Improved Prediction

IMPROVED PREDICTION OF WAX DEPOSITION IN SUBSEA CRUDE-OIL
PIPELINES IN NIGERIA
BY
OLUSIJI AYOADE ADEYANJU,
B.Sc, M.Sc. Petroleum Engineering (University of Ibadan, Ibadan, Nigeria)
A THESIS SUBMITTED TO THE SCHOOL OF POSTGRADUATE STUDIES IN
PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY (Ph.D.), IN CHEMICAL ENGINEERING, AT THE
UNIVERSITY OF LAGOS, LAGOS, NIGERIA.
DECEMBER 2016
i
SCHOOL OF POSTGRADUATE STUDIES
UNIVERSITY OF LAGOS
CERTIFICATION
This is to certify that the thesis:
IMPROVED PREDICTION OF WAX DEPOSITION IN SUBSEA CRUDE-OIL
PIPELINES IN NIGERIA
Submitted to the
School of Postgraduate Studies
University of Lagos
For the award of the degree of

DOCTOR OF PHILOSOPHY (Ph.D.)
Is a record of original research carried out
By:
ADEYANJU, OLUSIJI AYOADE
In the Department of Chemical and Petroleum Engineering
………………………………….. ……………………… ………………

AUTHOR’S NAME SIGNATURE DATE
………………………………….. ……………………… ………………

1ST SUPERVISOR NAME SIGNATURE DATE
………………………………….. ……………………… ………………

2ND SUPERVISOR NAME SIGNATURE DATE
………………………………….. ……………………… ………………

1ST INTERNAL EXAMINER SIGNATURE DATE
………………………………….. ……………………… ………………

2ND INTERNAL EXAMINER SIGNATURE DATE
………………………………….. ……………………… ………………

EXTERNAL EXAMINER SIGNATURE DATE
………………………………….. ……………………… ………………

SPGS REPRESENTATIVE SIGNATURE DATE
ii
DEDICATION
I dedicate this thesis to the Awesome and Almighty GOD
iii
ACKNOWLEDGEMENTS
I want to thank God Almighty for the inspiration, insights, wisdom and ability given to me from
the start to the completion of this research work, including the provision of all I needed to
successfully finish the research.
My appreciation goes to my supervisors, Prof. L.O. Oyekunle and Dr. J.U. Nwalor. Their help
thorough out the period of my research work were of high value. Thank you very much sirs. God
bless you.
My gratitude goes to my wife Mrs. Oluwakemi Elizabeth Adeyanju whom words cannot explain
the level of support, encouragement and love she offered throughout the period of the work.
Also to my lovely children: Titilayo, Ayomide, Olumide, Olamide, Iremide and Ademide. Also
my sweet mother Mrs. Felicia Adeyanju for all her help and encouragement, also my late father
may his gentle soul rest in perfect peace, my siblings and in-laws.
I appreciate the effort and support offered by Prof. D. S. Aribike, Prof. O.A. Olafadehan, Prof. O.
Aiyesimoju, Dr. B. A. Olufemi, Dr. F. U. Babalola, Dr. M. Usman, Dr. A. Akinola and my
wonderful colleagues Mr. O. A. Olatunde, Mr. Y.B. Adeboye, and Mr. R.U. Owolabi. All those
on study leave like Mr. O. Adekomaya, Mr. A. Okwanaike and Mr. O. Olanipekun. Also, Prof.
O. Olafuyi of the University of Benin for his words of encouragements. God bless you all.
I appreciate the assistance of the member of staff of Chemical and Petroleum Engineering
Department, Igbinedion University, Okada, Edo State.
iv
ABSTRACT
One of the problems faced by the petroleum industry is the wax deposition during crude oil
transportation in pipelines. The challenge is even more precarious in subsea pipelines as the
precipitation and deposition of wax components in the crude oil are more pronounced. Key
factors affecting wax deposition in water-cooled pipelines were studied in a series of flow loop
experiments. Seven crude oil samples from the Nigerian sub-sea oil fields were used in the study.
The Response Surface Methodology (RSM) was used to evaluate the individual and interactive
effects of five variables (coolant temperature (Tc), bulk oil temperature (Tb), percentage of wax
inhibitor (Iw), flow rate (Q) and the wax content (Wc)) on the wax deposition rate using the
Central Composite Design (CCD). A model that relates the individual and interactive effect of
these variables to the wax deposition rate was developed. The model gave excellent fits to the
deposition data obtained from in the laboratory scale flow loop tests under different flow regimes
with R2 of 0.9751 was in reasonable agreement with the adjusted R2 of 0.9618. The volume of
wax deposit from each of the crude-oil sample also fitted to the predicted volume as the
predicted R2 = 0.9333 - 0.9790 were in reasonable agreement with adjusted R2 = 0.9155 - 0.9634.
The agreement between model and experimental wax deposits for all the crude oil samples with
Average Absolute Deviation (AAD) of 2.47% - 6.18 % validates the relevance of the developed
model and its applicability within the coolant temperature range of 4oC – 32oC, and bulk oil
temperature range of 48oC – 66oC. In the formulation of optimal wax inhibition mixtures,
addition of demulsifier (Butyl-acrylate) to each of the crude oil samples and inhibitors (mixture
of acrylate ester polymers and xylene) increased the efficiency of the wax inhibitors at low
cooling temperatures (4oC - 16oC). The average wax inhibition (WIav) values of 67% – 82 %
were obtained with a solution of the demulsifier, the inhibitors and the oil compared to values of
46% – 65 % without the demulsifier. The demulsifier term and its interaction with other terms
also had significant influence on the WIav,with p-values of less than 0.05 (P < 0.05) at the
optimum AEP/Xylene/Butyl-acrylate formulation. In the turbulent flow regimes (Re > 3000), the
terms Tc, Tb, Iw, Q,TcTb, TbIw, TbIw, Tc2, Tb2, and Iw2 had significant influence on wax deposition
rates, while the effects of TcQ, TbQ, IwQ and Q2, were insignificant as their P-values were more
than 0.05 (p > 0.5). While in the laminar flow regime (Re < 2000) the results were similar to
those observed in the turbulent flow regime (Re > 3000) except for the flow rate term (Q) which
was insignificant (p > 0.05). This was due to the insignificant effect of shear removal in the
laminar flow region. Experimental verification of the empirical model showed that the wax
deposition was more in the light crude oils grade (APIo > 30), with predicted optimum deposition
rates of 6.2cm3/hr to 9.1cm3/hr in the laminar flow regime, and 6.3cm3/hr to 8.6cm3/hr in the
turbulent flow regime, compared to the wax deposition rates in the medium grade crudes (20 <
APIo < 30) of 5.0cm3/hr to 6.1cm3/hr in the laminar flow regime, and 4.4cm3/hr to 5.8cm3/hr in
the turbulent flow regime. The small error between the predicted and actual laboratory volume of
wax deposit: with percentage average absolute deviation (AAD %) of 2.60 to 3.91% indicated
the validity of the model in predicting the rate and inhibition of wax deposit. The model
predicted value was scaled up using the critical wax tension as a scaler to predict the actual wax
deposition rate in real subsea pipelines. The developed model predicts the field data better than
the Matzain et al., model (2001) implemented in OLGA software currently in use in the
petroleum industry judging by the percentage average absolute deviation (AAD%) of 9.57%
returned by the model as against the AAD% value of 24.46% returned by Matzain et al. model.
Keywords: Empirical model, demulsifier, wax inhibitors, wax deposition, significant influence.
v
TABLE OF CONTENTS
TITLE i
CERTIFICATION ii
DEDICATION iii
ACKNOWLEDGEMENTS iv
ABSTRACT v
LIST OF FIGURES xviii
LIST OF TABLES xxvi
GLOSSARY OF TERMS xxix
NOTATIONS xxxii
SYMBOLS xxxvi
LIST OF ABBREVIATIONS xxxvii
CHAPTER ONE
1.0: INTRODUCTION 1
1.1: Background of the study 2
1.2: Statement of the Problem 3
1.3: Aim and objective of the Study 4
1.4 Scope of the Study 4
1.5: Significance/Justification of the Study 5
vi
1.6: Operational Definition of Terms 6
CHAPTER TWO
2.0: LITERATURE REVIEW 7
2.1: Wax Deposition Mechanisms 7
2.2 Effects of Wax Deposition 8
2.2.1 Technical Effect 8
2.2.2 Economic Effect 9
2.3: Wax Deposition Process 9
2.3.1: Factors Leading to Wax Precipitation and Deposition 11
2.3.1.1: Bulk Oil and Pipe Wall (Coolant) Temperatures 12
differential
2.3.1.2: Flow rate 13
2.3.1.3: Pipe-wall Surface Roughness 14
2.3.1.4: Gas-Oil Ratio 14
2.3.1.5: Crude Oil Composition 14
2.3.1.6: Pressure 16
2.3.2: Wax Crystallization Process 17
2.3.3: Wax Deposition Mechanisms 18
2.3.3.1: Molecular Diffusion Mechanism 19
2.3.3.2: Shear Dispersion Mechanism 19
2.3.4: Production Problems of Wax Deposition 21
2.3.4.1: High Viscosity and Pressure Loss of Crude Oil Flow 22
vii
2.3.4.2: High Yield Stress for Restoration of Flow 22
2.3.4.3: Deposition of Wax Crystals on Surfaces 23
2.3.5: Wax Treatment Techniques 23
2.3.6: Wax Removal Techniques 24
2.3.6.1: Mechanical Removal 24
2.3.6.2: Heat Application 25
2.3.6.3: Application of Wax Removal Chemicals 25
2.3.6.4: Hot Oiling Flow 25
2.3.6.5: Application of Wax Solvent (Inhibitors) 26
2.3.6.6: Application of Thermochemical Packages 27
2.3.7: Wax Preventive Techniques 27
2.3.7.1: Major Methods 27
2.3.7.2: Minor Methods 28
2.3.8: Crude Oil Wax Deposition in Subsea Environment 29
Pipelines
2.3.9: Gulf of Guinea Subsea Environment 31
2.4: Laboratory Studies of Wax Deposition 33
2.4.1: Cold Finger Experiments 34
2.4.2: Flow Loop Experiments 34
2.5: Chemical Wax Prevention 37
2.5.1: Wax Dispersants 38
2.5.2: Wax inhibitors and Pour Point Depressants (PPDs) 39
2.5.2.1: Ethylene Copolymers 40
viii
2.5.2.2: Miscellaneous polymers 41
2.5.2.3: Comb polymers 42
2.5.3: Pour Point Depressants (PPDs) 42
2.5.4: Design of Pour Point Depressants 42
2.5.5: Acrylate Ester polymer 43
2.5.6: Wax Deposit Optimization 43
2.6: Mathematical Modeling of Wax Deposition 45
2.6.1: Rygg, Ryddah and Ronningsen (RRR) Model 47
2.6.2: The Matzain’s Model 49
2.6.3: Hydro Model 53
2.6.4: University of Michigan Model 55
2.7: Wax Deposit Thickness Determination 59
2.8: Evaluation of Wax Inhibitor Performance 59
CHAPTER THREE
3.0: EXPERIMENTAL METHODOLOGY 60
3.1: General Chemicals 61
3.1.1: Chemical Demulsifiers 62
3.1.2: Crude Oil Samples 62
3.2: 62
Experimental Facility (Rig)
3.3: Experimental Section 63
3.4: Analytical Techniques 64
3.5: Pour Point Determination 66
ix
3.6: Wax Content Determination 68
3.7: Wax Deposition Experimental Methods 68
3.8: Sensitivity Analyses of Different Wax Deposition Effluent 69
Variables
3.9: Wax Appearance Temperature (WAT) 69
3.10: Process Optimization 71
3.11: Response Surface Methodology (RSM) 71
3.12: The Central Composite Design (CCD) 73
3.13: Statistical analysis 73
3.14: Determination of Optimum Conditions 74
3.14.1: Optimization of Second-Degree Model 75
3.14.2: Analyzing the Stationary Points 76
3.15: The Test for Significance of Regression 78
3.15.1: Residual plots 79
3.15.2: The F-test 79
3.15.3: The P-value 79
3.15.4: The R2 and Adjusted-R2 79
3.15.5: The sum of square error (SSE) 81
CHAPTER FOUR
4.0: EXPERIMENTAL RESULTS AND DISCUSSION. 82
4.1: Characterization of Crude Oils and Solvents 82
4.2: Wax Deposition Effluent Variables 83
x
4.2.1: Effect of Inlet Coolant Temperature 84
4.2.2: Effect of Bulk Oil Temperature 85
4.2.2.1: Case 1: The Bulk Oil Temperature (Above its WAT) 86
4.2.2.2: Case 2: The Bulk Oil Temperature (Below its WAT) 87
4.2.3: Effect of Flow Rate 88
4.2.3.1: The Flow Rate in Laminar Flow Regime 89
4.2.3.2: The Flow Rate in Turbulent Flow Regime 90
4.3: Comparison of Shear dispersion and Molecular Diffusion 91
Contribution to Wax Deposition
4.4: Sensitivity Analyses of Shear Dispersion Influent Factors 92
to Wax Deposition
4.4.1: Effect of Bulk Oil and Wall Temperature differential 92
(below the WAT)
4.4.2. Effect of Flow Rate 93
4.4.3 Effect of Wax Content 94
4.5: Variation of Wax Content with Time. 95
4.6 Variation of Outlet Temperature of Test Section with Time 96
4.7: Bulk Crude Oil and Wax deposit composition analyses 97
4.8: Analysis of Pour Point Experimental Results 99
4.8.1: Effect of Petroleum fractions as pour point depressants 100
4.8.2: Effect of Acrylate Ester Polymers as Pour Point 102
Depressants
4.8.3: Verification of Different Acrylate Ester Co-polymers, AEP 103
xi
(solvent) in Xylene Solution as Wax Inhibitor.
