall refinery scheduling consists of the scheduling of three main sections: the

scheduling for the operation of crude oil inventory, the process unit operation and the

operations of finished product inventory. The scope of this work has been focused on

the first one that is very critical for keeping a good and stable operation in an oil

refinery and therefore, meeting its products demand. Furthermore, some important

_____________________________________________________________________

123

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

operation conditions downstream this section, for process plants and finished product

inventories, depend on the well managed scheduling condition of this first section.

Nowadays, the work has been continued in all of the main described sections; but

mostly modellers have worked each section separately. Each work developed has

been subjected to be improved for instance reducing the computer time running the

model; in most cases this has been the subsequent step after a new model has been

developed. Much more time and efforts should be dedicated in the future to the

process unit production scheduling that could have much more non linear

formulations that must be combined with discrete variables to get the optimal

operational modes in the scheduling horizon of each process plant; however, a good

knowledge of each process plant is necessary for improving and optimising its

production scheduling.

Finally, general efforts should be continued trying to develop a well developed

package for managing the production scheduling not only using MILP, also MINLP

techniques and covering all the overall oil refinery production scheduling as well.

Later, there will be time to think, design and developed a real software integration of

production planning and scheduling that at the end of the day should be sharing one

common data base.

_____________________________________________________________________

124

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

BIBLIOGRAFY

Barton, Paul I.; Allgor, Russell J.; Feehery, William F., and Galan, Santos. Dymamic

optimization in a discontinuous world. Industrial and Engineering Chemistry

Research. 1998; 37, 966-981.

Breiner Avi and Maman, Rafi. Refinery reaps the benefits of new. Oil & Gas Journal.

2001 Jan; 46-48.

Brooke, Anthony; Kendrick, David; Meeraus, Alexander, and Raman, Ramesh.

GAMS (General Algebraic Modeling System), a user's guide. Washington: GAMS

Development Corporation; 1998.

Douglas, James M. Conceptual design of chemical engineering processes. Chemical

Engineering series. USA: Mc Graw Hill; 1988.

Edgar, Thomas F. and Himmelblau, David M. Optimization of Chemical Processes.

Second ed. Boston: Mc Graw Hill; 2001. ISBN: 0-07-039359-1.

Geankoplis, Christie. Transport processes and unit operations . USA: 1978.

Goethe-Lundgren, Maud; Lundgren, Jan T., and Person, Jan A. An optimization

model for refinery production scheduling. Int.J. Production Economics. 2002; 78:255270.

Grossman, Ignacio E. and Biegler, Lorenz T. optimizing chemical processes.

Chemical Technology. 1995 Dec; 27-35.

Hartmann, Hans. Distinguish between scheduling and planning models. Hydrocarbon

Processing. 1998 Jul; 93-100.

Jia, Zhenia; Ierapetritou, Marianthi, and Kelly, Jeffrey D. Refinery short-term

scheduling using continuous time formulation:crude-oil operations . Industrial and

Engineering Chemistry Research. 2003: 3085-3097.

Jia, Zhenya and Ierapetritou, Marianthi. Mixed-Integer Linear Programming Model

for Gasoline Blending and Distribution Scheduling . Industrial and Engineering

Chemistry Research . 2003; 42:825-835.

Joly, M.; Moro, L. F. L, and Pinto, J. M. Planning and scheduling for petroleum

refineries using mathematical programming. Brazilian Journal of Chemical

Engineering. 2002; 19(02):207-228.

Lee, Heeman; Pinto, Jose M; Grossmann, Ignacio E., and Park, Sunwon. MixedInteger Linear Programming Model for Refinery Short-Term Scheduling of Crude Oil

Unloading with Inventory Management. Industrial and Engineering Chemistry

Research. 1996; 35:1630-1641.

Magalhaes, Marcos Vinicius de Oliviera; Moro, L. F. L.; Smania, P.; Hassimotto, M.

_____________________________________________________________________

125

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

K.; Pinto, Jose M., and Abadia, G. J. SIPP- A solution for refinery scheduling . NPRA

- National Petrochemical & Refiners Association ; San Antonio, TX. Boston; 1998.

Mas, Rodrigo and Pinto, Jose M. A mixed-integer optimization strategy for oil supply

in distribution complexes. Optimization and Engineering. 2003; 4:23-64.

McCabe, Warren L. sixth ed. New York: 2001.

Moro, L. F. L.; Zanin, A. C., and Pinto, J. M. A planning model for refinery diesel

production. Computers Chemical Engineering. 1998; 22,suppl:S1039-S1042.

Park, Hyungjin; Bok, Jin-Kwang, and Park, Sunwon. Scheduling of refinery processes

with optimal control approach. Journal of Chemical Engineering of Japan. 2001;

34(3):411-422.

Pelham, Roger and Pharris, Chuck. Refinery operations and control: A future vision.

Hydrocarbon Processing. 1996 Jul: 89-94.

Peters, Max S. and Timmerhaus, Claus D. Plant design and economics for chemical

engineers . fourth edition ed. New York: Mc Graw-Hill ; 1991.

Pinto, J. M. ; Joly, M., and Moro, L. F. L. Planning and scheduling models for

refinery operations. Computers and Chemical Engineering. 2000; 24:2259-2276.

Pinto, J. M. and Mas, R. A mixed-integer optimization strategy for oil supply in

distribution complexes . Optimisation and Engineering. 2003; 4:23-64.

Pinto, J. M. and Moro, L. F. L. a planning model for petroleum refineries . Brazilian

Journal of Chemical Engineering. 2000 Dec; 17(04-07):575-585.

Quesada, Ignacio and Grossmann, Ignacio E. a globlal optimization algorithm for

linear fractional and bilinear programs . Journal of Global Optimization. 1995; 6:3976.

Reklaitis, G. V.; Ravindran, A., and Ragsdell, K. M. Engineering optimization

methods and application. USA: John Wiley and sons; 1983.

Shah, N. mathematical programming techniques for crude oil scheduling. Computers

Chemical Engineering. 1996; 20:S1227-S1232.

Song, Jehoon ; Park, Hyungjin; Lee, Dong-Yup, and Park, Sunwon. Scheduling of

actual size Refinery Processes considering environmental impacts with multiobjective

optimization. Industrial and Engineering Chemistry Research . 2002; 41:4794-4806.

Zhang, N and Zhu, X. X. A novel modelling and decomposition strategy for overall

refinery optimisation. Computers and Chemical Engineering. 2000; 24:1543-1548.

Zhang, Nan. Novell modelling and decomposition for overall refinery optimisation

and debottlenecking. Doctor of Philosofy ed. Manchester: University of Manchester

Institute of Science and Technology; 2000 Sep.

_____________________________________________________________________

126

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

APENDIX A

A.1 Example 1 Computer hardcopy output

MODEL STATISTICS

BLOCKS OF EQUATIONS

BLOCKS OF VARIABLES

NON ZERO ELEMENTS

GENERATION TIME

40

20

1445

=

SINGLE EQUATIONS

SINGLE VARIABLES

DISCRETE VARIABLES

552

337

116

0.120 SECONDS

EXECUTION TIME

=

0.160 SECONDS

VERID WAT-25-089

AMS 2.25.089 DOS Extended/C

08/15/03 23:39:16

PAGE

44

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 443

S O L V E

MODEL

TYPE

SOLVER

SCHEDULING

MIP

OSL

**** MODEL STATUS

**** OBJECTIVE VALUE

OBJECTIVE

DIRECTION

FROM LINE

CO

MINIMIZE

443

1 NORMAL COMPLETION

8 INTEGER SOLUTION

206.9500

ITERATION COUNT, LIMIT

OSL Release 2,

S U M M A R Y

5.398

20000.000

4393

1000000000

---

386/486 DOS

1.3.051-030

Using defaults instead.

Work space requested by user

-Work space requested by solver -OSL step: Reading data

10.00 Mb

0.61 Mb

0.00 Seconds

0.00 Seconds

Range of matrix coeffients:

before scaling :

0.0100 50.0000

after scaling :

0.0476 1.0000

OSL step: Presolve

Size reduction:

rows

:

columns :

nonzeroes:

0.01 Seconds

552 (old)

337 (old)

1445 (old)

354 (new)

337 (new)

1013 (new)

_____________________________________________________________________

127

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Crash option : 1

0.00 Seconds

0.13 Seconds

Status:

Successful (optimal)

Iterations:

309

Objective :

73.7155

OSL step: Branch&Bound

Status:

Interrupt

Iterations:

4079

Nodes

:

366

Objective :

206.9500

5.06 Seconds

0.01 Seconds

0.00 Seconds

Status:

Successful (optimal)

AMS 2.25.089 DOS Extended/C

08/15/03 23:39:16

PAGE

45

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 443

Iterations:

Objective :

5

206.9500

Bound on best integer solution:

Objective value of this solution:

73.715476

189.91861

206.95000

Optcr

: .10000 Optca:

17.031385

0.0

0

NONOPT

0 INFEASIBLE

0 UNBOUNDED

PAGE

71

G e n e r a l

A l g e b r a i c

E x e c u t i o n

---time

1 2.000,

---1 3.000,

445 VARIABLE

08/15/03 23:39:16

M o d e l i n g

S y s t e m

TF.L

TL.L

departure time

VV.L

2 7.000

445 VARIABLE

2 8.000

---445 VARIABLE

vessel v at

time t

_____________________________________________________________________

128

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

1

100.000

2

100.000

100.000

50.000

100.000

100.000

50.000

---445 VARIABLE

tank i at

VS.L

time t

1

25.000

2

75.000

10.000

5.000

45.000

15.000

35.000

25.000

35.000

15.000

35.000

---445 VARIABLE

tank j at

VB.L

time t

50.000

90.000

100.000

50.000

10.000

6

1

50.000

80.000

2

50.000

50.000

+

100.000

1

70.000

60.000

2

50.000

90.000

AMS 2.25.089 DOS Extended/C

PAGE

72

G e n e r a l

A l g e b r a i c

E x e c u t i o n

---445 VARIABLE

v starts

XF.L

08/15/03 23:39:16

M o d e l i n g

0-1

S y s t e m

unloading at time t

2

1

2

1.000

1.000

_____________________________________________________________________

129

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

---v

445 VARIABLE

XL.L

0-1

3

1

2

1.000

---v is

445 VARIABLE

1.000

XW.L

0-1

2

1

2

1.000

1.000

---tank j

445 VARIABLE

VKB.L

1.000

1.000

at time t

1.000

1.400

1.500

2.250

0.450

6

1

1.200

2

2.250

1.000

2.500

4.500

1

2

1.050

2.250

0.900

4.650

---v to

445 VARIABLE

FVS.L

storage tank i at time t

1.1

2.2

---tank i

50.000

50.000

445 VARIABLE

FSB.L

50.000

50.000

to charging tank j at time t

40.000

40.000

10.000

7

1.1

1.2

2.1

20.000

15.000

10.000

_____________________________________________________________________

130

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

2.2

40.000

30.000

---445 VARIABLE

charging tank

35.000

FBC.L

from

j to CDU l at time t

1

6

1.1

10.000

2.1

50.000

20.000

50.000

40.000

1.1

10.000

10.000

AMS 2.25.089 DOS Extended/C

PAGE

74

G e n e r a l

A l g e b r a i c

E x e c u t i o n

---from

445 VARIABLE

10.000

Z.L

08/15/03 23:39:16

M o d e l i n g

S y s t e m

crude blended j to y at time t in

CDU l

INDEX 1 = 1

2

2.1

1.000

INDEX 1 = 2

5

1.1

1.000

MW2

Academic License

G951213:1201CR-

INPUT

OUTPUT

Z:\GAMS\CASE-LEE1.TXT

D:\GAMS\CASE-LEE1.LST

MODEL STATISTICS

BLOCKS OF EQUATIONS

40

SINGLE EQUATIONS

1807

_____________________________________________________________________

131

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

BLOCKS OF VARIABLES

NON ZERO ELEMENTS

GENERATION TIME

20

5018

=

SINGLE VARIABLES

DISCRETE VARIABLES

1138

396

0.220 SECONDS

EXECUTION TIME

=

0.230 SECONDS

VERID WAT-25-089

AMS 2.25.089 DOS Extended/C

08/16/03 11:09:38

PAGE

50

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 463

S O L V E

MODEL

TYPE

SOLVER

SCHEDULING

MIP

OSL

**** MODEL STATUS

**** OBJECTIVE VALUE

OBJECTIVE

DIRECTION

FROM LINE

CO

MINIMIZE

463

1 NORMAL COMPLETION

8 INTEGER SOLUTION

331.7400

ITERATION COUNT, LIMIT

OSL Release 2,

S U M M A R Y

6283.582

20000.000

1598388

1000000000

---

386/486 DOS

1.3.051-030

Using defaults instead.

Work space requested by user

-Work space requested by solver --

10.00 Mb

1.68 Mb

0.03 Seconds

0.03 Seconds

Range of matrix coeffients:

before scaling :

0.0100 60.0000

after scaling :

0.0390 1.0000

OSL step: Presolve

Size reduction:

rows

:

columns :

nonzeroes:

OSL step: Crash

Crash option : 1

0.05 Seconds

1807 (old)

1138 (old)

5018 (old)

1281 (new)

1138 (new)

3643 (new)

0.02 Seconds

1.06 Seconds

Status:

Successful (optimal)

Iterations:

698

Objective :

104.2389

OSL step: Branch&Bound

6282.19 Seconds

_____________________________________________________________________

132

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

Status:

Interrupt

Iterations:

1597684

Nodes

:

38978

Objective :

331.7400

OSL step: Postsolve

0.09 Seconds

0.03 Seconds

Status:

Successful (optimal)

AMS 2.25.089 DOS Extended/C

08/16/03 11:09:38

PAGE

51

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 463

Iterations:

Objective :

6

331.7400

Bound on best integer solution:

Objective value of this solution:

104.23889

301.63589

331.74000

Optcr

: .10000 Optca:

30.104111

0.0

0

NONOPT

0 INFEASIBLE

0 UNBOUNDED

PAGE

119

G e n e r a l

A l g e b r a i c

E x e c u t i o n

---time

1 1.000,

---1 2.000,

465 VARIABLE

2 4.000,

465 VARIABLE

2 5.000,

---465 VARIABLE

vessel v at

TF.L

08/16/03 11:09:38

M o d e l i n g

S y s t e m

3 8.000

TL.L

departure time

3 9.000

VV.L

time t

1

100.000

2

3

100.000

50.000

100.000

50.000

100.000

_____________________________________________________________________

133

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

50.000

---465 VARIABLE

tank i at

VS.L

time t

1

20.000

82.100

2

50.000

100.000

3

70.000

21.900

39.950

89.950

82.100

82.100

10.050

10.050

70.000

70.000

21.900

50.000

10

1

2

3

69.200

90.800

69.200

90.800

69.200

90.800

50.000

69.200

90.800

100.000

---465 VARIABLE

tank j at

VB.L

21.900

time t

6

1

30.000

100.000

80.000

60.000

40.000

20.000

2

50.000

30.000

10.000

3

30.000

10.000

10.000

76.000

56.000

36.000

AMS 2.25.089 DOS Extended/C

08/16/03 11:09:38

PAGE

120

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

465 VARIABLE

tank j at

VB.L

time t

2

3

44.000

16.000

24.000

8.000

4.000

4.000

---465 VARIABLE

v starts

XF.L

0-1

unloading at time t

_____________________________________________________________________

134

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

1

2

3

1.000

---v

465 VARIABLE

1.000

1.000

XL.L

0-1

2

1

2

3

1.000

---v is

465 VARIABLE

1.000

1.000

XW.L

0-1

1

1

2

3

1.000

1.000

1.000

---tank j

465 VARIABLE

1.000

1.000

1.000

VKB.L

at time t

1.600

1.200

0.800

2.400

1.800

1.200

0.300

0.250

0.400

3.185

2.385

0.180

1.176

0.976

6

1.1

0.501

2.000

0.400

1.2

0.999

3.000

0.600

2.1

1.500

0.800

2.2

1.150

0.610

3.1

1.299

0.400

1.585

3.2

0.399

0.180

0.616

AMS 2.25.089 DOS Extended/C

PAGE

121

G e n e r a l

A l g e b r a i c

E x e c u t i o n

465 VARIABLE

VKB.L

08/16/03 11:09:38

M o d e l i n g

S y s t e m

tank j

at time t

_____________________________________________________________________

135

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

2.1

2.2

3.1

3.2

1.500

0.919

0.768

0.256

0.800

0.432

0.384

0.112

0.100

0.072

0.192

0.072

---v to

465 VARIABLE

FVS.L

storage tank i at time t

50.000

50.000

50.000

50.000

9

1.1

2.2

3.3

50.000

---tank i

50.000

465 VARIABLE

FSB.L

to charging tank j at time t

1

1.1

1.2

1.3

2.1

2.2

2.3

3.2

3.3

30.050

12.900

7.850

39.950

9.200

10.050

21.900

48.100

---465 VARIABLE

charging tank

FBC.L

from

j to CDU l at time t

1

6

1.2

20.000

20.000

20.000

20.000

20.000

2.1

20.000

20.000

10.000

3.1

20.000

20.000

20.000

3.2

20.000

AMS 2.25.089 DOS Extended/C

08/16/03 11:09:38

PAGE

123

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

_____________________________________________________________________

136

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

charging

465 VARIABLE

tank

FBC.L

from

j to CDU l at time t

+

10

2.2

3.1

20.000

8.000

20.000

4.000

4.000

4.000

6.000

4.000

---from

465 VARIABLE

Z.L

crude blended j to y at time t in

CDU l

INDEX 1 = 1

7

2.2

1.000

INDEX 1 = 2

4

3.1

1.000

INDEX 1 = 3

2

1.2

1.000

MW2

Academic License

G951213:1201CR-

INPUT

OUTPUT

Z:\GAMS\CASE-LEE2.TXT

D:\GAMS\CASE-LEE2.LST

MODEL STATISTICS

BLOCKS OF EQUATIONS

BLOCKS OF VARIABLES

NON ZERO ELEMENTS

GENERATION TIME

47

23

7760

=

SINGLE EQUATIONS

SINGLE VARIABLES

DISCRETE VARIABLES

2991

1399

582

0.230 SECONDS

_____________________________________________________________________

137

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

EXECUTION TIME

=

0.240 SECONDS

VERID WAT-25-089

AMS 2.25.089 DOS Extended/C

08/16/03 22:22:43

PAGE

58

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 530

S O L V E

MODEL

TYPE

SOLVER

S U M M A R Y

SCHEDULING

MIP

OSL

**** MODEL STATUS

**** OBJECTIVE VALUE

OBJECTIVE

DIRECTION

FROM LINE

1 NORMAL COMPLETION

8 INTEGER SOLUTION

248.6400

ITERATION COUNT, LIMIT

OSL Release 2,

CO

MINIMIZE

530

18113.945

200000.000

4236122

1000000000

---

386/486 DOS

1.3.051-030

Using defaults instead.

Work space requested by user

-Work space requested by solver --

10.00 Mb

2.48 Mb

0.02 Seconds

0.02 Seconds

Range of matrix coeffients:

before scaling :

0.0100 50.0000

after scaling :

0.0317 1.0000

OSL step: Presolve

Size reduction:

rows

:

columns :

nonzeroes:

0.07 Seconds

2991 (old)

1399 (old)

7760 (old)

Crash option : 1

2296 (new)

1399 (new)

6132 (new)

0.02 Seconds

1.70 Seconds

Status:

Successful (optimal)

Iterations:

816

Objective :

107.6704

OSL step: Branch&Bound

Status:

Interrupt

Iterations:

4235306

Nodes

:

87253

Objective :

248.6400

OSL step: Postsolve

18111.19 Seconds

0.40 Seconds

_____________________________________________________________________

138

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

0.07 Seconds

Status:

Successful (optimal)

AMS 2.25.089 DOS Extended/C

08/16/03 22:22:43

PAGE

59

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 530

Iterations:

Objective :

0

248.6400

Bound on best integer solution:

Objective value of this solution:

107.67042

226.06606

248.64000

Optcr

: .10000 Optca:

22.573941

0.0

0

NONOPT

0 INFEASIBLE

0 UNBOUNDED

PAGE

158

G e n e r a l

A l g e b r a i c

E x e c u t i o n

---time

532 VARIABLE

1 1.000,

---1

2 5.000,

532 VARIABLE

2.000,

TF.L

M o d e l i n g

S y s t e m

3 9.000

TL.L

6.000,

---532 VARIABLE

vessel v at

08/16/03 22:22:43

departure time

3 10.000

VV.L

time t

1

50.000

2

3

25.000

25.000

50.000

25.000

10

---532 VARIABLE

tank i at

50.000

VS.L

time t

_____________________________________________________________________

139

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

1

20.000

57.000

2

20.000

20.000

3

20.000

40.000

45.000

70.000

57.000

57.000

20.000

20.000

20.000

20.000

20.000

20.000

15.000

15.000

10

11

1

57.000

54.375

2

20.000

45.000

3

65.000

78.625

57.000

54.375

54.375

54.375

20.000

20.000

20.000

45.000

65.000

53.625

78.625

78.625

+

12

---532 VARIABLE

tank j at

VB.L

time t

1

30.000

28.000

2

50.000

6.000

3

30.000

30.000

30.000

48.000

38.000

40.000

30.000

20.000

10.000

20.000

10.000

10

11

1

18.000

8.000

6.000

4.000

2.000

2

4.000

2.000

3

14.000

4.000

2.000

AMS 2.25.089 DOS Extended/C

08/16/03 22:22:43

PAGE

159

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

---532 VARIABLE

v starts

XF.L

0-1

unloading at time t

1

1

2

3

1.000

---v

532 VARIABLE

1.000

1.000

XL.L

0-1

_____________________________________________________________________

140

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

2

1

2

3

1.000

---v is

532 VARIABLE

10

1.000

1.000

XW.L

0-1

1

1.000

1.000

1.000

1.000

10

1

2

3

1.000

1.000

---532 VARIABLE

tank i at

VKS.L

time t

0.400

0.650

0.900

0.770

0.770

1.000

1.000

1.000

1.000

1.000

1.600

1.600

1.600

1.150

1.150

10

11

0.744

0.744

0.744

1.000

1.000

2.500

4.376

5.876

5.876

6

1

0.770

2

1.000

3

3.275

+

12

1

0.770

0.770

0.744

2

1.000

1.000

2.500

3

5.400

5.400

5.876

AMS 2.25.089 DOS Extended/C

PAGE

160

G e n e r a l

A l g e b r a i

E x e c u t i o n

---tank j

532 VARIABLE

VKB.L

08/16/03 22:22:43

c

M o d e l i n g

S y s t e m

at time t

_____________________________________________________________________

141

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

1

0.880

2

0.270

3

0.900

0.900

0.900

1.480

1.130

2.500

2.000

1.550

0.900

0.450

2.400

1.550

0.800

10

11

1

2

3

0.530

0.180

0.280

0.090

0.210

0.140

0.070

1.050

0.300

0.150

---v to

532 VARIABLE

FVS.L

storage tank i at time t

25.000

25.000

25.000

25.000

10

1.1

2.3

3.2

25.000

3.3

---tank i

25.000

532 VARIABLE

FSB.L

to charging tank j at time t

3

1.1

1.3

3.1

3.3

13.000

2.625

5.000

11.375

---532 VARIABLE

charging tank

FBC.L

from

j to CDU l at time t

1

10.000

10.000

10.000

4.000

6

1.2

10.000

2.1

2.000

3.2

+

10.000

10.000

10.000

10.000

10.000

10.000

10

11

10.000

2.000

2.000

2.000

2.000

12

1.2

2.000

_____________________________________________________________________

142

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

2.1

2.000

2.000

3.1

10.000

2.000

2.000

6.000

AMS 2.25.089 DOS Extended/C

08/16/03 22:22:43

PAGE

162

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

---from

532 VARIABLE

Z.L

crude blended j to y at time t in

CDU l

INDEX 1 = 2

9

3.1

1.000

INDEX 1 = 3

4

1.2

1.000

MW2

Academic License

G951213:1201CR-

INPUT

OUTPUT

Z:\GAMS\CASE-LEE3.TXT

D:\GAMS\CASE-LEE3.LST

MODEL STATISTICS

BLOCKS OF EQUATIONS

39

SINGLE EQUATIONS

1682

BLOCKS OF VARIABLES

20

SINGLE VARIABLES

863

NON ZERO ELEMENTS

4779

DISCRETE VARIABLES

261

GENERATION TIME

=

0.240 SECONDS

EXECUTION TIME

=

0.260 SECONDS

VERID WAT-25-089

GAMS 2.25.089 DOS Extended/C

07/29/03 22:21:40

PAGE

49

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 460

S O L V E

MODEL

TYPE

SOLVER

SCHEDULING

MIP

OSL

S U M M A R Y

OBJECTIVE

DIRECTION

FROM LINE

CO

MINIMIZE

460

_____________________________________________________________________

143

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

**** MODEL STATUS

**** OBJECTIVE VALUE

1 NORMAL COMPLETION

8 INTEGER SOLUTION

338.9642

ITERATION COUNT, LIMIT

OSL Release 2,

1913.820

20000.000

397133

1000000000

---

386/486 DOS

1.3.051-030

Using defaults instead.