4.9: Wax Inhibition Evaluation with Fixed Quantity of (PPDs) 105
at Varying Ambient Conditions.
4.10: Evaluation of Wax Inhibition Performance 110
4.11: Demulsifier Screening Process 113
4.12: Effect of Demulsifier on Overall Wax Inhibition 115
4.13: Effects of Demulsifier on Crude Oil Wax Deposits 117
4.13.1: Crude Oil Mixed with Pour Point Depressant 117
4.13.2: Crude Oil Mixed with Pour Point Depressant and a 118
Demulsifier.
4.14: Modeling of Wax Inhibitor (AEP/xylene) Mixture 118
Formulation
4.15: Interaction plot of Overall Wax Inhibition (WIav) for 122
AEP/Xylene Formulation
4.16: Response Surface Graphs for Acrylate Ester 124
Copolymer/Xylene Interactions
4.17: Residual Plot of Overall Wax Inhibition (W.Iav) in 125
AEP/Xylene Formulation
4.18: Optimization of AEP/Xylene Formulations 126
4.19: Analysis of the stationary point of Acrylate Ester 127
Polymer/Xylene Mixture Formulation
4.20: Modeling of AEP/Xylene/Butyl acrylate (Demulsifier) 129
Mixture Formulation
xii
4.21: Interaction Plot of Overall wax inhibition (WIav) for 134
AEP/Xylene/Butyl-acrylate Formulation.
4.22: Response Surface Graphs for Acrylate Ester 136
Copolymer/Xylene/Butyl-acrylate Interactions
4.23: Residual Plot of Overall Wax Inhibition (W.Iav) in 141
AEP/Xylene/Butyl-acrylate Formulation
4.24: Analysis of the stationary point of Acrylate Ester 142
Polymer/Xylene/Butyl Acrylate Mixture Formulation
4.25: Comparison of Optimum AEP/Xylene to Optimum 144
AEP/Xylene/Butyl acrylate Formulation
4.26: Modeling of Wax Deposition Process for Each Crude Oil 144
Sample.
4.26.1: Modeling of Wax deposition in the Laminar Flow Regime 145
(Re < 2000)
4.26.2: Modeling of Wax deposition in the Turbulent Flow 148
Regime (Re > 3000)
4.27: Analyses of Variance (ANOVA) and Model Fitting (Re < 149
2000)
4.28: Interaction Plot of Wax Deposition Process (Laminar 151
Flow) for Crude-Oil A
4.29: Response Surface Plots (Laminar Flow Regime) 153
4.30: Residual Plot of Wax Deposition for Crude Oil A in 163
Laminar Flow Regime
xiii
4.31: Models optimization (Laminar flow) 164
4.32: Analyses of Variance (ANOVA) and Model fitting 168
(Turbulent Flow Regime (Re > 3000))
4.33: Interaction Plot of Wax Deposition Process (Turbulent 171
Flow) for Crude-Oil A
4.34: Response Surface Plots (Turbulent flow regime) 172
4.35: Residual Plot of Wax Deposition for Crude Oil A in 175
Turbulent Flow Regime
4.36: Models Optimization (Turbulent Flow Regime) 176
4.37: Modeling of General Wax Deposition Process in Laminar 180
Flow Regime (Five Variables)
4.38: Analyzing the Stationary Point of Wax Deposition Process 182
in Laminar Flow (Five Variables)
4.39: Variables Interaction Analyses of Wax Deposition Process 185
4.40: Modeling of Wax Deposition Process in Turbulent Flow 189
Regime (Five Variables)
4.41: Response Surface Analyses of Five Variables Wax 191
Deposition (Turbulent Flow Regime)
4.42: Analyzing the Stationary Point of Wax Deposition Process 198
in Laminar Flow (Five Variables)
4.43: Model Validation 201
4.43.1: Experimental versus Predicted Values 201
4.43.2: The Residual Plots for General Wax Deposition Model 204
xiv
4.44: Scale-Up of Wax deposition Model 204
4.45: Modeling attempts with critical wax tension 205
4.46 Validation of Developed Model with Field Data and 209
Comparison with Existing Model in Use
CHAPTER FIVE
5.0: CONCLUSIONS AND RECOMMENDATIONS 212
5.1: Conclusions 212
5.2: Summary of Findings 213
5.3: Contribution to Knowledge 215
5.4: Publications from the Study 216
5.5 Future Studies 217
5.5.1 Multiphase Wax Deposition Experiments and Modeling 217
5.5.2 Impact of Surface Roughness on Wax Deposition 217
5.5.3 Single Phase Wax Deposition under Low Driving 218
Force/Heat Flux Conditions
REFERENCES 219
APPENDIX 1: CONTOUR PLOTS OF OVERALL WAX INHIBITION, 229
W.Iav (%) FOR CRUDE OILS B – G (AEP/XYLENE
FORMULATION)
APPENDIX 2: CONTOUR PLOTS OF OVERALL WAX INHIBITION, 230
xv
W.Iav (%) FOR CRUDE OILS B – G
(AEP/XYLENE/BUTYL-ACRYLATE FORMULATION)
APPENDIX 3: RESIDUAL PLOTS OF OVERALL WAX INHIBITION, 233
W.Iav (%) FOR CRUDE-OIL B – G (AEP/XYLENE
FORMULATION)
APPENDIX 4: RESIDUAL PLOTS OF OVERALL WAX INHIBITION 235
W.Iav FOR CRUDE-OIL B – G (AEP/XYLENE/BUTYL-
ACRYLATE FORMULATION)
APPENDIX 5: CONTOUR PLOTS OF WAX DEPOSITION FOR 237
CRUDE OIL B –G (4-VARIABLES)
APPENDIX 6: RESIDUAL PLOTS OF WAX DEPOSITION FOR 243
CRUDE OIL B – G (4 VARIABLES, LAMINAR FLOW)
APPENDIX 7: RESIDUAL PLOTS OF WAX DEPOSITION FOR 245
CRUDE OIL B – G (4 VARIABLES, TURBULENT
FLOW)
APPENDIX 8: ANALYSIS OF VARIANCE (ANOVA) RESULTS OF 247
OVERALL WAX INHIBITION FOR CRUDE OIL A-G
AEP/XYLENE FORMULATION
APPENDIX 9: ANALYSIS OF VARIANCE (ANOVA) RESULTS OF 253
OVERALL WAX INHIBITION FOR CRUDE OIL A-G
AEP/XYLENE/BUTYL-ACRYLATE FORMULATION
APPENDIX 10: ANALYSIS OF VARIANCE RESULTS (ANOVA) OF 264
WAX DEPOSITION FOR CRUDE OIL A-G (4
xvi
VARIABLES, LAMINAR FLOW REGIME)
WAX DEPOSITION FOR CRUDE OIL A-G (4
VARIABLES, TURBULENT FLOW REGIME)
WAX DEPOSITION FOR CRUDE OILS FLOW IN
SUBSEA STRAIGHT PIPELINE (FIVE (5)
VARIABLES, LAMINAR FLOW REGIME)
WAX DEPOSITION FOR CRUDE OILS FLOW IN SUB
SEA STRAIGHT PIPELINE (FIVE (5) VARIABLES,
TURBULENT FLOW REGIME)
xvii
LIST OF FIGURES
Figure 1: Wax deposit in a retrieved pipeline 10
Figure 2a: Shear dispersions mechanism during wax deposition. 20
Figure 2b: Schematic diagram of particles movement in shear dispersion process 21
Figure 3: Typical Subsea Pipelines 30
Figure 4: Schematic diagram of crude oil flow in subsea environment 31
Figure 5: The Map of Gulf of Guinea 32
Figure 6: Temperature Variation in Typical Ultra Deep Location in the Gulf of 33
Guinea
Figure 7: The structural formulae of alkyl ester 44
Figure 8: Laboratory wax deposition test flow loop rig. 65
Figure 9: Schematic diagram of wax deposition test flow loop 66
Figure 10: The flow chart of the procedure for the research study 70
Figure 11: Viscosity profile of the oil sample at different shear rate and 82
temperature
Figure 12: Wax deposit versus time at different wall temperatures for oil A 85
(flow rate of 1.4ltr/min., Temp. 55oC).
Figure 13: Wax deposit versus time at different inlet oil temperatures above 86
WAT at inlet coolant temperature of 37oC and oil flow rate of 1.4
liter/min.
Figure 14: Dimensionless thickness of wax deposit versus time at inlet coolant 88
xviii
temperature of 28oC (below WAT) at oil flow rate of 1.4 liter/min.
Figure 15: Effect of Flow Rate (Laminar/Laminar-turbulent flow region) on 89
Dimensionless Wax Thickness.
Figure 16: Effect of Flow Rate (Turbulent flow region) on Dimensionless Wax 90
Thickness.
Figure 17: Plot of Dimensionless wax thickness due shear dispersion and 92
molecular diffusion effect against time.
Figure 18: 93
Plot of Dimensionless wax thickness due shear dispersion effect
against time at different temperatures
Figure 19: 94
Plot of Dimensionless wax thickness due shear dispersion effect
against time at different flow rates
Figure 20: Plot of Dimensionless wax thickness due shear dispersion effect 95
against time at different wax contents.
Figure 21: The change in density of the deposited wax deposit with time at oil 96
flow rate of 1.00 liters/min
Figure 22: Outlet temperatures of the bulk oil and the water/wall temperature at 97
different time.
Figure 23: Mass percentage change of n-alkanes in crude-oil A and its wax 98
deposit
Figure 24: Mass percentage change of n-alkanes in crude-oil D and its wax 98
deposit
Figure 25: Mass percentage change of n-alkanes in crude-oil E and its wax 99
xix
deposit
Figure 26: Pour point depressions of Kerosene on oil samples. 100
Figure 27: Pour point depressions of PMS (Petrol) on oil samples. 101
Figure 28: Effect of petroleum fractions (Kerosene) on Wax deposition 101
Figure 29: Pour point depressions on oil samples by different percentage of 102
polymers.
Figure 30: Effect of Acrylate Ester Polymer on Wax deposition 103
Figure 31: Effect of different quantities of solvent P1 on wax deposition 106
Figure 33: Further effect of different quantities of solvent P2 on wax deposition 107
Figure 35: Effect of different coolant temperature on wax deposition 109
Figure 36: Further effect of different coolant temperatures on wax deposition 109
Figure 37: Wax deposition profiles at different injection time using sample A 110
Figure 38: Cumulative wax deposition rate with and without the P1 inhibitors 111
using sample A.
Figure 39: Cumulative wax deposition rate with and without the P2 inhibitors 112
using sample A.
Figure 40: Cumulative wax deposition rate with and without the P2 inhibitors on 112
sample B.
Figure 41: Cumulative wax deposition rate with and without the polymer, P3 on 113
sample C.
Figure 42: Effect of different demulsifiers on wax deposit using crude-oil 114
xx
sample A
sample D
sample E
Figure 45: Effect of demulsifier on inhibition of polymer P1 on sample A 116
Figure 46: Effect of demulsifier on inhibition percentage of polymer, P2 on oil 117
sample B
Figure 47: Interaction plot for overall wax inhibition percentage (%) for crude 123
oil A (AEP/xylene)
Figure 48: Surface plots of WIav for varying AEP and xylene for Oil A in 124
AEP/Xylene formulation
Figure 49: Contour plots of WIav for varying AEP and xylene for Oil A in 125
AEP/Xylene formulation.
Figure 50: Residual plots of WIav for Oil A AEP/Xylene formulation. 126
Figure 51: Interaction plot for overall wax inhibition percentage W.I.av (%) for 135
crude oil A (AEP/xylene/butyl-acrylate)
Figure 52: Surface plot of WIav at varying AEP and butyl-acrylate for Oil A in 137
AEP/Xylene/butyl acrylate formulation
Figure 53: Contour plots of WIav for varying AEP and butyl-acrylate for Oil A 137
in AEP/Xylene/butyl acrylate formulation
Figure 54: Surface plots of WIav for varying xylene and butyl acrylate for Oil A 138
; in AEP/Xylene/butyl acrylate formulation
xxi
Figure 55: Contour plots of WIav for varying xylene and butyl acrylate for Oil A 139
in AEP/Xylene/butyl acrylate formulation
Figure 56: Surface plots of WIav for varying AEP and xylene for Oil A in 140
Figure 57: Contour plots of WIav for varying AEP and xylene for Oil A in 140
Figure 58: Residual plots of WIav for Oil A AEP/Xylene formulation 141
Figure 59: Optimal overall wax inhibition for AEP/Xylene and 144
AEP/Xylene/Butyl acrylate.