Work space requested by user

-Work space requested by solver --

10.00 Mb

1.56 Mb

0.01 Seconds

0.09 Seconds

Range of matrix coeffients:

before scaling :

0.0025 250.0000

after scaling :

0.0251 1.0000

OSL step: Presolve

Size reduction:

rows

:

columns :

nonzeroes:

0.02 Seconds

1682 (old)

863 (old)

4779 (old)

Crash option : 1

1021 (new)

863 (new)

3197 (new)

0.01 Seconds

0.51 Seconds

Status:

Successful (optimal)

Iterations:

650

Objective :

78.3530

OSL step: Branch&Bound

Status:

Interrupt

Iterations:

396480

Nodes

:

19128

Objective :

338.9642

OSL step: Postsolve

1912.54 Seconds

0.11 Seconds

0.02 Seconds

Status:

Successful (optimal)

GAMS 2.25.089 DOS Extended/C

07/29/03 22:21:40

PAGE

50

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 460

Iterations:

Objective :

3

338.9642

Bound on best integer solution:

78.353009

308.17602

_____________________________________________________________________

144

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

338.96425

Optcr

: .10000 Optca:

30.788228

0.0

**** REPORT SUMMARY :

0

NONOPT

0 INFEASIBLE

0 UNBOUNDED

GAMS 2.25.089 DOS Extended/C

07/29/03 22:21:40

PAGE

110

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

---time

1

1.000,

---1

462 VARIABLE

2

6.000,

462 VARIABLE

2.000,

TF.L

3 11.000

TL.L

7.000,

---462 VARIABLE

vessel v at

departure time

3 12.000

VV.L

time t

1

400.000

2

3

400.000

250.000

400.000

250.000

10

11

12

250.000

400.000

---462 VARIABLE

tank i at

VS.L

time t

1

30.000

291.000

2

200.000

5.000

3

140.000

70.000

115.000

316.500

291.000

291.000

135.000

135.000

70.000

5.000

138.500

138.500

73.500

70.000

_____________________________________________________________________

145

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

+

12

10

11

1

226.000

142.250

2

90.000

198.750

3

70.000

105.000

161.000

161.000

142.250

142.250

300.000

300.000

235.000

198.750

70.000

70.000

5.000

5.000

14

15

13

1

77.250

27.250

27.250

2

133.750

83.750

83.750

3

350.000

350.000

350.000

GAMS 2.25.089 DOS Extended/C

07/29/03 22:21:40

PAGE

111

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

---462 VARIABLE

tank j at

VB.L

time t

1

70.000

15.000

2

190.000

250.000

201.500

250.000

168.000

90.000

108.000

26.000

181.500

250.000

10

11

1

145.000

15.000

2

168.000

250.000

250.000

250.000

168.000

90.000

90.000

15.000

163.750

200.000

+

12

13

14

15

1

2

150.000

168.000

250.000

90.000

250.000

15.000

---462 VARIABLE

v starts

XF.L

0-1

unloading at time t

1

1

2

3

11

1.000

1.000

1.000

_____________________________________________________________________

146

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

---v

462 VARIABLE

XL.L

0-1

2

1

2

3

1.000

---v is

462 VARIABLE

12

1.000

1.000

XW.L

0-1

1

11

12

1

1.000

1.000

2

1.000

1.000

3

1.000

1.000

GAMS 2.25.089 DOS Extended/C

07/29/03 22:21:40

PAGE

112

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

---tank j

462 VARIABLE

VKB.L

at time t

1.106

3.781

4.266

2.626

1.066

7.391

4.040

1.170

6.625

8.750

10

11

2.750

4.600

4.600

2.960

1.650

5.880

3.150

0.525

5.912

7.000

13

14

15

1

2

3.000

6.630

5.000

3.900

5.000

0.675

6

1

0.150

2

8.750

+

12

1

0.150

2

9.500

PAGE

113

G e n e r a l

A l g e b r a i c

E x e c u t i o n

07/29/03 22:21:40

M o d e l i n g

S y s t e m

_____________________________________________________________________

147

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

---v to

462 VARIABLE

FVS.L

storage tank i at time t

1.1

150.000

2.2

3.3

250.000

250.000

150.000

250.000

11

12

---tank i

462 VARIABLE

150.000

FSB.L

to charging tank j at time t

65.000

48.500

7

1.1

65.000

1.2

2.1

40.000

2.2

3.1

3.2

+

1.1

1.2

2.1

2.2

3.1

3.2

65.000

25.500

65.000

65.000

65.000

65.000

65.000

3.500

11

12

13

65.000

50.000

65.000

50.000

1.500

9

10

18.750

65.000

36.250

5.000

65.000

---462 VARIABLE

charging tank

50.000

FBC.L

from

j to CDU l at time t

1

82.000

82.000

78.000

75.000

82.000

78.000

75.000

10

11

82.000

78.000

75.000

6

1.1

2.1

82.000

+

12

1.1

2.1

82.000

_____________________________________________________________________

148

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

13

14

15

2.1

78.000

75.000

82.000

GAMS 2.25.089 DOS Extended/C

07/29/03 22:21:40

PAGE

114

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

---462 VARIABLE

crude oil

D.L

blended j charges CDU l at time t

1.000

1.000

1.000

1.000

1.000

10

11

1.000

1.000

1.000

1.000

1.000

13

14

15

2.1

1.000

1.000

1.000

6

1.1

2.1

1.000

+

12

1.1

2.1

1.000

---from

462 VARIABLE

Z.L

crude blended j to y at time t in

CDU l

INDEX 1 = 1

6

12

2.1

1.000

1.000

GAMS 2.25.089 DOS Extended/C

PAGE

115

G e n e r a l

A l g e b r a i c

E x e c u t i o n

462 VARIABLE

Z.L

07/29/03 22:21:40

M o d e l i n g

S y s t e m

from

crude blended j to y at time t in

CDU l

INDEX 1 = 2

3

_____________________________________________________________________

149

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

1.1

1.000

1.000

MW2

Academic License

G951213:1201CR-

INPUT

OUTPUT

Z:\GAMS\CASE1.TXT

D:\GAMS\CASE1.LST

MODEL STATISTICS

BLOCKS OF EQUATIONS

39

SINGLE EQUATIONS

1682

BLOCKS OF VARIABLES

20

SINGLE VARIABLES

863

NON ZERO ELEMENTS

4779

DISCRETE VARIABLES

261

GENERATION TIME

=

0.310 SECONDS

EXECUTION TIME

=

0.350 SECONDS

VERID WAT-25-089

GAMS 2.25.089 DOS Extended/C

07/30/03 00:47:40

PAGE

49

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 458

S O L V E

MODEL

TYPE

SOLVER

SCHEDULING

MIP

OSL

**** MODEL STATUS

**** OBJECTIVE VALUE

OBJECTIVE

DIRECTION

FROM LINE

CO

MINIMIZE

458

1 NORMAL COMPLETION

8 INTEGER SOLUTION

450.8241

ITERATION COUNT, LIMIT

OSL Release 2,

S U M M A R Y

11438.500

160000.000

886901

900000000

---

386/486 DOS

1.3.051-030

Using defaults instead.

Work space requested by user

-Work space requested by solver -OSL step: Reading data

10.00 Mb

1.56 Mb

0.05 Seconds

0.02 Seconds

Range of matrix coeffients:

before scaling :

0.0025 250.0000

_____________________________________________________________________

150

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

after

scaling :

Size reduction:

rows

:

columns :

nonzeroes:

0.0251 -

1.0000

0.07 Seconds

1682 (old)

863 (old)

4779 (old)

Crash option : 1

1021 (new)

863 (new)

3197 (new)

0.01 Seconds

1.04 Seconds

Status:

Successful (optimal)

Iterations:

624

Objective :

87.0352

OSL step: Branch&Bound

Status:

Interrupt

Iterations:

886272

Nodes

:

61981

Objective :

450.8241

11436.02 Seconds

0.58 Seconds

0.06 Seconds

Status:

Successful (optimal)

GAMS 2.25.089 DOS Extended/C

07/30/03 00:47:40

PAGE

50

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 458

Iterations:

Objective :

5

450.8241

Bound on best integer solution:

Objective value of this solution:

Relative gap: .07415 Absolute gap:

Optcr

: .10000 Optca:

87.035176

419.70283

450.82410

31.121275

0.10000000

Extended/C

07/30/03 00:47:40 PAGE

110

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

---time

1

1.000,

---1

460 VARIABLE

2

6.000,

460 VARIABLE

5.000,

TF.L

9.000,

3 10.000

TL.L

departure time

3 12.000

_____________________________________________________________________

151

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

---460 VARIABLE

vessel v at

VV.L

time t

1

550.000

2

550.000

510.000

510.000

475.000

250.000

10

11

2

540.000

3

250.000

475.000

250.000

550.000

500.000

+

12

---460 VARIABLE

tank i at

VS.L

time t

1

30.000

350.000

2

200.000

18.800

3

140.000

65.200

5.000

5.000

5.000

165.000

135.000

93.000

93.000

28.000

130.200

65.200

65.200

65.200

10

11

1

350.000

220.000

2

5.000

315.000

3

55.000

350.000

350.000

350.000

285.000

220.000

5.000

165.000

350.000

315.000

55.000

55.000

50.000

100.000

+

12

13

14

15

1

2

3

193.400

250.000

535.000

164.000

250.000

486.000

164.000

250.000

486.000

PAGE

111

G e n e r a l

A l g e b r a i c

E x e c u t i o n

07/30/03 00:47:40

M o d e l i n g

S y s t e m

_____________________________________________________________________

152

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

---460 VARIABLE

tank j at

VB.L

time t

1

70.000

250.000

2

190.000

86.000

209.800

127.800

45.800

175.800

108.000

215.000

250.000

168.000

10

11

1

168.000

250.000

2

120.000

15.000

90.000

15.000

150.000

250.000

185.000

250.000

168.000

90.000

+

12

13

14

15

1

2

168.000

171.600

90.000

250.000

15.000

250.000

---460 VARIABLE

v starts

XF.L

0-1

unloading at time t

1

1

2

3

1.000

---v

460 VARIABLE

10

1.000

1.000

XL.L

0-1

5

1

2

3

1.000

---v is

460 VARIABLE

12

1.000

1.000

XW.L

0-1

1

1.000

1.000

1.000

1.000

7

1

2

1.000

1.000

_____________________________________________________________________

153

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

PAGE

112

G e n e r a l

A l g e b r a i c

E x e c u t i o n

460 VARIABLE

XW.L

07/30/03 00:47:40

M o d e l i n g

0-1

S y s t e m

v is

unloading its crude oil at time t

+

2

3

1.000

1.000

---tank j

460 VARIABLE

VKB.L

10

11

12

1.000

1.000

1.000

at time t

1.106

4.196

2.556

0.916

3.516

7.391

4.521

9.031

9.381

6.511

10

11

2.802

1.650

0.150

3.000

4.700

4.865

6.815

8.765

5.895

3.165

13

14

15

1

2

3.060

6.006

1.500

8.750

0.150

8.750

---v to

460 VARIABLE

6

1

4.442

2

3.641

+

12

1

4.700

2

0.540

FVS.L

storage tank i at time t

40.000

35.000

225.000

250.000

7

1.1

2.2

65.000

10.000

2.2

3.3

225.000

250.000

10

11

12

50.000

250.000

250.000

_____________________________________________________________________

154

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

---tank i

460 VARIABLE

FSB.L

to charging tank j at time t

65.000

65.000

65.000

9.200

10

12

65.000

65.000

6

1.1

1.2

2.1

2.2

23.800

3.1

3.2

10.200

+

65.000

35.000

65.000

42.000

9.800

65.000

7

13

1.1

1.2

29.400

2.1

2.2

3.1

3.2

49.000

26.600

65.000

65.000

35.000

65.000

65.000

5.000

65.000

PAGE

114

G e n e r a l

A l g e b r a i c

E x e c u t i o n

---460 VARIABLE

charging tank

FBC.L

07/30/03 00:47:40

M o d e l i n g

S y s t e m

from

j to CDU l at time t

1

82.000

82.000

82.000

82.000

10

11

82.000

78.000

75.000

6

1.1

82.000

2.1

+

82.000

7

78.000

75.000

12

1.1

82.000

2.1

+

13

14

15

1.1

78.000

75.000

82.000

_____________________________________________________________________

155

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

---460 VARIABLE

crude oil

D.L

blended j charges CDU l at time t

1.000

1.000

1.000

1.000

10

11

1.000

1.000

1.000

6

1.1

1.000

2.1

1.000

1.000

1.000

12

1.1

1.000

2.1

+

13

14

15

1.1

1.000

1.000

1.000

---from

460 VARIABLE

Z.L

crude blended j to y at time t in

CDU l

INDEX 1 = 1

2.1

1.000

1.000

INDEX 1 = 2

1.1

12

1.000

1.000

1.000

MW2

Academic License

G951213:1201CR-

INPUT

OUTPUT

Z:\GAMS\CASE2.TXT

D:\GAMS\CASE2.LST

MODEL STATISTICS

_____________________________________________________________________

156

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

BLOCKS OF EQUATIONS

39

SINGLE EQUATIONS

1682

BLOCKS OF VARIABLES

20

SINGLE VARIABLES

863

NON ZERO ELEMENTS

4779

DISCRETE VARIABLES

261

GENERATION TIME

=

0.430 SECONDS

EXECUTION TIME

=

0.450 SECONDS

VERID WAT-25-089

GAMS 2.25.089 DOS Extended/C

07/30/03 01:13:21

PAGE

49

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 460

S O L V E

MODEL

TYPE

SOLVER

S U M M A R Y

SCHEDULING

MIP

OSL

**** MODEL STATUS

**** OBJECTIVE VALUE

OBJECTIVE

DIRECTION

FROM LINE

1 NORMAL COMPLETION

8 INTEGER SOLUTION

809.9055

ITERATION COUNT, LIMIT

OSL Release 2,

CO

MINIMIZE

460

34308.601

200000.000

4631527

900000000

---

386/486 DOS

1.3.051-030

Using defaults instead.

Work space requested by user

-Work space requested by solver --

10.00 Mb

1.56 Mb

0.01 Seconds

0.02 Seconds

Range of matrix coeffients:

before scaling :

0.0025 250.0000

after scaling :

0.0220 1.0000

OSL step: Presolve

Size reduction:

rows

:

columns :

nonzeroes:

0.04 Seconds

1682 (old)

863 (old)

4779 (old)

Crash option : 1

1021 (new)

863 (new)

3197 (new)

0.01 Seconds

1.19 Seconds

Status:

Successful (optimal)

Iterations:

597

Objective :

267.0352

OSL step: Branch&Bound

Status:

Interrupt

Iterations:

4630923

Nodes

:

361178

Objective :

809.9055

34306.05 Seconds

_____________________________________________________________________

157

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

0.67 Seconds

0.12 Seconds

Status:

Successful (optimal)

GAMS 2.25.089 DOS Extended/C

07/30/03 01:13:21

PAGE

50

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 460

Iterations:

Objective :

7

809.9055

Bound on best integer solution:

Objective value of this solution:

267.03518

736.29773

809.90550

Optcr

: .10000 Optca:

73.607773

10.000000

PAGE

110

G e n e r a l

A l g e b r a i c

E x e c u t i o n

---time

1

1.000,

---1

462 VARIABLE

2

6.000,

462 VARIABLE

5.000,

TF.L

M o d e l i n g

S y s t e m

3 10.000

TL.L

9.000,

---462 VARIABLE

vessel v at

07/30/03 01:13:21

departure time

3 12.000

VV.L

time t

1

550.000

2

550.000

526.000

461.000

461.000

250.000

10

11

2

540.000

3

250.000

475.000

250.000

550.000

500.000

+

12

_____________________________________________________________________

158

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

---462 VARIABLE

tank i at

VS.L

time t

1

30.000

350.000

2

200.000

5.000

3

140.000

39.000

5.000

5.000

5.000

165.000

135.000

134.000

69.000

62.000

140.000

140.000

75.000

39.000

10

11

1

318.000

219.000

2

5.000

350.000

3

39.000

305.000

318.000

318.000

253.000

219.000

5.000

165.000

350.000

350.000

5.000

5.000

5.000

55.000

+

12

13

14

15

1

2

3

193.000

285.000

490.000

193.000

225.750

470.250

193.000

225.750

470.250

PAGE

111

G e n e r a l

A l g e b r a i c

E x e c u t i o n

---462 VARIABLE

tank j at

VB.L

07/30/03 01:13:21

M o d e l i n g

S y s t e m

time t

1

70.000

208.000

2

190.000

168.000

184.000

250.000

168.000

86.000

108.000

26.000

156.000

250.000

10

11

1

250.000

250.000

2

86.000

15.000

168.000

86.000

216.000

250.000

185.000

250.000

168.000

90.000

14

15

+

12

13

_____________________________________________________________________

159

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

1

2

168.000

171.000

90.000

250.000

---462 VARIABLE

v starts

XF.L

15.000

250.000

0-1

unloading at time t

1

1

2

3

1.000

---v

462 VARIABLE

10

1.000

1.000

XL.L

0-1

5

1

2

3

1.000

---v is

462 VARIABLE

12

1.000

1.000

XW.L

0-1

1

7

1

1.000

1.000

1.000

1.000

2

1.000

1.000

GAMS 2.25.089 DOS Extended/C

07/30/03 01:13:21

PAGE

112

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

462 VARIABLE

XW.L

0-1

v is

unloading its crude oil at time t

+

2

3

---tank j

1.000

1.000

462 VARIABLE

VKB.L

10

11

12

1.000

1.000

1.000

volume of component k in charging

at time t

_____________________________________________________________________

160

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

1

3.306

2

6.020

1.106

3.546

4.226

2.586

0.946

7.391

4.040

1.170

6.370

8.890

10

11

3.926

2.730

1.090

3.690

4.030

3.150

6.800

8.750

5.880

3.150

13

14

15

1

2

3.210

5.985

1.650

8.750

0.150

8.750

+

12

1

4.030

2

0.525

PAGE

113

G e n e r a l

A l g e b r a i c

E x e c u t i o n

---v to

462 VARIABLE

FVS.L

07/30/03 01:13:21

M o d e l i n g

S y s t e m

storage tank i at time t

24.000

65.000

211.000

250.000

7

1.1

2.2

65.000

10.000

2.2

3.3

225.000

250.000

---tank i

462 VARIABLE

FSB.L

10

11

12

50.000

250.000

250.000

to charging tank j at time t

49.000

65.000

65.000

1.000

6

1.1

32.000

1.2

2.1

10.000

2.2

3.2

65.000

51.000

57.000

65.000

65.000

7.000

36.000

_____________________________________________________________________

161

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

10

65.000

34.000

12

13

1.1

1.2

2.1

2.2

59.250

3.2

19.750

26.000

65.000

65.000

65.000

65.000

34.000

---462 VARIABLE

charging tank

65.000

FBC.L

from

j to CDU l at time t

1

6

1.1

82.000

82.000

2.1

82.000

82.000

82.000

82.000

GAMS 2.25.089 DOS Extended/C

07/30/03 01:13:21

PAGE

114

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

charging

462 VARIABLE

tank

FBC.L

from

j to CDU l at time t

+

82.000

82.000

10

11

82.000

78.000

75.000

12

1.1

82.000

2.1

+

13

14

15

1.1

78.000

75.000

75.000

---462 VARIABLE

crude oil

D.L

blended j charges CDU l at time t

1.000

1.000

1.000

1.000

6

1.1

2.1

1.000

+

1.000

9

10

11

12

_____________________________________________________________________

162

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

1.1

1.000

2.1

1.000

1.000

1.000

13

14

15

1.1

1.000

1.000

1.000

PAGE

115

G e n e r a l

A l g e b r a i c

E x e c u t i o n

---from

462 VARIABLE

Z.L

1.000

1.000

07/30/03 01:13:21

M o d e l i n g

S y s t e m

crude blended j to y at time t in

CDU l

INDEX 1 = 1

2.1

1.000

1.000

INDEX 1 = 2

1.1

12

1.000

1.000

1.000

EXECUTION TIME

19.730 SECONDS

MW2

Academic License

VERID WAT-25-089

G951213:1201CR-

INPUT

OUTPUT

Z:\GAMS\CASE3.TXT

D:\GAMS\CASE3.LST

MODEL STATISTICS

BLOCKS OF EQUATIONS

39

SINGLE EQUATIONS

1682

BLOCKS OF VARIABLES

20

SINGLE VARIABLES

863

NON ZERO ELEMENTS

4779

DISCRETE VARIABLES

261

GENERATION TIME

=

0.240 SECONDS

EXECUTION TIME

=

0.290 SECONDS

VERID WAT-25-089

GAMS 2.25.089 DOS Extended/C

07/31/03 08:53:26

PAGE

49

_____________________________________________________________________

163

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

Solution Report

SOLVE SCHEDULING USING MIP FROM LINE 460

S O L V E

MODEL

TYPE

SOLVER

S U M M A R Y

SCHEDULING

MIP

OSL

**** MODEL STATUS

**** OBJECTIVE VALUE

OBJECTIVE

DIRECTION

FROM LINE

1 NORMAL COMPLETION

8 INTEGER SOLUTION

329.1860

ITERATION COUNT, LIMIT

OSL Release 2,

CO

MINIMIZE

460

7532.652

120000.000

350904

900000000

---

386/486 DOS

1.3.051-030

Using defaults instead.

Work space requested by user

-Work space requested by solver --

10.00 Mb

1.56 Mb

0.03 Seconds

0.02 Seconds

Range of matrix coeffients:

before scaling :

0.0025 250.0000

after scaling :

0.0251 1.0000

OSL step: Presolve

Size reduction:

rows

:

columns :

nonzeroes:

0.05 Seconds

1682 (old)

863 (old)

4779 (old)

Crash option : 1

1021 (new)

863 (new)

3197 (new)

0.01 Seconds

0.72 Seconds

Status:

Successful (optimal)

Iterations:

650

Objective :

78.3530

OSL step: Branch&Bound

Status:

Interrupt

Iterations:

350251

Nodes

:

15853

Objective :

329.1860

OSL step: Postsolve

7524.74 Seconds

4.02 Seconds

0.94 Seconds

Status:

Successful (optimal)