Figure 60: Interaction Plot for wax deposition process (cm3) 151
Figure 61: Surface Plot for Oil A (Coolant Temp.Vs. Oil temp.) in laminar flow 154
regime
Figure 62: Contour Plot for Oil A (Coolant Temp. vs Oil temp.) during laminar 155
flow regime
Figure 63: Surface Plot for oil A (Coolant temp. Vs wax inhib. Percent) in 156
laminar flow regime
Figure 64: Contour plots for oil A (Coolant temp. vs wax inhib. Percent) in 156
laminar flow regime
Figure 65: Surface Plot for oil A (Coolant Temp. Vs. Flow Rate) in Laminar 157
flow regime
Figure 66: Contour Plot for oil A (Coolant Temp. Vs Flow Rate) in Laminar 158
flow regime
Figure 67: Surface Plot for Oil A (Oil Temp. Vs. Wax Inhibitor Percent) 159
xxii
inlaminar flow regime
Figure 68: Contour Plot for Oil A (Oil Temp. Vs. Wax Inhibitor Percent) during 159
laminar flow regime
Figure 69: Surface Plot for Oil A (Oil Temp. Vs. Flow rate) in Laminar flow 160
regime
Figure 70: Contour Plot for Oil A (Oil Temp. Vs. Flow rate) during Laminar 161
flow regime
Figure 71: Surface Plot for Oil A (Wax Inhibitor Percentage. Vs. Flow Rate) in 162
Laminar flow regime
Figure 72: Contour Plot for Oil A (Wax Inhibitor Percentage Vs. Flow Rate) in 162
Laminar flow regime
Figure 73: Residual plots of Wax Deposition for Oil A in Laminar flow regime 163
Figure 74: Interaction plot of wax deposition process (cm3) (Four variables) 171
Figure 75: Contour plot for Oil A (Coolant temperature Vs. Bulk oil 173
temperature) in turbulent flow regime
Figure 76: Contour plot for Oil A (Coolant Temp. Vs. Wax inhibitor 173
percentage.) in turbulent flow regime
Figure 77: Contour plots for Oil A (Bulk oil Temperature Vs. Wax inhibitor 173
Percentage) in turbulent flow regime
Figure 78: Contour plot for Oil A. (Flow rate Vs. Wax inhibitor Percentage.) in 174
turbulent flow regime
Figure 79: Contour plot for Oil A. (Flow rate Vs. Coolant temp.) in Turbulent 174
flow regime
xxiii
Figure 80: Contour plot for Oil A. (Flow rate Vs. Bulk oil temp.) in Turbulent 175
flow regime
Figure 81: Residual plots of Wax Deposition for Oil A in Turbulent flow regime 176
Figure 82: Interaction plot for wax deposition process (cm3) (Five Variables) 186
Figure 83: Contour and Surface Plots of Wax Deposit Vs. Coolant, Bulk Oil 192
Temperature.
Figure 84: Contour and Surface Plots of Wax Deposit Vs. Coolant Temperature, 193
Wax Inhibitor Percentage.
Flow Rate.
Wax Content.
Figure 87: Contour and Surface Plots of Wax Deposit Vs. Bulk Oil 195
Temperature, Wax Inhibitor Percentage.
Temperature, Flow Rate.
Temperature, Wax Content.
Figure 90: Contour and Surface Plots of Wax Deposit Vs. Wax Inhibitor 196
Percentage, Flow Rate.
Figure 91: Contour and Surface Plots of Wax Deposit Vs. Wax Inhibitor 197
Percentage, Wax Content.
Figure 92: Contour and Surface Plots of Wax Deposit Vs. Flow Rate, Wax 297
xxiv
Content.
Figure 93: Model validation within studied range. 201
Figure 94: Model validation outside studied range 202
Figure 95: Residual plots of general wax deposition model 204
Figure 96 Comparison of total wax deposit against time of different models 211
with field data
xxv
LIST OF TABLES
Table 1: Highlights of considered mechanisms by different models. 46
Table 2: Analyses of different models advantages and dis-advantages 57
Table 3: Values of wax deposition effluent factors used in sensitivity 69
analyses
Table 4: Crude Oil Characteristics 83
Table 5: Kerosene and PMS Properties 83
Table 6: Experimental Wax Deposition flow loop results. 104
Table 7: Experimental flow loop results for sample A at varying ambient 105
temperatures.
Table 8: Levels of factors selected for the central composite design for 119
AEP/xylene mixture.
Table 9: Design matrix of the central composite design and corresponding 120
responses using AEP/xylene mixture
Table 10: P-values for Crude oil A-G in AEP/Xylene mixture formulations 121
Table 11: Individual term coefficients for Crude oil A-G in coded units in 121
AEP/xylene mixture formulations
Table 12: Individual term coefficients for Crude oil A-G in AEP/xylene 122
mixture Formulations
Table 13: The PPD Formulation Optimization results using copolymer 129
(AEP)/Xylene
AEP/xylene/butyl acrylate mixture.
xxvi
responses using AEP/xylene/butyl acrylate mixture
Table 16: Individual term coefficients for Crude oil A-G in 133
AEP/xylene/butyl acrylate mixture formulations in un-coded
units.
Table 17: P-values for Crude oil A-G in AEP/xylene/butyl acrylate mixture 133
formulations
Table 18: The Optimization results of PPD Formulation using 143
copolymer/xylene/Butyl acrylate
wax deposition inhibition.
responses of wax deposits inhibition process (Laminar flow, Re <
2000).
responses of wax deposits inhibition process (Turbulent, Re >
3000).
Table 22: The Models equations coefficients for oil A-G in un-coded units 150
for laminar flow regime.
Table 23: The regression equations (Models) coefficients for oil A-G 164
(Laminar flow).
Table 24: Optimum wax inhibition variables for Oil A-G in the laminar 167
flow regime.
xxvii
Table 25: Regression equations (Models) coefficients for Oil A – G. 168
(turbulent transition flow), in coded units.
Table 26: Regression equations (Models) coefficients for Oil A – G. 169
(turbulent flow), in un-coded units.
Table 27: P-values in the turbulent flow regime for Oil A – G. 170
Table 28: Optimum wax inhibition variables for Oil A-G in turbulent flow 179
regime.
Table 29: Range of minimum wax deposit (cm3) at different flow regimes 179
for different Crude oils
Table 30: Optimum wax inhibition variables for Oil A-G in laminar flow 185
regime. (5 variables)
Table 31: Wax Deposit at different Variables Interaction (Five Variables) 192
Table 32: Optimum wax inhibition variables for Oil A-G in turbulent flow 200
regime. (5 variables)
Table 33: Operating parameters within the studied range used for model 202
validation
Table 34: Operating parameters outside the studied range used for model 203
validation
Table 35: Wax deposition conditions for a Niger Delta pipelines 209
Table 36: Comparison of the developed model and Matzain et al (2001) 210
model with field data
Table 37: Summary of findings of the study 213
xxviii
GLOSSARY OF TERMS
Aromatics: Hydrocarbon characterized by unsaturated ring structures of
carbon atom e.g.: acrylate ester polymer.
Asphaltenes: These are the polar components of crude oil.
Brownian diffusion: Deposition of continuously bombarded wax particles due to
resulting net transport.
Bulk oil temperature: This is the temperature of the transported oil in the flow loop
experiment.
Coolant Temperature: This is the temperature of the cooling water used in the flow
loop experiment.
Cold flow: Crude oil flow in relatively lower temperature (below
25oC) pipeline.
Critical wax tension: The force required to break a unit thickness of wax deposit in
radial direction.
Deep sea: A sea of relatively higher depth
Demulsifier: Chemicals added to neutralize the effect of emulsifying agent
and de- emulsified the emulsion. E.g. butyl acrylate.
Gelation: The formation of jelly like liquid on the pipeline during crude
oil flow as a result of cooler pipe wall.
Gravity settling: Deposition of precipitated wax crystal due to gravitational
force.
Heavy components: These are C30+ components of crude oil.
Hot flow: Crude oil flow in relatively higher temperature(Above 25oC)
xxix
range.
Inhibitors: A solvent added to crude oil in order to reduce or completely
eliminate the amount of wax deposition during the oil flow in
pipeline e.g.: Acrylate ester copolymer dissolved in xylene.
Mathematical model: Mathematical expression relating a dependent variable to
set of independent variables.
Molecular diffusion: Net flow of dissolved wax towards the pipe wall due to
temperature gradient in radial direction
Naphthenes: These are cyclo-paraffins contained in the crude oil
Optimization: Scientific method of selecting independent variable
combination that gives optimal objective function from series
of possible variable combinations.
Overall Average Wax inhibition: Quantitative measure of the wax deposit inhibition within the
studied temperature range (%).
Phase: A state of matter
Pigging: Passage of circular disc (pig) through the pipeline in order to
removes the sludge inside the pipeline.
Plugging: The stacking of the pig in the pipeline during pigging.
Pour point: Temperature at which the crude oil sample no longer flow.
Pour point depressant (PPD): A chemical that reduce the pour point temperature e.g.
acrylate ester polymer.
Response surface modeling: Scientific technique of regression analysis using response
surface graphs.
xxx
Response surface Methodology: Method of using response surface graph to model the effect of
(RSM) different independent variables on the objective functions.
Shear dispersion: Net flow of wax crystals towards the pipe wall due to velocity
radient in the radial direction.
Shear removal: Erosional removal of deposited wax due to shear stress.
Solubility limit: The condition at which the wax particle will no longer be
soluble in the Crude oil.
Subsea oil fields: Off-shore oil fields
Temperature differential: The difference in temperature between the bulk oil and the
pipe all temperature.
Wax: A ”shoe polish” like solid that get deposited in pipeline
during flow of warm crude through a relatively colder
pipeline.
Wax appearance temperature: The temperature at which wax particles begin to precipitate in
(WAT) crude oil mixture.
Wax Deposit: The wax volume that clings to the inner part of the pipeline
during the flow of waxy crude in the pipeline
Wax inhibition: A process that prevents wax deposit in pipeline.
Wax thickness: Thickness of wax deposit in the pipeline (dimensionless).
Xylene: Aromatic solvent containing carbon ring with delocalized
electron double bonding.
xxxi
NOTATIONS
A = The surface area available for deposition (cm2) = πDL
b = The volume of the xylene (centiliters)
B = symmetric matrix of order k x k whose ith diagonal element is  ii (i  1,2,.................k ) and its
(i,j)th off diagonal element.
c = The volume of butyl acrylate (centiliters) in the wax inhibitor,
c ib , ciw = The model concentration of the wax component i dissolved in the oil phase in the bulk
oil and at the wall respectively (mole/m3).
C = composition
Cp = Crude oil heat capacity
Cwall = The volume fraction of the precipitated wax in the oil at the inner wall temperature.
C1, C2, and C3 = The three empirical constants which are correlated from the single phase and
two phase flow data. (Their values were found to be: C1 = 15.0, C2 = 0.055 and C3 = 1.4).
d = The hydraulic diameter or effective inside diameter,
dw = The inside diameter as a result of wax build-up (m)
D = The diffusion coefficient.
D’ = internal diameter of the full scale pipeline.
Dow = Turbulent mass diffusion.
xxxii
dC/dr = radial concentration gradient of the precipitated wax
dC/dT = concentration gradient of the precipitated wax with respect to temperature
dT/dr = radial temperature gradient of the precipitated wax.
E = The liquid hold up ,
f2 = The stress on the same wax filament on the deposit surface,
hwall = The inner wall surface heat transfer coefficient (W/(m2K)).
Iw = The percentage of wax inhibitor,
ko = Crude oil thermal conductivity
k* = Shear dispersion rate constant (kg/m2)
L = The length of pipe/tube section (m),
MWi = The molar weight of wax component i (kg/mole),
NWAX = The number of wax components,
Q = flow rate
r = The radial distance (m)
R = The radius of the clean pipe
S = sample mass, g;
Swet = The fraction of the wetted circumference.
xxxiii
SSE = Sum of square of error
SSR = sum of square of regression model
SST = Total sum of square
t = time
T = the temperature (oC),
Tb = The bulk average flow temperature (K),
Tw = the inner wall surface temperature (K)
v = flow velocity
vsl = The liquid superficial velocity (m/s)
V = The predicted response (volume of deposited wax)
WI =The group of AEP/xylene/butyl acrylate mixture in the crude oil sample,
Wb= Wax deposit by volume of the blank oil (cm3) and
Wwi = Wax deposit by volume during the same period of time of inhibitor/chemically treated oil
(cm3).
(WI)Ti = Percentage Wax Inhibition (%) at temperature Ti
W.I. av = overall average wax inhibition percentage, ais the volume of group of acrylate ester
polymers (centiliters) (W.I) mixture,
ww = The concentration of wax in solution (weight %),
xxxiv
x = The wax content of the deposit,
xxxv
SYMBOLS
 = vector of m unknown constant coefficient referred to as parameters,
 = The shear rate at the wall (s-1).
 = The thickness of wax layer deposited on the wall (m).
 = a random experimental error assumed to have a zero mean
o = The oil thermal conductivity (W/(mK)).
 = The apparent viscosity of the crude oil.
μo = The oil viscosity (kg/m.s)
ρ = The density of wax component i (kg/m3).
ρo = The oil density (kg/m3).
ρm = the average density of the gas-oil mixture (kg/m3).
 wax = The average wax density (kg/m3).
ϕ = The porosity of the deposited wax.
 1 = the empirical relation for the rate enhancement due to oil being trapped in the deposited
wax layer. The relation also accounts for any positive deposition rate which is not accounted for
by the diffusion constant,
 2 = The empirical relation for the rate reduction due to shear stripping.
xxxvi
 = wax deposition tendency, and it’s a function of oil temperature, oil composition and/or
shear rate
xxxvii
LIST OF ABBREVIATIONS
AAD: Average Absolute Deviations
AEP: Acrylate Ester Polymers
CCD: Central Composite Design
CCN: Critical Carbon Number
JIP: Joint Industry Programme
RRR: Rygg, Ryddah and Ronningsten
RSM: Response Surface Methodology
WAT: Wax Appearance Temperature
WI: Wax Inhibitors
xxxviii
CHAPTER ONE
1.0: INTRODUCTION
Most major crude oil fields world-wide have reached the mature stage, and oil production has
been on decline. As a result, oil companies are exploring oil fields situated in deep and ultra deep
water. These together with local challenges of oil communities’ unrest as a result of oil pollution
on the farmlands make explorations of subsea fields attractive to Nigeria’s oil companies.