_____________________________________________________________________

164

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

0

NONOPT

0 INFEASIBLE

0 UNBOUNDED

PAGE

110

G e n e r a l

A l g e b r a i c

E x e c u t i o n

---time

1

1.000,

---1

462 VARIABLE

2

TF.L

6.000,

462 VARIABLE

2.000,

M o d e l i n g

S y s t e m

3 10.000

TL.L

7.000,

---462 VARIABLE

vessel v at

07/31/03 08:53:26

departure time

3 11.000

VV.L

time t

1

400.000

2

3

250.000

250.000

400.000

250.000

10

11

400.000

---462 VARIABLE

tank i at

VS.L

time t

1

30.000

253.100

2

200.000

5.000

3

140.000

7.900

115.000

300.000

263.600

253.100

135.000

102.700

37.700

5.000

137.300

137.300

72.300

7.900

10

11

1

188.100

114.500

2

90.000

145.000

3

7.900

314.500

137.100

137.100

126.760

114.500

275.000

275.000

210.000

145.000

7.900

7.900

5.000

90.000

+

12

_____________________________________________________________________

165

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

13

14

15

1

2

3

49.500

80.000

308.000

5.000

15.000

308.000

5.000

15.000

308.000

---462 VARIABLE

tank j at

VB.L

time t

6

1

70.000

202.700

300.000

218.000

136.000

54.000

2

190.000

108.000

26.000

192.400

300.000

300.000

GAMS 2.25.089 DOS Extended/C

07/31/03 08:53:26

PAGE

111

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

462 VARIABLE

tank j at

VB.L

time t

+

12

10

11

1

184.000

54.000

2

218.000

300.000

300.000

300.000

218.000

136.000

136.000

54.000

132.240

274.500

13

14

15

1

2

190.500

218.000

300.000

136.000

300.000

54.000

---462 VARIABLE

v starts

XF.L

0-1

unloading at time t

1

1

2

3

1.000

---v

462 VARIABLE

10

1.000

1.000

XL.L

0-1

2

11

_____________________________________________________________________

166

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

1

2

3

1.000

---v is

462 VARIABLE

1.000

1.000

XW.L

0-1

1

1

2

3

1.000

1.000

1.000

---tank j

462 VARIABLE

1.000

1.000

10

11

1.000

VKB.L

at time t

6

1

1.106

3.841

5.460

3.820

2.180

0.540

2

7.391

4.521

1.170

6.734

11.040

11.040

GAMS 2.25.089 DOS Extended/C

07/31/03 08:53:26

PAGE

112

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

462 VARIABLE

VKB.L

tank j

at time t

+

12

10

11

3.140

5.600

5.600

3.960

2.320

8.170

5.300

2.430

4.628

9.951

13

14

15

1

2

3.605

8.356

6.000

4.760

6.000

1.890

---v to

462 VARIABLE

1

0.680

2

11.226

FVS.L

storage tank i at time t

_____________________________________________________________________

167

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

10

11

1.1

150.000

250.000

2.2

150.000

250.000

3.3

150.000

250.000

GAMS 2.25.089 DOS Extended/C

07/31/03 08:53:26

PAGE

113

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

---tank i

462 VARIABLE

FSB.L

to charging tank j at time t

65.000

65.000

65.000

32.300

7

1.1

51.000

1.2

2.1

65.000

2.2

3.1

3.2

+

1.1

1.2

2.1

2.2

3.1

3.2

65.000

36.400

10.500

65.000

65.000

32.700

65.000

64.400

11

12

13

65.000

44.500

65.000

65.000

2.700

9

10

10.340

12.260

65.000

65.000

2.900

65.000

6.500

---462 VARIABLE

charging tank

FBC.L

25.500

volumetric flow rate

from

j to CDU l at time t

1

82.000

82.000

82.000

82.000

82.000

82.000

82.000

10

11

82.000

82.000

82.000

6

1.1

2.1

82.000

+

12

1.1

2.1

82.000

_____________________________________________________________________

168

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

13

14

15

2.1

82.000

82.000

82.000

---462 VARIABLE

crude oil

D.L

blended j charges CDU l at time t

6

1.1

1.000

1.000

1.000

2.1

1.000

1.000

1.000

GAMS 2.25.089 DOS Extended/C

07/31/03 08:53:26

PAGE

114

G e n e r a l

A l g e b r a i c

M o d e l i n g

S y s t e m

E x e c u t i o n

462 VARIABLE

crude oil

D.L

blended j charges CDU l at time t

10

11

1.000

1.000

1.000

1.000

1.000

13

14

15

2.1

1.000

1.000

1.000

12

1.1

2.1

1.000

---from

462 VARIABLE

Z.L

crude blended j to y at time t in

CDU l

INDEX 1 = 1

2.1

12

1.000

1.000

INDEX 1 = 2

1.1

1.000

1.000

EXECUTION TIME

7.090 SECONDS

VERID WAT-25-089

_____________________________________________________________________

169

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

GAMS 2.25.089

PAGE

115

G e n e r a l

DOS Extended/C

A l g e b r a i c

07/31/03 08:53:26

M o d e l i n g

MW2

Academic License

S y s t e m

G951213:1201CR-

INPUT

OUTPUT

Z:\GAMS\CASE6.TXT

D:\GAMS\CASE6.LST

APENDIX B

* Program in GAMS developed by Cassio Tamara for the Research Project in

Optimisation of Oil refineries, MSc Process System Engineering, Cranfield University

*3 vessels, 3 storage tanks, 2 charging tanks and 1 CDU. Scheduling horizon for this

problem is 15 time intervals.

* This program is configurated for running case 1.

option iterlim =1000000000

sets

i "crude oil storage tank" /1,2,3/

j "crude oil charging tank" /1,2/

y "crude oil charging tank" /1,2/

l "crude distillation unit" /1/

t "time interval" /1*15/

m "time interval" /1*15/

v "crude vessel" / 1,2,3/

parameters

VSmax (i)

_____________________________________________________________________

170

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

/ 1

350

350

350 /

VSmin (i)

VBmax (j)

VBmin (j)

/ 1

5/

/ 1

250

250/

/ 1

2

15

15 /;

parameters

CU(v) "unloading cost of vessel v per unit time interval"

/ 1

10

10

10 /

/ 1

8/

CSINV(i) "inventory cost of storage tank i per unit time per unit volume"

/ 1

0.008

0.008

0.008 /

CBINV(j) "inventory cost of charging tank j per unit time per unit volume"

/ 1

0.005

0.005/

_____________________________________________________________________

171

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

TARR(v) "crude oil vessel v arrival time around the docking station"

TLEA(v)

/1

10 /

/1

10

15 /

/ 1

0.01

0.03

0.05 /

charging tank j"

/ 1

2

0.01

0.035

charging tank j"

/

1 0.02

2 0.045

s per time interval"

/1

75

75

75

75

75

_____________________________________________________________________

172

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

75

75

75

75

10

75

11

75

12

75

13

75

14

75

15

75 / ;

variables

TF(v) vessel v unloading initiation time

TL(v) vessel v unloading completion and departure time

VV(v,t) volume of crude oil in crude vessel v at time t

VS(i,t) volume of crude oil in storage tank i at time t

VB(j,t) volume of crude oil in charging tank j at time t

XF(v,t) 0-1 variable to denote if vessel v starts unloading at time t

XL(v,t) 0-1 variable to denote if vessel v completes unloading at

time t

XFF(v,m) transitory variable to search XF

XLL(v,m) transitory variable to search XL

XW(v,t) 0-1 variable to denote if vessel v is unloading its crude

oil at time t

D(j,l,t) 0-1 variable to denote if the crude oil blended j charges

CDU l at time t

_____________________________________________________________________

173

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

CDU l at time t

Z(j,y,l,t) 0-1 variable to denote transition from crude blended j to y

at time t in CDU l

FVS(v,i,t) volumetric flow rate from vessel v to storage tank i at

time t

FSB(i,j,t) volumetric flow rate from storage tank i to charging tank

j at time t

FBC(j,l,t) volumetric flow rate from charging tank j to CDU l at

time t

FKSB(i,j,t)

FKBC(j,l,t)

VKB(j,t) volume of component k in charging tank j at time t

EB(j,t)

chargin tank j

co

Positive

variables

VV,VS,VB,FVS,FSB,FBC,X,W,XFF,XLL,

FKSB,FKBC,VKB,EB;

integer variables TF,TL;

VV.fx('

1'

,'

1'

) = 400;

VV.fx('

2'

,'

6'

) = 400;

VV.fx('

3'

,'

10'

) = 400;

VS.fx('

1'

,'

1'

) = 30;

VS.fx('

2'

,'

1'

) = 200;

_____________________________________________________________________

174

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

VS.fx('

3'

,'

1'

) = 140;

FVS.up(v,i,t) = 250;

FSB.up(i,j,t) = 65;

FBC.up(j,l,t) = 82;

VB.fx('

1'

,'

1'

) = 70;

VB.fx('

2'

,'

1'

) = 190;

EB.fx('

1'

,'

1'

) = 0.0158;

EB.fx('

2'

,'

1'

) = 0.0389;

VKB.fx('

1'

,'

1'

) = 1.106;

VKB.fx('

2'

,'

1'

) = 7.391;

Equations

cost

vearrive

veleave

unloit

unloct

unlost

vedurat

vearco

ffd

lld

fd

ld

intvess

lastvess

nounloa

vemasba

vevol1

vevol2

vtoiatt

stmasba

material balance equation for storage tank

_____________________________________________________________________

175

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

volimst1

volimst2

ratestk

ncandfst1

operating constraint

blmasba

volimbt1

volimbt2

ncandfst2

totprod

demand fulfilment

kbmasba

kitojatt

kjtolatt1

kjtolatt2

volimkbt1

volimkbt2

cdubyone

compar

makes D equal to G

changeov

calculation of Z ;

cost..

co

=e=

sum(v,(TL(v)-TF(v)+1)*CU(V))

sum(v,(TF(v)-TARR(v))*CSEA(v))

sum((j,t)$(ord(t)

le

14),((VB(j,t)

VB(j,t+1))*CBINV(j)/2))

sum((j,y,l,t),Z(j,y,l,t)*50);

vearrive(v)..

sum(t,XF(v,t))

veleave(v)..

sum(t,XL(v,t))

=e= 1.0 ;

=e= 1.0 ;

unloit(v)..

TF(v)

=e= sum(t,XF(v,t)*ord(t)) ;

unloct(v)..

TL(v)

=e= sum(t,XL(v,t)*ord(t)) ;

unlost(v)..

TF(v)

=g= TARR(v) ;

vedurat(v)..