Forecasters expect that, by 2030, oil production from deep sea area will exceed 48 million
barrels per day which is about three times greater than deep sea production in year 2012 (12.4
million barrels per day) (Moritis, 2002; Singh et al., 2004).
There are many of such fields in Nigeria,located in the subsea environment of the Gulf of
Guinea, which play host to major crude oil and natural gas reserves in sub-Sahara Africa,
accounting for 35% of the continent’s oil production (Ajienka and Ikoku, 1997; Adewusi, 1998)
The problem of wax deposition has plagued the petroleum industry for decades. The earlier the
problem is diagnosed in the life of a reservoir, well or pipeline, the easier it will be to design a
preventive or control management plan that will reduce or eliminate some of the technical
problems associated with wax deposition (Adewusi, 1998).
During pigging, pig sometimes gets stuck inside the pipeline in the presence of thick wax deposit
and one of such replacements cost forty million U.S. dollars ($40,000,000) (Fung et al., 2006).
Wax deposition also leads to abandonment of projects, one of such was by Shell Oil Company as
a field in the North sea with an investment of one hundred million U.S. dollars was abandoned
(Singh et al., 2000: Nguyen et al., 2001). The earlier the problem is diagnosed in the life of a
reservoir, well or pipeline, the easier it will be to design a preventive or control management plan
1
that will reduce or eliminate some of the technical and economic problems associated with wax
deposition.
Other depositions in pipelines include:
a. Asphaltene: These are heavy hydrocarbon molecules in petroleum reservoir fluids that
precipitate and are deposited in pipelines during pressure depletion or during gas
injection processes.
b. Mineral Scales: These are insoluble salts deposited during water-flooding as a result of
chemical reaction between the formation water and injected water.
c. Gas Hydrates: These are ice-crystalline compound that are deposited in gas pipelines as a
result of water molecules in the flowing gas through the pipeline.
1.1: Background of the study
Wax deposition is one of the critical challenges facing the oil and gas field operators as the
industry explores increasingly challenging environments, such as deep water especially during
the cold season. Wax deposition during crude oil transportation is problematic to the industry
due to reduction in flow area as more wax continues to get deposited (Zhang et al., 2011).
According to Lee, (2008) crude oil is a complex mixture of saturates (paraffins/waxes),
aromatics, naphthenes, asphaltenes and resins. The heavy paraffinic hydrocarbon(waxes) and
asphaltene components are responsible for solid deposition during crude oil flow in subsea
pipeline systems. At reservoir conditions (temperatures of 70 – 150 oC and pressures of 50 – 100
MPa) these heavy components are dissolved in the crude oil (Adeyanju and Oyekunle, 2011).
However, as the crude oil flows from the production platform to the export terminal through
subsea pipelines at temperatureof 20 -29oC in shallow sea, at sub-sea temperature that can be as
2
low as 4oC in deep sea, the wax compounds precipitated from the oil mixture and are deposited
in the pipeline. In most of these scenarios, the crude oil temperature falls below its Wax
Appearance Temperature (WAT) or the cloud point temperature due to heat loss to the
surroundings.
1.2: Statement of the Problem
Exploration and production technologies in deep sea areas have made drilling in deep water more
attractive (Leiroz and Azevedo, 2007; Lee, 2008). As more and deeper sea reserves are
exploited, developed and transported, wax problems are expected to become more severe and
extensive due to installations of more transportation lines on the cold ocean floor. Hence, there is
obvious need for continuous research efforts to reduce the wax deposition rates in these pipelines
to the barest minimum.The wax deposition experimental study in this work will provide the data
needed for the development of realistic models for scheduling the removal of the deposited gel.
Challenges associated with wax deposition include:
i. Increase in capital investment and operating cost.
ii. Permeability reduction and formation damage when it occurs around the wellbore and
its vicinity
iii. Reduction in the interior diameter and eventual plugging of production pipelines and
flow channels
iv. Changes in the reservoir fluid composition and fluid rheology due to phase separation
as wax solid precipitates.
v. Additional strain on pumping equipment owing to increased pressure drop along flow
channel consequent to rheological changes as wax begins to crystalize.
3
vi. Limiting influence on the operating capacity of the entire production system.
vii. Loss in production.
1.3: Aim and objectives of the Study
The aim of this research is to optimize the operating parameters for wax deposition process in
crude oil flow in straight sub-sea pipelines. Thus, this research work has the following
objectives:
i.. To study the effects of temperature, flow rate, wax content, wax inhibitors (petroleum
fractions and polymeric compounds) and de-emulsifier on wax deposition rates.
ii. To develop empirical models relating the wax deposition rates to the above variables for the
effective scheduling of wax deposit removal.
iii. To determine the operating parameters for minimum wax depositionrate in straight subsea
flow-lines.
iv. Employ the model to field data (validate) and compare the model with existing ones currently
in use.
1.4: Scope of the Study
This study is concerned with the modeling and optimization of the operating parameters that
influence the paraffin wax deposition rates in subsea pipeline using response surface
methodology. The study considered the flow of single phase crude oil from production
platform/Floating production storage and overloading (FPSO) vessel to the export terminals
through subsea horizontal pipelines. Hence, the study was limited to the following:
i. The used of single phase crude oil (dead crude).
4
ii. Seven crude oils of different API gravity from different subsea reservoir in the Gulf
of Guinea were used.
iii. The crude oil flow is in horizontal pipeline.
iv. The operating parameters considered are within the following range:
Coolant temperature (277 – 297 K),
Bulk oil temperature (321 – 329 K)
Wax inhibitors percentage (1 – 5 %)
Flow rate (Laminar flow: 0.3 – 0.7 liter/min, turbulent flow: 1.0 -1.4 liter/min)
Wax content (28 – 40 %)
v. Polymeric compounds (acrylate ester co-polymer of different alkyl length) dissolved
in xylene were used as pour point depressant/wax inhibitors.
vi. The effect of petroleum products (kerosene and premium motor spirit) on the crude
oil pour points were studied.
vii. The evaluation of the de-emulsifiers were restricted to acrylic acid, butyl acrylate and
methyl mathacrylate due to their solubility in xylene.
viii. Response surface methodology central composite design was used to design the
experiments, develop the models, analyze data and optimize the operating parameters.
1.5: Significance/Justification of the Study
This research work has the following significance:
1. The developed empirical models can be used for the prediction of wax deposition rate in
crude oil flow in subsea pipelines and also for the scheduling of wax deposit removal.
5
2. The developed wax inhibitor evaluation expression can be used to evaluate the
effectiveness of different wax inhibitors at field operating conditions.
3. The study will also strengthen the understanding of the wax deposition and inhibition
processes in cooled pipelines and also the surface response method of experimental
design, models formulation and results analyses.
1.6: Operational Definition of Terms
De-emulsifier Chemicals added to neutralize the effect of emulsifying agent
and de- emulsified the emulsion. E.g. butyl acrylate.
Empirical Model Model based on Experimental Data
Flow Rate The volume of crude oil in liters that flows through a point in
the pipeline in 1 second
Petroleum Fractions These are Refineries products of Crude Oil
Polymeric Compounds These are Polymers based Compounds
Significance Influence These are variables that have significance influence on the
response due to their returned P-value of less than 0.05 (based
on 95% confidence interval)
Temperature Differential The difference between the coolant and bulk oil temperature
Wax Content The percentage volume of wax particle in the crude oil sample
Wax Deposition The deposition of wax deposit at the pipeline’s wall during
crude oil flow.
Wax Deposition Rate The volume of deposited wax per unit time in cm3/sec
Wax Inhibitors The chemicals/solvents that prevent the deposition of wax
6
CHAPTER TWO
2.0: LITERATURE REVIEW
2.1:Wax Deposition Mechanisms
From the literature the general consensus is that the major wax deposition mechanisms in sub-
cooled pipelines are: molecular diffusion, shear dispersion, shear stripping reduction, Brownian
diffusion, and gravity settling (Bern et al., 1980; Burger et al. 1981; Majeed et al., 1990).
Mechanism such as shear dispersion, Brownian diffusion, and gravity settling had been identified
to play a role only for particulate deposition of wax. However, all models studied in the literature
neglected the effect of particulate deposition as this effect is considered inconsequential. Hence,
molecular diffusion and shear stripping are currently considered as the predominant mechanisms
underlying the wax deposition process. A series of experimental studies had been carried out on
the mechanisms affecting wax deposition: Vankatesan and Creek, (2007) articulated the
difference between the laboratory conditions and those prevailing in the fields and recommended
that the then existing laboratory methods should be reviewed.
Risk associated with the transportation of waxy crudes is one of the most critical operational
hazardof deepwater offshore pipelines. The challenge that engineers will face in offshore
operation is thus on how to design the pipeline and subsea system to assure that multiphase waxy
crudes will be safely and economically transported from the bottom of the wells through
deepwater pipelines all the way to the downstream processing plant. The practice of identifying,
quantifying and mitigating all the flow risks associated with offshore pipelines and subsea
systems is called flow assurance (Guo et al., 2005). Flow assurance is tedious for deepwater
pipelines and system operations. In deepwater, the seawater temperature is usually much colder
7
than the fluid temperature inside the pipeline (Al-Yaari, 2011). If the fluid temperature inside the
pipelines becomes too low due to heat loss, wax deposit will form on the pipe wall. Thus,
effective protection of fluid heat is one of the most important design parameters for offshore
pipeline (Guo et al., 2005). Hence the effect of wax deposition and the factors affecting wax
deposition mechanisms need to be extensively studied.
2.2: Effect of Wax Deposition
The problem of wax deposition has plagued the petroleum industry for decades, arousing two
main concerns: technical and economical upon its occurrence.
2.2.1:Technical Effect
Technical issues associated with wax deposition include:
viii. Permeability reduction and formation damage when it occurs around the wellbore and
its vicinity
ix. Reduction in the interior diameter and eventual plugging of production pipelines and
flow channels
x. Changes in the reservoir fluid composition and fluid rheology due to phase separation
as wax solid precipitates.
xi. Additional strain on pumping equipment owing to increased pressure drop along flow
channel consequent to rheological changes as wax begins to crystalize.
xii. Limiting influence on the operating capacity of the entire production system.
8
2.2.2: Economic Effect
The critical role of economies in crude oil production makes wax deposition a significant
economic concern to the industry due to the following.
i. Capital investment and operating costs are increased when developing paraffinic crude
oil field. This could cause serious financial strain on the operator of such field or even
lead to abandonment when it becomes uneconomical due to blockage of facilities by wax
deposit.
ii. Lost production to system shutdown and/or reduction in production due to plugging of
pipelines by wax deposits.
iii. Risk element in development, a problem that could jeopardize the development of
marginal field given the prevailing economic situation. The additional cost of controlling
and managing wax puts a greater risk of abandonment on such fields.
To address the issue of wax deposition, three important phenomena have to be considered: wax
precipitation, dynamic wax deposition and heat transfer from the pipeline. Wax plug, which is
obtained from wax deposition, is a gel that contains solid wax crystal and trapped oil
(Venkatesan et al., 2007). Wax deposition is a thermodynamic phenomenon in which a gel is
formed with wax crystal and liquid. Wax precipitation should be described accurately in order to
develop the solutions to control wax deposition in wellbores.
2.3: Wax Deposition Process
Wax deposition occurs when paraffinic components in crude oil (high molecular weight alkanes
with carbon numbers greater than 20) precipitate and deposit on cold pipeline wall when the
inner wall temperature falls below the Wax Appearance Temperature (WAT). Whereas wax
9
precipitation during oil flow results in wax deposition and flow restriction, wax precipitation
during a production shutdown results in problems when attempting to restart the flow ( Al-Yaari,
2011). When the transportation in a pipeline is stopped due to a planned maintenance or an
emergency situation such as severe weather conditions on offshore platform the temperature and
solubility of wax decrease and wax molecules precipitate out of liquid phase in static condition
(Fung et al, 2006, Thomason, 2000)
When waxy oil flowing in cold lines is cooled, it gels due to the formation of a network of wax
crystals. Unlike in the case of inorganic solutions, where there is hardly any interaction among
the salt crystals, the wax crystals have a strong interaction and affinity, resulting in the formation
of the network. Although oil (solvent) and wax (solute) have a similar chemical nature, their
molecular weights are quite different. Waxes have a higher molecular weight and they tend to
form stable wax crystals that interlock to form a solid network. The network of wax traps a large
quantity of oil (Holder and Winkler, 1965). Hence, the initial stage of the deposition of the waxy
oil mixture on a cold surface is the formation of a gel layer with large fraction of trapped oil.
Wax deposition can be mild or it can be severe enough to become unmanageable (Figure1)
Figure1: Wax deposit in a retrieved pipeline
10
The wax deposition process can be described by the following steps (Hernandez et al., 2004).
1. Gelation of the waxy oil (formation of incipient gel layer) on the cold surface.
2. Diffusion of waxes (hydrocarbon with carbon numbers greater than the critical number)
towards the gel layer from the bulk oil.
3. Internal diffusion of these molecules through the trapped oil.
4. Precipitation of these molecules in the deposit.
5. Counter-diffusion of de-waxed oil (hydrocarbon with carbon numbers lower than the
critical carbon number) out of the gel layer.
The last three steps result in an increase of the solid wax content of the deposit. Though there is
ongoing efforts to gain insight into the physical phenomenon of wax deposition, a model to give
reliable guesses of the wax build up is still lacking (Gjermundsen and Duenas, 2006; Mehrotra
and Bhat, 2007). A reliable wax deposition experimental study will be an invaluable tool for the
development of such effective model for optimal scheduling and removal of the deposited gel.
2.3.1: Factors Leading to Wax Precipitation and Deposition
Wax precipitation occurs when the wax molecules contained in the crude oil reach their
solubility limit due to change in equilibrium conditions in the crude, resulting in loss of paraffin
solubility. Temperature is the major factor affecting solubility of wax in crude oil, and hence it is
defined in term of temperature at other specified conditions. They influence wax precipitation by
influencing the solubility limit through the shifting of the temperature upward and downward.
Other factors such as oil composition and the available solution gas, oil pressure which affect the
gas in solution, flow rate, completion, and deposition surface roughness affect the deposition
process by creating favourable environment for deposition to occur.
11
2.3.1.1: Bulk Oil and Pipe Wall (Coolant) Temperature differential
Temperature is undisputedly the predominant and most important factor in wax precipitation and
deposition due to its effect on wax solubility. Sadeghazad et al. (1998) concluded that
temperatures together with the amount of light hydrocarbon are the two most prominent factors
affecting wax precipitation and deposition. Solubility of paraffin (wax) increases with increasing
temperature and decreases with decreasing temperature.
Singh et al., (2000) observed the relationship between wax solubility and temperature. Wax has
been reported to precipitate from crude oil when the bulk oil temperature falls below the Wax
Appearance Temperature (WAT), or cloud point temperature (Erickson et al., 1993). All other
factors contribute to wax deposition when the crude oil temperature is below the WAT. The
surrounding temperature around the pipe is generally less than the bulk crude oil temperature in
the pipe. Thus there is heat loss to the surrounding water through the pipe-wall due to
temperature gradient in the radial direction of the pipeline i.e. between the bulk oil and the colder
pipe-wall. The temperature gradient resulted in wax deposition as the pipe wall temperature falls
below the WAT of the crude oil. Eaton et al., 1976 observed that the rate of wax deposition is in
direct proportion to the temperature difference between the bulk oil and the cold pipe. However,
Haq (1981) concluded studies that maintaining the coolant (pipe wall) temperature at a constant
value below the WAT of the oil and varying the bulk oil temperature reduce the amount of wax
deposited as the temperature difference between the bulk oil and the pipe wall (coolant)
temperature increases.
12
2.3.1.2: Flow rate
The rate at which fluid flow in the pipe line also affect the rate of wax deposition, this is through
the shear removal of the deposited wax as a result of shear stress on the surface of deposited
wax. Flow rates also affect the residence time of the crude oil in the pipeline as higher residence
time results in more wax deposit on the pipe wall. However, in term of shear removal, the higher
the flow rates the less the amounts of wax deposit. The resultant effect of the flow rates depends
on which of these two contradicting effects is overwhelming. In the laminar flow region the
effect of residence time is the controlling effect, while in the pseudo-turbulent and turbulent
regime the effect of shear removal is overwhelming. Increasing the flow rate from laminar flow
through the pseudo-turbulent flow to turbulent flow reduces maximum deposition rate and at the
same time lower the temperature at which the maximum deposition rate occurs (Hsu et al.,
1994). Low flow rates offer the moving oil stream longer residence flow channel. The increase
residence time allows more heat loss to the surrounding, leading to higher chance of bulk
temperature falling below the WAT and enough time for wax precipitation and deposition.
However wax deposition decreases as the flow moves to the turbulent regime. The turbulent
stream exerts a kind of viscous force, which tends to drag or slough the wax deposit from the
pipe wall. When this viscous drag exceeds the resistance to shear in the deposit, the wax then
sloughs and is lodged back to the liquid. This removal mechanism has a significant impact on the
wax deposition rate (Hsu et al., 1994). There is a difference in texture between wax deposited at
high flow rates and wax deposited at low flow rates (Jessen and Howell, 1958; Tronov, 1969;
Haq, 1981). The deposited paraffin wax at high flow rates appears harder, being more
13
compactedand more firmly attached to the deposition surface, the molecules having good
cohesion among them.
2.3.1.3: Pipe-wall Surface Roughness
Hunt (1962) study on the effect of deposition surface on paraffin deposition concluded that the
deposits do not adhere to metal themselves, but are held in place by surface roughness which act
as wax nucleating sites. Jorda (1966) observed that paraffin deposition increases with greater
surface roughness. In the wax deposition study with pipes of different materials, Jessens and
Howell, (1958) concluded that the amount of wax deposited on a smooth surface is less than that
deposited on steel. However, Patton and Casad (1970) could not see any correlation between the
wax deposition and surface roughness, but opined that adhesion bond at a surface should be
proportional to the total contact area and therefore related to surface roughness.
2.3.1.4: Gas-Oil Ratio
Gas-oil ratio influences wax deposition in a manner that depends on the pressure regime. Above
the bubble point, where all gases remain in solution, solution gas helps to keep wax in in
solution. Luo et al., (2001) reported that wax appearance temperature will be high with low gas
oil ratio, while Singh et al., (2004) observed that injection of lift gas in a closed loop reduced
wax deposition by causing a depression in wax appearance temperature as a function of pressure.
High gas oil ratio (GOR) would result in more expansion and subsequent cooling as pressure of
the oil system depletes, a situation that can aggravate the wax deposition problem. In a study to
reduce wax appearance temperature by injecting of diluent lift gas, Singh et al., (2004) noted
good results were not obtained in high GOR wells.
14
2.3.1.5: Crude Oil Composition
Crude oil is composed of saturates, aromatics, resins and asphaltenes (SARA), the distribution of
which in a particular oil system is shown by the SARA analysis. SARA determines the ability of
crude oil to deposit wax solid on the pipe wall and thus the stability of crude oil. Saturates are
flexible in nature, the flexibility being highest in normal paraffins because they are straight chain
compounds. The very high flexibility of normal paraffins makes it possible for them to cluster
and crystallize. The iso-paraffin equally enjoy a high level of flexibility, but form a more
unstable wax. Cyclo-paraffin (Naphthenes) are least flexible due to their cyclic nature and do not
contribute much to wax deposition.
The components of crude oil are in thermodynamic equilibrium at initial reservoir conditions. It
is known that aromatics serve as solvents for high molecular weight saturates, which are the
sources of paraffin waxes in crude oil while the polar components, especially asphaltenes induce
wax nucleation (Hammami et al., 1999). Singh et al., (2001) however reported that the solubility
of paraffin in aromatic, naphtenic, and other organic solvents becomes low at room temperature
(low temperatures). Light ends of saturates equally help to keep the high molecular weight heavy
ends in solution. The onset of production results in the loss of these light ends, as they are first to
leave the reservoir. This alters the original composition of the oil system, resulting in decreased
solubility of the paraffin waxes. This loss of solubility could lead to precipitation and deposition
of wax. In a model study, Huanquan et al.,(1997) reported that increasing the percentage of light
(C5) in a synthetic oil system decreased the cloud point temperature, reducing the chance of
deposition. Generally, the weight percent of the saturates in the crude oil, the structural
distribution of the paraffin components and occurrence of other solid like formation fines,
corrosion materials and presence of asphaltene which could form nucleating sites-all contribute
15
to wax precipitation and deposition. Oils containing high C30+ (especially normal paraffin C30+)
concentrations exhibit high cloud point temperature (Ferworn et al., 1997).
Therefore, knowledge of the oil composition (SARA) gives a fair idea of the wax deposit
potential of the crude oil and hence, the oil stability. Stability of the oil is dependence on its solid
content and the balance between the aromatics and saturates. By SARA analysis of oils,
distribution by weight percent of saturates, aromatics, resins and asphaltene component, for
stable and unstable crude is as follows:
Unstable crude: Saturates > Aromatics > Resins > Asphaltenes
Stable crude: Aromatics > Saturates > Resins > Asphaltenes
This distribution is as expected since the aromatics keep the heavy paraffin wax in solution,
while the crude oil system that displays a large amount of saturates (paraffin) is likely to be
unstable (Carbognani et al., 1999) and thus precipitate and deposit wax (Oyekunle and
Adeyanju, 2011).
2.3.1.6: Pressure
Pressure being one of the important parameter in the production of reservoir fluids, plays an
important role in wax precipitation and deposition. As reservoir fluids are being produced, the
pressure drops as the reservoir fluid moves from the reservoir to the surface equipment. The
lighter components of the reservoir fluid are first to leave the reservoir as pressure drops, this
causes an increase in the solute solvent ratio, since the light ends serve as solvent to the wax
crystals in the oil. Hence the loss of these lights ends leads to reduction in wax solubility.
16
Brown et al., (1994) Studied the effect of pressure on the cloud point of dead and live oil through
the cloud point measurement at atmospheric and higher pressure. The WAT was observed to
increase with pressure above bubble point, at constant composition. This implies that pressure
increment during one phase flow (above bubble point) will favour wax deposition. This is
contrary to the phenomenon during the two phase flow (below the bubble point pressure). The
wax appearance temperature decreases with increase in pressure up to bubble point pressure
(Brown et al., 1994) due to dissolution of light ends back to the liquid phase. The WAT increases
with increase in pressure for stock tank oil referred to as dead oil (Brown et al., 1994; Karan et
al., 2000). Huanquan et al.,(1997) also reported that the WAT increases with pressure for a fixed
component of liquid mixture.
2.3.2: Wax Crystallization Process
The precipitation of the wax molecules occurs when the temperature of the crude oil decreases
below the wax appearance temperature (WAT), or sometimes called cloud point. This is the
temperature when the first wax crystal begins to precipitate out from the crude oil. The
crystallization of the wax molecules can further develop an ordered solid structure, which traps
the crude oil causing it to experience a gel consistence with low flow ability. The crystallization
process can be divided into two stages called nucleation and growth.
When the temperature decreases the molecular motion of the molecules in the crude oil are being
hindered, due to less energy to move around freely. This will make it easier for the wax
molecules to come closer and aligned together, and eventually attached to each other and reach a
critical and stable size. The cluster of the wax molecules is called nuclei, and the formation
process nucleation. The nuclei will be stable as long as the temperature is below the melting
17
point for the wax.However increasing the temperature, the nuclei structure will be disrupted by
thermal motion. Once the nuclei is formed and the temperature is kept low, more wax molecules
will precipitate and continue to grow on the nucleation site, which is referred to as the growth
process (Ajienka, 1981; Ahmed, 1991, 2007; Oyekunle and Adeyanju, 2011).
The nucleation process can be further characterized as either homogeneous or heterogeneous.
Homogeneous nucleation refers to the formation of a pure nuclei cluster, which means that the
wax molecules have spontaneously and randomly crystallized together forming stable wax
nucleus. Heterogeneous nucleation happens when there are provided an active site for the
nucleation process to occur, caused by impurities in the solution. The latter process is the most
common nucleation process for crude oil, because crude oils normally contains a lot of
impurities such as asphaltenes, formation fines, clay and corrosion products, which can act as a
nucleation site for the wax crystals (Hammami and Raines, 1999). However asphaltenes present
in the crude oil have also shown to inhibite wax deposition, because of disruption they cause to
the growth process of the wax crystallization network and thereby increase the mobility of the
crude oil (Dobbs, 1999).
2.3.3: Wax Deposition Mechanisms
The most important factor governing the wax deposition is the reduction in the temperature of
the crude oil, due to the cooling of the flow lines by the cold seawater. The flow rate of the well
stream fluid in the flow line is important to the settling of the wax crystal on the wall surface,
and the stripping of wax deposition off the pipeline walls. The surface properties of the pipeline
also play an important role. If the surface material is rough this can lead to more wax crystal
adhering to the surface wall, and thereby increasing the deposition (Dobbs, 1999).
18
Many different wax deposition mechanisms have been proposed, such as molecular diffusion,
Brownian diffusion, shear dispersion and gravitational settling (Misra, et al., 1995). However,
the mechanisms are still not fully confirmed, but molecular diffusion and shear dispersion are
regarded as primary wax deposition mechanisms (Kelland, 2009; Leiroz and Azevedo, 2005).
2.3.3.1: Molecular Diffusion Mechanism
The wax starts to precipitate when the temperature of the pipeline surface becomes lower than
the WAT of the crude oil. The crude oil near the wall surface will crystallize and adhere to the
surface wall, which will create a concentration gradient between the wax concentration in the
bulk liquid. The concentration gradient means that the wax molecules will move from high
concentration to low concentration. The low wax concentration is towards the pipe wall, due to
the wax precipitating out of the solution and deposits on the pipe wall. The molecular diffusion
mechanism has widely been accepted as dominant wax deposition mechanism.
2.3.3.2: Shear Dispersion Mechanism
The stream flowdoes normallyexperience a difference in the velocity, depending on where in the
fluid it is being measured. The bulk fluid will often have a higher velocity compared to the fluid
near the pipe wall, due to the friction between the fluid and the wall. Already precipitated wax
near the wall will therefore move to the pipe wall where the velocity are lower and deposit.
However increasing the pumping pressure will increase the shear rate, thus decreasing the wax
deposition rate. Figure 2a below shows an illustration of the concept of the different velocity
profiles that can be experienced in the flow lines, laminar and turbulent flow, which depend on
the pressure applied to the well stream flow.
19
Figure 2a: Shear dispersions mechanism during wax deposition.
High shear environment, found in turbulent flow, will disrupt the adhering and growing of wax
deposits on the pipe wall. The low shear environment, laminar flow, will cause the wax
molecules to easily adhere to the pipe wall, and cluster together (Hammami and Raines, 1995).