TL(v)-TF(v)

vearco(v)$(ord(v) le 2)..

=g= 1 ;

TF(v+1)

=g= TL(v) + 1 ;

_____________________________________________________________________

176

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

ffd(v,m)..

ord(m) ge TARR(v))),XF(v,t));

lld(v,m)..

fd(v,t,m)$(ord(t) le ord(m)$(ord(m) lt TLEA(v) and ord(m) ge TARR(v)))..

XW(v,t) =l= XFF(v,m) ;

ld(v,t,m)$(ord(t) gt ord(m)$(ord(m) le TLEA(v) and ord(m) ge TARR(v) ))..

XW(v,t) =l= XLL(v,m) ;

intvess(v)$(ord(v) le 2)..

lastvess(v)$(ord(v) eq 3)..

vemasba(v,t)$(ord(t) le 14 and ord(t) ge TARR(v))..

VV(v,t+1)

=e=

VV(v,t) - sum(i,FVS(v,i,t));

vevol1(v)..

ord(v)),FVS(v,i,t));

vevol2(v)..

sum((i,t),FVS(v,i,t))

vtoiatt(v,i,t)..

=e= 400;

* all vessel have the same capacity, then VVv = 400 (equat.3.17, chapter 3)

* "Material balance equations for the storage tank"

stmasba(i,t)$(ord(t) le 14)..

sum(j,FSB(i,j,t));

volimst1(i,t)..

volimst2(i,t)..

ratestk(i,t)..

ncandfst1(i,j,l,t)..

FSB(i,j,t) =l= FSB.up(i,j,t)*(1-D(j,l,t)) ;

_____________________________________________________________________

177

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

* all storage tanks have the same pumping rate maximum capacity, then PUMPCAPi

= 70 (equat.3.25, chapter 3)

blmasba(j,t)$(ord(t) le 14)..

FBC(j,l,t));

volimbt1(j,t)..

volimbt2(j,t)..

* Solving direction

ncandfst2(j,l,t)..

totprod(t)..

kbmasba(j,t)$(ord(t) le 14)..

sum(l,FKBC(j,l,t));

kitojatt(i,j,t)..

kjtolatt1(j,l,t)..

kjtolatt2(j,l,t)..

volimkbt1(j,t)..

volimkbt2(j,t)..

_____________________________________________________________________

178

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

cdubyone(l,t)..

sum(j,D(j,l,t)) =e= 1;

compar(j,y,l,t)$(ord(j) eq ord(y))..

+ G(y,l,t+1) - 1 ;

scheduling.optfile = 1;

scheduling.workspace = 10;

option reslim = 20000;

option mip = osl;

solve scheduling using mip minimizing co;

display

TF.l,TL.l,VV.l,VS.l,VB.l,XF.l,XL.l,XW.l,VKB.l,FKBC.l,EB.l,FVS.l,FSB.l,FBC.l,D.l

,G.l,Z.l;

_____________________________________________________________________

179

Cassio R. Tamara

Oil Refinery Scheduling Optimisation

MSc Process & System Engineering

Cranfield University 2003

- What Are Opportunity Crude OilsTransféré parDebasish
- Kuwait Oil and Gas Sector_Oil & Gas Directory ME_2010Transféré parKanika Kapoor
- Chapter15 - EconomicsTransféré parAsad Raza
- PM1 Plant MaintenanceTransféré parSyedNadeemAhmed
- Oil Refining Industry in South East Asia, 2011Transféré parrajeshgunnam
- Grace Fact Sheet.Transféré parureachanit
- MB0048 Operations ResearchTransféré pardolodeep
- FullModuleLectures.pdfTransféré parvadje pritam
- CATPAPER.pdfTransféré parmustafa Insel
- Solving Difficult Multicommodity Problems (PPRN, CPLEX, EMNET, Castro, 2003)Transféré parchadspen
- Crude Oil RefiningTransféré parVijaya Kumar
- The Significance of Reducing Setup Times_setup CostsTransféré parAdrienn Pánovics
- Syllabus for New or Modified CoursesTransféré parkingraaja
- 2012 - Weight Minimization of Frames With SGA and PGATransféré parRadu Zoicas
- Lingo_15_Users_Manual.pdfTransféré parEdnilson Santos
- QP Null Space MethodTransféré parvinzklorthos
- PetroPowerChileTransféré parjojumathew
- APCOM_Haulage Impact on Mine Cut-Off Grade StrategyTransféré parJorge
- 1-s2.0-S0098135411001001-mainTransféré parFSBoll
- 01 IntroductionTransféré parLTE002
- FP_00238Transféré parAnand Retharekar
- Oil Pet ChemTransféré parAmr Mokhtar
- Ecuador Oil SectorTransféré parBryan Armas
- Branch and Bound Technique for Three Stage Flow Shop Scheduling Problem Including Breakdown Interval and Transportation TimeTransféré parAlexander Decker
- compedium of Energy Laws_volume_1_DOE_3.pdfTransféré parStarr Weigand
- Optimization Plate Fin Heat Exchanger.pdfTransféré parEmilio José Berciano Rueda
- video_2Transféré parW00W
- Operations Research-rony Andrews.Transféré parRony Andrews
- Course OutlineTransféré parfour threepio
- Consistency and Optimality in ManagerialTransféré parSanyam Madaan

- Organizational Communication ProjectTransféré parDanAlexandruJoita
- EJMiller_Urban Transport 21st CenturyTransféré parDanAlexandruJoita
- ArceeTransféré parDanAlexandruJoita
- OIML 2007 R117-1-e07Transféré parchacon123
- How to Initiate Work Agency Briefing SBlaga Dec 3 2015_SNSPATransféré parDanAlexandruJoita
- Creative Assignment Brief Agency Internal - ExampleTransféré parDanAlexandruJoita
- SPARK09tvaquero.pdfTransféré parDanAlexandruJoita
- bun bibTransféré parDanAlexandruJoita
- Semester PaperworkTransféré parDanAlexandruJoita
- Agency Assignment Brief - ExampleTransféré parDanAlexandruJoita
- Book on Pipeline Best PracticeTransféré parSkazemi7
- 1fmatterTransféré parDanAlexandruJoita
- Adrian PascuTransféré parDanAlexandruJoita
- PipeFlow2Transféré parClement Chima

- sdcTransféré parSalman Aslam
- PfcTransféré paradhyartha
- Pharmaceutical and Software Development ProjectsTransféré parAlexandar123
- CompressorTransféré paragrocel_bhv5591
- ae bd consulting - uh-60a - aircraft for sale - vip configuration 160801 a4 sizeTransféré parapi-325281362
- Direct Simple Shear Software ManualTransféré parFilipe
- Operacion Epson 3520Transféré parAnonymous lSEnEklZ7
- HowToAspNetMVCTransféré parRkdeepLove
- manual transmissionTransféré parEmad Satari
- Plastic Ruler Project Status ReportTransféré parAmin ZD
- Filtration for Aseptic ProcessingTransféré parhugo737977
- Samsung Refrigerator Manual DA68-02481A-11Transféré parKristina
- seminartopic880325897_[EDocFind.com]Transféré parsmritisingh12
- Hive and Presto for Big DataTransféré parparashar1505
- Spin Characteristics of CessnasTransféré parbft23
- Ing Diba Case StudyTransféré parAmy Adams
- Economic Impact of Derated Climb on Large Commercial EnginesTransféré parBobsunoy
- Debug ItemTransféré parnrusingh
- Compressor Packager Guide WEB NewTransféré parJaroslav Kuruc
- Hero Honda MARKETING PPT MBATransféré parBabasab Patil (Karrisatte)
- Tasdeer Exporters Directory 2013-English.pdfTransféré parMel
- ForecastInternational_AGM-119 Penguin_F658_CompleteSample.pdfTransféré parScott_B4U
- Twp Parallel Execution Fundamentals 133639Transféré paryogeeswar
- Siemens 01 Sinorix Overview BrochureTransféré parmthuya
- moti-strategyTransféré parmarklung
- Maintenance Schedule for WARTSILA 20 B-C Engines Wfi 2004N007 02gbTransféré parDimitris Paliktoglou
- livebook ACBTransféré parShubhangi Jagtap
- Dell PowerEdge C4130 Spec SheetTransféré parjeffjozo
- RN362 Digital Maturity Model Release Notes R17.0.1Transféré pardeepak_sachdeva81
- RJ Tech Data at a Glance April 2004Transféré pargabiwen