When small particles are suspended in a fluid that is in laminar motion, the particles (n-paraffin,
iso-paraffin, and cyclo-paraffin) tend to move at the mean speed and in the direction of
surrounding fluid (Figure 2b). The particle speed is that of streamline at its center, and the
particle rotates with an angular velocity which is half the fluid shear rate. If the particles
approach a solid boundary, both linear and angular velocities will be reduced. Because of fluid
viscosity, rotating particles will impart a circulatory motion to a layer of fluid adjacent to the
particle. This rotating fluid region exerts a drag force on the neighboring particles. In a shear
field, each particle passes and interacts with nearby particles in slower or faster moving
streamlines. When only two particles are present, far from a wall and at a very low Reynolds
number, these passing encounters result in large temporary displacement. As the particles pass,
their trajectories are such that the particles curve around one another and return to their original
20
streamline. Thus, there is no net lateral displacement. If the particle concentration is high,
however, then a significant number of multi-particle interactions will occur. These multi-particle
collision results in net lateral transport and a dispersing of particles toward the pipe wall in the
contained pipelines. The same is true for turbulent flow except the additional effect of shear
removal (Bhattacharya, 1991).
Figure 2b: Schematic diagram of particles movement in shear dispersion process
2.3.4: Production Problems of Wax Deposition
The wax deposition problems occur when the temperature of the crude oil decreases below the
WAT. The WAT is different for each crude oil, and is dependent on the pressure and the crude
oil composition. The concentrations of light ends in the crude oil have a significant effect on the
WAT, because they act as a solvent for the wax in the crude oil. The light ends are light
hydrocarbon, such as methane and ethane, and other gases such as carbon (IV) oxide and
nitrogen. When the pressure is reduced this can lead to the release of the light ends to the
21
gaseous phase (Carnahan, 1989). A decrease in the light ends can thereby reduce the solubility of
the wax in the crude oil, resulting in an increase in the WAT. When the WAT increases the
temperature at which wax begins to precipitate get higher, and it can be as high as 50oC for some
oils. In deep water flow lines the temperature is usually low, and if the distance from the
reservoir to the process facilities is long, the cold sea water will eventually start to cool the well
stream flow, leading to wax precipitation inside the flow line walls. Three major problems
areassociated with the wax crystallization of the crude (Kelland, 2009; Misra et al.,1999):
2.3.4.1: High Viscosity and Pressure Loss of Crude Oil Flow
When the wax begins to crystallize this will resultin higherviscosity of the crude oil and loss of
the pressure in the pipeline, which can reduce the effective capacity of the line. A high viscosity
and wax deposition in the flow lines will require a higher pumping pressure to ensure flow of the
crude oil.However, severe gelling of the crude oil can, in worst case, lead to production stop.
2.3.4.2: High Yield Stress for Restorationof Flow
When the crude oil is allowed to be stored in the pipeline over a time period, such as during shut-
in, the crude oil can be cooled down to below the pour point. The pour point being the
temperature at which the crude oil begins to gel, causing no movement of the crude oil. At the
pour-point a high extent of wax crystallization has resulted in a wax crystal matrix, which traps
the liquid crude oil and prevents it from flowing. The wax crystal matrix is formed when the
waxes precipitate out as plates like crystals, which interact with each other and form a three
dimensional structure. The temperature where the crude oil begins to gel is often 10 – 30oC
below the WAT. To restore the flow of the crude oil a high pressure must be applied to break the
22
gel. However, sometimes the pressure needed to resume the flow is not great enough, leaving the
pipeline chocked.
2.3.4.3: Deposition of Wax Crystals on Surfaces
If the flow lines are cooled to the temperature below the WAT, the wax in the crude oil near the
pipeline can start to precipitate and deposit on the wall surface. The build up of the wax deposit
inside the pipeline will cause a reduction in the wall stream flow, due to a smaller pipeline
diameter and roughness of the pipeline surface.
2.3.5: Wax Treatment Techniques
In flow assurance the most important task is to maintain the well stream flow, and prevent any
blockage from occurring that can stop the production. In order to sustain an optimal production
at all time, thorough analysis and research have to be implemented in the planning phase of the
field, to achieve better understanding and predict any possible challenges that can occur. This
include important knowledge about pressures, temperatures, flows, fluid samplings, solid
samplings and laboratory testing, which will all be helpful in determining the processes that can
take place in the flow lines, and be able to come up with solutions to the challenges that may
arise. In most cases the flow assurance solution would add a substantial amount of cost for the
project, however these costs would be small compared to loss caused by reduction or shutdown
of the production due to pipeline blockage (Abney et al., 2003).
The problems related to wax deposition are often associated with deep-water flow line, and tend
to involve more difficult flow assurance challenges. However there have been evolved numerous
techniques to remove and reduce the wax build-up of wax deposition down-holes and in the
flow-lines. Many of the methods have achieved great results when applied.However, their
23
success rate is dependent on the wax composition and crude oil found in the specific reservoir,
and therefore will vary for the different fields. In many cases a combination of two or more
treatment methods have been used to more efficiently remove and reduce the wax depositions.
The wax treatments techniques are divided into two categories, the wax removal techniques and
the wax prevention techniques (Mokhatab, 2003).
2.3.6: Wax Removal Techniques
The most common methods used for removing already deposit wax are by mechanical removal,
heat application and chemical removal. However some of these techniques are also regarded as
wax preventive techniques, such as with heat application.
2.3.6.1: Mechanical Removal
Pigging is the first widely used mechanical removal technique used to remove wax deposits
formed at the pipeline walls, and is a good and cost effective method. However in recent years
more difficult offshore fields have been exploited, where deep and long tieback pipelines are
being applied, which have lead to more extensive wax problems. These conditions, have required
more frequent cleanings, thereby making pigging difficult and costly operation. In many fields it
is therefore usual to perform pigging in combination with for example wax inhibitor chemicals or
hot oil (Al-Yarri, 2011, Mokhatab, 2009).
The pigging operations are performed by using a scraper or cutter, which will scrape the wax
deposition of the walls inside the pipelines. There are developed different type of pigs with
different shapes and materials. The most convectional used pigs are disc, cup and polly, which
all have different wax removal performance depending on the wax hardness and thickness. The
frequency of the pigging operations depends on the severity of the wax depositions, and can vary
24
from weeks to months. However, too much wax deposition in the flow lines will require more
pressure to operate the pig, and too much wax cuttings in the flow lines can reduce or block the
well stream fluid (Wang et al., 2001).
2.3.6.2: Heat Application
Different thermal solutions for removing or reducing wax deposition have shown to be
successful for flow lines in deep water fields. Active heating has especially been regarded as an
efficient flow assurance solution against wax due to the advantages to control the temperature
above the formation region (Esaklul, 2003). The thermal management system is normally broken
down to passive and active heating. Passive insulation is mainly techniques used to prevent the
warm well stream fluid to lose its heat by using different insulation methods. Active heating use
2.3.6.4: Hot Oiling Flow
The hot oiling is a technique that has been used a lot in the early days in the oil and gas industry
for removing wax depositions down hole and in the flow lines. In the process hot oil is heated to
a temperature above the melting point for wax and then pumped into the well, normally through
the annulus space. The heated oil will circulate around the tubing and melts the wax deposits
inside the tubing, and then be brought back to the surface via a heating system in the production
tubing. As a result of the process the melted wax inside the tubing will be transported with the
well stream fluids through the flow lines to the production facilities. During the hot-oil process a
wax dispersant is usually added to the crude, to enhance the dispersion of the melted wax with
the crude oil. The hot-oil technique is a fast and simple method, which will effectively remove
the wax deposition. However, this method can be dangerous, especially when applied to wells
25
that have crude with low flash point. There have been operations where the hot-oil has been
replaced with steam or hot water to melt the wax, but is rarely used due to risk of corrosion and
emulsion problems.
The hot oiling method is not usually used in subsea flow lines, due to high expenses related to
heating the oil, there have also been reports showing formation damage due to hot-oiling
(Kelland, 2009; Ashton et al., 1989; Barker, 1989).
2.3.6.5: Application of Wax Solvent (Inhibitors)
The adding of solvent down-hole and to the flow lines have shown to be effective to remove wax
deposition, as it will help to resolve the precipitated wax and make it easier to transport the crude
oil to the surface (Ferworn et al., 1997). The different solvents have different abilities to dissolve
wax deposit, which depends on the wax, the location in the system and the temperature.
Normally the solvents are applied in frequent batch treatments orcontinuously. Aliphatics and
aromatic solvents are two main group of solvents used in the oil fields. The aliphatics solvents
are straight or ring formed hydrocarbons, and the most common aliphatic used are diesel,
kerosene and condensate. The aromatic solvents are chemical containing a carbon ring with
delocalized electron double bonding, such as xylene and toluene, which are the most popular
solvents used today. A mixture of toluene and xylene together with an aliphatic solvent has
shown to increase the wax removal. Adding a surfactant can also enhance the performance of the
solvents by helping it to disperse the waxes (Thierheimer, 1990). In subsea systems or cold
climates, solvents alone may not be effective, and usually have to be used in combination with
other methods, such as hot oiling. More faster and effective wax removal has been achieved by
heating xylene and toluene when employed (Kelland, 2009; Straub et al., 1989). Other
26
solventsthat have been applied with good success are benzene, chlorinated hydrocarbon, and
carbon disulphide. However, many of the solvents used are not environmental friendly and some
highly toxic. Many also possess dangerous risks related to low flash points, such as aromatic
solvent and corrosion problems, as seen with chlorinated hydrocarbon (Woo et al., 1984).
2.3.6.6: Application of Thermochemical Packages
Thermochemical packages are wax removal methods that utilizes the heat of an exothermic
reaction to melt the wax deposits down-hole or in subsea flow lines. A common reaction used is
the acid-catalyzed decomposition of ammonium nitrate. The reaction starts by mixing a sodium
nitrate or chloride in an aqueous solution, in the presence of HCl as catalyst, as shown in the
equation below
NH4X + NaNO2  N2 + 2H2O + NaX + Heat
This developed a huge amount of heat as the reaction rate increases, which will be transported
upwards the pipelines and melt the waxes. The reaction is controlled by keeping the pH in range
between 5.0 and 8.0. However, keeping the reaction under control is very important, as the
reaction can cause great risks to the pipelines and materials, or worst case damages the pipeline
system (Kelland, 2009; Ashton et al., 1989).
2.3.7: Wax Preventive Techniques
Wax preventive techniques can be divided into major and minor methods depending on the
applied method.
27
2.3.7.1: Major Methods
Wax preventive techniques are different solutions intended to prevent the wax from precipitate
and deposit in the flow lines. The preventive solutions can be as simple as (Mokhatab, 2003):
a. Pressure Method: Increasing the pipeline pressure, this will lead to an increase in the
shear rate, thereby making it harder for the wax to deposit inside the pipeline walls.
b. Lamination Method:Insulate or bury the flow lines, so that no heat will be lost from the
pipeline stream fluid, thereby keeping the temperature above WAT. Different insulation
materials used are polypropylene foam, rubber, glass and plastic, all of which have low
heat transfer coefficients assuring less heat lost to the surroundings. Burying the flow-
lines are effective due to high heat capacity of the soil, which will absorbed and store the
heat from the flow lines (Thant et al., 2011).
c. Coating Method:Coating the inside of the flow line with material that will keep the wax
from adhering on the pipeline wall. Types of coating material that have been proposed
are plastic pipes or plastic-coated pipes. These have shown to decrease the deposition rate
on the pipeline walls.However, when a layer of wax deposition have covered the pipe, the
growing rate will be the same as for steel pipelines (Al-Yaari, 2011).
2.3.7.2: Minor Methods
Some other preventive methods are not commonly used and these include:
a. Cold flow: This is a new technology that is being exploited as a method for removing
wax deposition in deep water and cold oil fields. The approach is to decrease the
temperature of the bulk fluid, to prevent the temperature gradient in the flow lines, and
thereby stop the wax precipitate from depositing on the pipeline walls. The whole well
28
stream fluid will be cooled down to the temperature of the surrounding sea water, thereby
making the wax precipitate inside the bulk fluid and be transported as solid dispersion
within the well stream fluid (Aiyejina et al., 2011).
b. Magnetic fluid conditioning (MFC): The technology utilizes a magnetic field to alter
the kinetics of the precipitated wax, which disturbs the wax agglomeration process,
making it harder for wax to precipitate and grow into larger crystals (Tung et al., 2001).
Studies have shown the WAT have not been changed by the MFC technique, but it had
an effect on the viscosity of the crude oil (Chow et al., 2000).
c. Microbial prevention: The uses of selectively isolated bacteria have shown to be
effective and environmental friendly method to both removal and prevention of wax
deposition in flow lines (Sadeghadezad and Ghaemi, 2003). The selected microbes used
are called paraffin degrading bacteria (PDB). These are tough bacterial that can survive
the extreme conditions found in the reservoir, well tubing and surface flow lines with
varying temperature, pressure and anaerobic environment (no oxygen present). The
micro-organism mechanisms are to biodegrade the long wax molecules into smaller
components, thereby preventing wax precipitation and deposition in the flow lines. The
mechanism will cause a decrease in the WAT as a result of the breakdown of the wax
molecules, which will decrease the wax content in the crude oil, and increasing the API
gravity of the oil. The API gravity is a measurement of how light or heavy oil compared
to water. The bacterial can also form a bio-film layer inside the pipeline wall, which will
prevent wax from clinging to the pipeline. (Rana et al., 2010).
29
2.3.8: Crude Oil Wax Deposition in Subsea Environment Pipelines
Crude oil is a complex mixture of saturates (parrafins/waxes), aromatics, naphthenes and resin.
Among these components, high molecular weight paraffin (wax) and asphaltenes are responsible
for production and transportation problems in subsea pipeline systems. At reservoir temperatures
(70 – 150oC) and pressure (50 - 100MPa), wax molecules, are dissolved in the crude oil,
However, the crude oil flows through a sub-sea pipeline resting on the ocean floor (Figure 3) at a
low temperature (can be as low as 4oC) (Lee, 2008).
Figure 3: Typical Subsea pipelines
Wax is a component of crude oil that remains in solution until operating conditions are
favourable to its precipitation, a condition caused by changes in the temperature-pressure
equilibrium of the crude oil. Upon precipitation (crystallization) wax is deposited on the
components of the production system which includes molecular diffusion, shear dispersion,
Brownian diffusion and gravity settling.
30
Wax deposition has been reported in all facets of the production system including the reservoir,
wellbore, tubing, flow lines, and surface facilities.Temperature reduction/heat loss is a dominant
factor in wax problems, as wax begins to precipitate from the crude when the temperature falls to
or below the cloud point (wax appearance temperature, (WAT)). Other factors such as pressure,
oil composition, gas-oil ratio, water oil ratio, flow rate, well completion, and pipeline roughness
also contribute to wax deposition problem. As crude oil flow through sub-cooled pipeline there
exist temperature gradient along the pipeline radial direction.The crude oil solubility is
dependent on the temperature of the crude oil.From Fick’s law, there exist dissolved wax
concentration gradient along the radial direction, leading to diffusion of dissolved wax towards
the pipe wall forming a gel-like structure which eventually transform to wax deposit (Figure4).
Figure 4: Schematic diagram of crude oil flow in subsea environment
31
Wax deposition can be mild or it can be severe enough that it is unmanageable. The earlier the
problem is diagnosed in the life of a field, the easier it will be to design a preventive or control
management plan.
2.3.9: Gulf of Guinea Subsea Environment
The Gulf of Guinea is the north easternmost part of the tropical Atlantic Ocean between Cape
Lopez in Gabon, north and west to Cape three points in the western region Ghana. Among the
many rivers that drain in the Gulf of Guinea are the Niger and the Volta. The coast lines on the
Gulf include the Bight of Benin and the Bight of Bonny. The Gulf of Guinea region (Figure 5)
along with the Congo river delta and Angola, further south are expected to provide a quarter of
the United states crude oil imports by 2030. This region is now regarded as one of the world’s
top crude oil and natural gas exploration hot sports. The region is currently the source of around
5.4 millions barrels of oil per day this is equivalent to more than the total amount imported by
EU 27 countries in 2008, (4.9 millions barrels per day).
Figure 5: The Map of Gulf of Guinea Region.

32
Oil supply from the region in 2011 was equivalent to 40% of total EU 27 and 29% of total
United State petroleum consumption in the same year respectively. Angola and Nigeria account
for 34% and 47% respectively of the region total crude oil supply (Zhu et al., 2008).
The crude oils produced from the fields in the Nigerian offshore are of API gravity of between
35-45o, low sulphur (0.5 - 1%) and high wax content (between 20 – 45%) similar to what is
obtainable in most fields in the Gulf of Guinea. The Gulf of Guinea subsea temperature ranges
between 20 -29oC in the shallow water and can be as low as 4oC for deep water (Ajienka, 1981).
The Figure 6 below shows the variation of the temperature with depth in an ultra deep subsea
environment in the Nigerian Niger delta part of Gulf of Guinea.
Figure 6: Temperature variation in a typical ultra deep location in the Gulf of Guinea.
33
The high wax content of these crudes makes their transportation from the production platform or
floating, production, storage and overloading (FPSO) platform to the export terminals through
subsea pipelines to be prone to wax deposition on the pipe wall.
2.4: Laboratory Studies of Wax Deposition
In order to study the wax deposition in crude oil flow in subsea (sub-cooled) pipelines, different
scholars have simulated the wax deposition process in the laboratory. The main apparatus are the
cold finger and flow loop. In cold finger experiments, small steel cylinder heat exchangers
maintained at lower temperature are immersed in higher temperature crude oil. The bulk oil and
the coolant temperatures are controlled by two different water baths. While most flow loop
devices, comprise of the annular pipe with hot crude oil flowing in inner pipe and the coolant
liquid flowing in the annulus between the inner pipe and the outer pipe. The flow rates of the
crude oils are controlled by regulating the pump, while the oil and coolant temperature are
regulated by the two separate water baths. Although the flow loop devices better simulate the
wax deposition in subsea flow-lines than the cold finger devices. The cold finger devices have
advantages in its simplicity, easy to operate and relatively smaller crude oil samples needed for
their operations.
2.4.1: Cold Finger Experiments
Paso and Fogler (2004) using cold finger device to study wax deposition at lower temperatures
(keeping the coolant temperature below the pour point temperature) observed an increase in wax
deposit volume and thickness at decreased temperature differential.
Jennings and Weispfennig (2005) observed similar results in their study to determine wax
deposition at different coolant temperature with constant bulk oil temperature using a crude oil
34
sample from the Gulf of Mexico in cold finger apparatus. It was concluded that the wax deposit
decreased for every increase in coolant temperature. The study also showed an increase in wax
deposit and the percentage of trapped oil in the deposit when the coolant temperature decreases.
2.4.2: Flow Loop Experiments
Creek et al., (1999), using flow loop apparatus to determine wax deposition at different coolant
temperatures while bulk oil temperature was kept constant, observed that experimental runs with
lower coolant temperatures (higher temperature differential) though yielded more amount of wax
deposit but the wax deposits are relatively softer when compared to wax deposit from higher
coolant temperature experiments.
Todi and Deo (2006), using a transparent experimental set-up studied wax deposition process in
relation to particle transport at three different pipe wall temperatures, showed that shear
dispersion and Brownian diffusion in the pipe section tend to cause a uniform dispersion across
the channel.The particle concentration distribution depended on the competition between the two
forces and wax deposition occurred as long as the temperature was below the crude oil wax
appearance temperature irrespective of the flux nature.
Edmonds et al., (2008), developed a wax deposition model taking the wax phase as a continuous
distribution of n-paraffin components. Shear stress was incorporated into the mass transfer model
and coupled with thermodynamic model to predict mass of the deposited wax. When validated
with flow loop experimental results it was concluded that incorporating compositional analysis
of the oil into the model predict the wax deposition in pipeline better than non compositional
treatment of the crude oil.
35
Bidmus and Mehrotra (2009), studied the effect of different inlet oil temperatures on wax
deposition process keeping the coolant temperature constant using flow-loop device. It was
discovered that the wax deposit volume decreased when the bulk oil temperature decreased in the
cold flow regimes. It was observed that wax deposit increased when the bulk oil temperature
decreased in the hot flow regimes. It was also observed that the quantity of wax deposit highly
decreased when the bulk oil temperature was close to the coolant temperature, with the coolant
temperature was higher than the crude oil pour point but lower than its wax appearance
temperature, (WAT).
Lashkarbolook et al., (2010) used flow loop device to study the effect of temperature on wax
deposition in a crude oil sample (WAT = 31oC) under laminar flow regime, while maintaining
the bulk oil at 40oC. It was shown that the wax deposit volume increases when the coolant
temperatures decrease at constant bulk oil temperature.
Kelechukwu et al., (2010) observed that the wax deposit is dependent on each of the studied
variables, with wax deposit decreasing with increasing flow rate and temperature differential, but
increased with residence time attaining a maximum value and then gradually tailed off.
Hoffmann and Amundsen, (2010) investigated waxy crude oil behaviour in subsea production
lines using a 2 inches flow loop device. A North Sea waxy gas condensate was used to
investigate wax deposition in turbulent single phase flow under different temperature and flow
conditions. Wax thickness and roughness was determined from pressure drop, weight and laser
measurement. It was observed that though the molecular diffusion is the central mechanisms
responsible for wax deposition the effect of deposited wax composition and shear stress cannot
be under-estimated.
36
Huang et al., (2011), while studying the flow of hot crude oil in relatively colder pipe in a flow
loop device, observed the decrease in wax deposit thickness as the bulk oil temperatures
increased while keeping the coolant temperature at 5oC (the temperature at which all wax was
believed to have precipitated out of the crude oil).
Osokogwu and Otung, (2012) developed semi empirical model for wax precipitation and
deposition along oil field installation under various flow conditions Though there was agreement
between the models output and published works, the model slightly under-predicted the wax
thickness when compared to experimental data from Banki et al., (2008a; b).
The effect of varying operating temperatures on wax deposition had been studied by different
scholars, various conclusions were made and consensus opinion was not reached. Some observed
the wax deposit to increase with the temperature differential (Creek et al, 1999; Jennings and
Weispfennig, 2005; Bidmus and Mehrotra, 2009; Burger et al., 1981, Lashkarbolook et al.,
2010; Huang et al., 2011), while others concluded that the wax deposit decreased as temperature
differential is increased (Paso and Fogler, 2004; Osokogwu and Otung, 2012).
2.5: Chemical Wax Prevention
In the recent years, there has been more focus on using chemical prevention against wax
deposition to minimize transportation problems with waxy crude oils. These wax reducing
chemicals are added early to the well stream fluid, normally by injection at the wellhead or down
hole (Bello et al., 2005; 2006). Here their presence will affect the wax crystallization and growth
process of the crude oils, which will occur as the temperature of the well stream fluid decreases
during the transportation in flow lines. In many cases the application of wax preventive
chemicals alone will not fully manage to prevent wax depositions in the whole flow-line,
37
however they can reduce the frequency of other wax removing techniques, such as mechanical
removal (Kelland, 2009).
There are three main different preventive chemicals (wax inhibitors) used against wax
deposition. These are
i. Wax dispersants
ii. Wax crystal modifiers
iii. Pour point depressants (PPDs) or flow improvers.
Wax dispersants are surface active chemicals, which prevent wax deposition by reducing their
tendency to adhere to the pipeline walls, and by keeping the precipitated wax dispersed as
separated particles (Pedersen and Ronningsen, 2003). The preventive mechanisms for wax
dispersants are quite different compared to the mechanisms for the wax inhibitors and PPDs.
The wax inhibitors are chemicals referred to as wax crystal modifiers, and will affect the WAT
of the crude oil. While the pour point depressants (PPDs) are chemical that affects the pour point,
which will reduce the gelling of the crude oil as temperature decreases, and therefore improve
the flow. However for both of these classes to be able reduce the wax deposition they have to be
applied at temperatures above the WAT, so they can interfere with the crystallization process of
the waxes. There are a lot of similarities between the two chemical classes in both the chemistry
and mechanisms, and in most cases the wax inhibitors will also function as PPDs (Kelland,
2009).
38
2.5.1: Wax Dispersants
Wax dispersants as preventive techniques alone are rarely used in the oilfields. However, in
combination with other wax inhibitors, such as PPDs, their presence has given better results in
preventing wax deposition.Wax dispersants are surfactants, which are amphophilic chemicals
that possess both a hydrophobic part, which is water repelling and a hydrophilic part, which is
water attractive, usually the dispersant molecules consist of a long hydrocarbon chain, the
hydrophobic part, with a hydrophilic head.
The wax dispersants ability to reduce wax deposition are mainly by adhering unto the pipeline
wall and creating a surface film, which make it difficult for the precipitated wax to stick to the
wall. Some dispersants can also create a surface film which will let the precipitated wax bind
weakly to the wall, but this bond will easily break off with the turbulent fluid flow, thereby
preventing wax deposition build-up inside the pipelines. Wax dispersants also absorb to the
growing wax crystal, and prevent them from agglomerating together, which will make it easier to
transport the precipitated wax with the well stream flow (Ahn et al., 2005).
Common wax dispersants are alkyl aryl sulphonates, fatty amine ethoxylates, other alkoxylated
product, and imidazolines. Imidazolines dispersants are good surface coating chemicals, which
can also prevent gelling of the crude oil in cold climates and acts as a good corrosion inhibitors
(Kelland, 2009; Martin et al., 2010).
2.5.2: Wax inhibitors and Pour Point Depressants (PPDs)
The chemicals used as wax inhibitors and PPDs are normally polymers with long alkyl chains,
which are intended to interfere with the wax crystallization and growth process. However the
mechanisms for the wax inhibitors and PPDs are not fully understood and different mechanisms
39
have been proposed. The general mechanism is that wax inhibitor polymers contain a structure
similar to the wax structure, which will allow the polymer to be incorporated into the wax crystal
growth. Beside the similar structure the polymer should also contain structural network, which
will cover the wax site and thereby prevent new wax molecules to attach to the wax and
continuing further growth (Jennings and Newberry, 2008). The effect of this alteration causes a
reduction in the three dimensional network the wax deposition can form, and promote the
formation of smaller wax aggregates. The overall effect will give a lower pour point and a
reduction in the viscosity of the crude oil, which makes it easier to transport the crude oil
(Pedersen and Ronningsen, 2003)..
The wax inhibitor chemicals are often waxy materials and solid at room temperature, and
therefore must dissolved before they are applied in the field, (Manka and Ziegler, 2001). The
wax inhibitors are usually dissolved in aromatic solvents when they are applied in cold deep
water fields. The use of a good solvent in cold climates can also avoid gelling of the wax
inhibitor, such as toluene, xylene and cyclo-pentane.
In general, good wax inhibitors, with concentration of 20-50 ppm, have shown to prevent wax
deposition at temperature from 10 – 15 oC below the WAT. While the PPDs, with relatively high
dosage have shown to reduce the pour point temperature with up to 30oC (, Bello, et al., 2005;
Kelland, 2009; Pedersen and Ronningsen, 2003).
The main chemical group used as wax inhibitors and pour point depressants are:
a. Ethylene copolymers
b. Miscellaneous polymers
c. Comb polymers.
40
2.5.2.1: Ethylene Copolymers
Different types of polyethylene polymers copolymerized with larger monomers are common and
effective wax inhibitors. These copolymers will consist of monomers of ethylene, and another
larger monomer, which will contain branches or side chain that can interfere with the wax
molecules in the crude oil, thereby preventing them from aligning together and form wax
deposits.Polyethylene polymers alone can crystallize together with wax molecules of similar
structure, and therefore have limited success preventing wax deposition (Kelland, 2009).
The most successful copolymer are ethylene/vinyl acetate (EVA). It is the percentage of vinyl
acetate in EVA, which determines how effective the polymer will inhibit the wax crystallization
growth. Increasing the vinyl acetate content will interrupt the crystallization process and lower
the WAT, and also makes the polymer more polar, which will enhance solubility. However too
much vinyl acetate will limit the co-crystallization with the wax, and cause a negative effect on
the wax inhibition (Lindeman and Allenson, 2005). The optimums vinyl acetate in EVA polymer
are around 25 – 30 percent (Machado et al., 2001).
Other ethylene copolymers used as wax inhibitors are ethylene/small alkene copolymer, such as
poly(ethylene butane) and poly(ethylene b-propylene) and ethylene/acrylonitrile copolymers
(Tinsley et al., 2007; Motz et al., 1990).
2.5.2.2: Miscellaneous polymers
The miscellaneous polymers are chemicals that have shown to be good wax inhibitors or have
synergetic effective with PPDs against wax deposition, which means their presence will enhance
the wax deposition removal. Some chemicals that have been used against wax deposition are
alkyl phenol-formaldehyde resin, long chain phosphoric ester surfactants with sodium aluminate,
41
and branched polymer, such as poly-ethyleneimine with pendant carbon 18 groups and
dendrimeric hyper branched poly esteramines with long alkyl chains (Martella and Jeruzelski,
1993; Gentili et al., 2005; Duncum et al., 2000; Van Bergen, et al., 2005).
2.5.2.3: Comb polymers
The comb polymers are regarded as the most effective wax inhibitor class today.These polymers
have got their name as a result of their chemical structure, because they look like a comb. The
comb polymer consists of polyvinyl backbone with different pendant chains, which are normally
long alkyl chains.
2.5.3: Pour Point Depressants (PPDs)
Pour point depressants (PPDs) or flow improvers being one of the chemicals that can be used as
wax inhibitors need to be designed in order to obtain optimum output during their application to
minimize the amount of wax deposit in the flow of crude oil in sub-cooled pipeline. Since the
composition of crude oil varies from oil field to oil field, and also at different period of crude oil
production for a particular field, there is need to optimize the wax inhibitors formulation for
different field and at different period of crude oil production.
2.5.4: Design of Pour Point Depressants
The structural designs of the PPDs normally contain three variable characteristic, which can
influence their performance, and should be chosen in relation to the wax structures in the specific
crude, and the following are considered (Manka and Ziegler, 2001; Soni et al., 2008):
42
a. The length on the polymers pendant chains is the most important variable, since the
interaction between the wax in the crude oil and the PPDs should be of similar length to
get the wax inhibition.
b. The polymer backbones have shown little effect on the performance of the wax PPDs,
but provide a structure for where the pendant chains are suspended from.
c. A polar part on the backbone will prevent the whole polymer to be co-crystallized with
alkane chains. This part on the polymer will prevent further agglomerating, and thereby
prevent the wax growth.
2.5.5: Acrylate Ester Polymer
Acrylate ester polymers (AEP) commonly referred to as comb polymers are regarded as most
effective wax inhibitor class used in the todays oil industry (Norland, 2012) , Having got their
names from their chemical structure, which look like a comb (Figure 7 ). The comb polymer
consists of a polyvinyl backbone with different pendant chains, which are normally long alkyl
chains.
43
CH3
H3C m Polymer backbone, where pendant chain are suspended from
O O Polar part, create hindrance in co-crystallization process
Paraffin like part, co-crystallize with wax crystals, the length of polymer
pendant chain is crucial, it should be of similar length as the wax structure
in the crude oil to get the best wax inhibition
CH3
Figure 7: The structural formulae of alkyl ester
2.5.6: Wax Deposit Optimization
Empirical or semi-empirical experimental models are capable of predicting response within the
range of experimental condition; hence there is no need for assumptions but quantification of
variables interactions in the models may be impossible. Response surface methodology central
composite design (RSMCCD) can be used to provide an avenue to overcome this inadequacy.
Response surface modeling (RSM) is an application of mathematical and statistical technique for
modeling and analysis of experimental responses influenced by several variables (Sabet et al.,
44
2012; Badwaik et al., 2012). It has been applied to various engineering models to study the
influences of variables interaction on the objective function in order to optimize the available
resources without compromising the targeted objectives (Lee et al., 2000)
Norland, (2012) used a response surface methodology (RSM) known as design of experiment
method to determine the best formulation for a group of acrylate ester polymers (AEP). Design
expert software was used to perform the two separate designs of experiment analyses on two
different synthesized mixture of group of AEP. However, none of the optimal formulation result
corresponds well to the generated response surface graph results from the used software.
Salam, et al., (2014) investigated the influence of wax deposition operating parameters on weight
of wax deposit in oil pipeline using RSM. The results showed that the developed model can be
used to simulate wax deposition within the range of variables studied. However, apart from the
fact that some of the studied variables are not completely independent, the effect of wax
inhibitors and wax content on wax deposition in crude oil flow was not studied as only one
sample of crude oil was used.
2.6: Mathematical Modeling of Wax Deposition
There is an agreement among researchers in flow assurance study that the major wax deposition
mechanisms in cooled pipelines are:
a. Molecular diffusion (MD)
b. Shear dispersion (SD)
c. Shear Stripping reduction (SR)
d. Brownian diffusion (BD)
e. Gravity settling (GS)
45
Mechanism such as shear dispersion, Brownian diffusion, and gravity settling had been identified
to play a role only for particulate deposition of wax. However, most models studied in the
literature neglected the effect of particulate deposition as they believe its effect is
inconsequential (Wang et al., 2015). The Table 1 shows an overview of the reviewed wax
deposition models.
Table 1: Highlights of considered mechanisms by different models.
S/N Model Published year Mechanisms considered

1 Burger et al. 1981 MD, SD
2 Majeed et al. 1990 MD
3 Svendsen 1993 MD
4 Brown et al. 1994 MD, SD
5 Hsu et al. 1995, 1998 MD
6 Rygg et al. 1998 MD
7 Creek et al. 1999 MD
8 Singh et al. 2000, 2001. MD, ID, gelation
9 MSI 2000 MD, Kinetic crystallization.
10 Solaimany et al. 2001 MD, SR
11 Banki et al. 2002 MD
12 Lindeloff 2002 MD, SD
13 Azevedo and Teixeira 2003 MD, SD
14 Fasano et al. 2004 MD
15 Ramirez et al. 2004 MD, SR, ID
16 Hernandez et al. 2004 MD, SR, ID
17 Rosvold 2006 MD, SD, SR, ID, PD, and Inertia.
18 Edmonds et al. 2008 MD,SR
19 Banki et al. 2008 MD, SD, Thermal Diffusion, aging
20 Botne 2011 MD,SD
21 Adeyanju and Oyekunle 2011 MD, SD, SR
22 Kjorass 2012 MD, SD, ID
23 Wang et al. 2014 MD, SR, Thermal diffusion
24 He and Huang 2014 MD, SD, SR, Thermal diffusion
M.D. = molecular diffusion, S.D. = Shear dispersion, I.D. = Internal diffusion, S.R. = Shear
removal, P.D. = Particle diffusion.
46
Wax deposition mathematical model can be categorized into four groups. These are:
i. Rygg, Ryddah and Ronningsen (RRR) model based on molecular diffusion and shear
dispersion (Rygg et al., 1998; Lindeloff, 2002; Azevedo, 2003; Fasano et al., 2004).
ii. The Matzain’s model based on molecular diffusion, shear dispersion and shear stripping
(Matzain et al., 2003; Ramirez et al., 2004; Hernandez et al., 2004; Rosvold, 2006; Adeyanju
and Oyekunle, 2014).
iii. The Hydro model based on molecular diffusion and shear removal described by asymptotic
fouling theory (Singh et al., 2000, 2001; Solaimany et al., 2001; Alana, 2003).
iv. The University of Michingan model based on molecular diffusion and deposits aging
(Venkatesan, 2004; Gjermundsen et al., 2006).
2.6.1: Rygg, Ryddah and Ronningsen (RRR) Model
The RRR model is multi-phase wax deposition model applicable to non laminar flow. It
consistsof standard steady state multi-phase point model for the prediction of pressure drop and
liquid hold up along the pipeline. The effect of deposition or pressure drop and temperature is
calculated by integration in time. The multi-component wax model continuously estimates the
wax precipitation along the pipeline and the viscosity of the composition. The wax deposition is
then estimated from the diffusion of wax from the bulk towards the surface of the pipeline, due
to temperature gradients and shear dispersion effect. The inner pipe wall friction is varied due to
wax deposition (Rygg et al., 1998).
The RRR wax deposition model is divided into separate sub-models. Which include:
a. Flow model that calculates pressure drop and flow regime.
47
b. Thermodynamic Wax Model (TWM) that determines the number and properties of the
different phases for each section.
c. Viscosity model that calculates the viscosity.
d. Wax deposition model that predicts the amount of wax that deposits in a section of the
pipeline.
Deposition in the RRR model is based on molecular diffusion and shear dispersion, both
mechanisms which enhance the wax deposition. The volume rate of wax deposition by molecular
diffusion for a wax forming composition is derived from
diff
NWAX
 
D cib  ciw S wet
Vol wax  i 1  i
2rL
(1)
Where c ib , ciw is the model concentration of the wax component i dissolved in the oil phase in the
bulk oil and at the wall respectively (mole/m3). Swet is the fraction of the wetted circumference.
NWAX is the number of wax components, MWi is the molar weight of wax component,i
(kg/mole), ρ is the density of wax component,i (kg/m3), r is the current inner pipe radius (m), L is
the length of pipe section (m), D is the diffusion coefficient. The Hayduk-Minhas correlation
(m2/s) (Hayduk and Minhas, 1980) is used to calculate the diffusion coefficient.  is the
thickness of the laminar sub layer (m) and Blasius equation (Blasius, 1982) is used when
calculating the thickness of the sub-layer in turbulent flow.
Also the volume rate of wax deposited from shear dispersion can be estimated from a correlation
by Burgler et al., (1981).
shear k *C wall A
Vol wax  (2)
 wax
48
Where k* is the shear deposition rate constant (kg/m2). Cwall is the volume fraction of the
precipitated wax in the oil at the inner wall temperature.  is the shear rate at the wall (s-1). A is
the surface area available for deposition (m2) and  wax is the average wax density (kg/m3).
When accounting for both mechanisms, the total rate of increase in thickness for wax layer is
given as
diff shear
Vol wax  Vol wax
I wax 
1   2rL (3)
Where ϕ is the porosity of the deposited wax. The is usually assumed to be in the range from 0.6-
0.9.
it should be noticed that no removal mechanisms are discussed in this model. Therefore, the
main drawback concerning the RRR model is the lack of shear incorporated into the model.
Therefore there is tendency for the model to give good prediction for low flow rates and when
wax starts to grow at a clean pipe, but after some time the results become less reliable due to no
release mechanism included in the model.
2.6.2: The Matzain’s Model
The Matzain model originated from the University of Tulsa Joint IndustryProject (JIP). One of
the aims for the JIP was to develop a computer program for the prediction of wax deposition in
multi-phase flow-lines and wellbores. Therefore, extensive experimental tests were conducted
for measuring wax deposition during multi- and single phase oil, (Matzain et al., 2001).
South Pelto oil from the Gulf of Mexico and Natural gas from Oklahoma Natural Gas Company
were used for the experiments (Gjermundsen, 2003). Deposition tests performed by Matzain et
49
al.,(2001) for the two phase flow showed that wax deposition depends on flow pattern. At low
mixture velocities, the trend for deposition build-up is similar to observations in laminar single-
phase flow tests. The build up trend at high mixture velocities is more similar to observations for
the turbulent single-phase flow.
The semi-empirical kinetic model predicts the wax thickness. The accuracy is acceptable,
especially at high flow rates. The Matzain and the RRR model were based on mechanisms as
molecular diffusion and shear dispersion, but shear dispersion is found to be of minor importance
(Gjermundsen, 2006). Matzain et al., (2001) proposed that the rate may be influenced by other
mechanisms as well. Shear stripping will result in a reduced deposition rate, and rate
enhancement occurs due to entrapment of oil and other mechanisms. Since Fick’s mass diffusion
theory does not account for such influences of the rate. Matzain et al. (2001) tried to take this
into account.
d 1  dw dT 
 Dow  w 
dt 1 2  dT dr  (4)
Where  is the thickness of wax layer deposited on the wall (m). ww is the concentration of wax
in solution (weight %), r is the radial distance (m) and T is the temperature (oC),  1 is the
empirical relation for the rate enhancement due to oil being trapped in the deposited wax layer.
The relation also accounts for any positive deposition rate which is not accounted for by the
diffusion constant, Dow as the turbulent mass diffusion.  2 is the empirical relation for the rate
reduction due to shear stripping.
for this by empirically modifying Fick’s law. Based on experiment, two empirical correction
terms are added.
50
The total wax deposition rate is then expressed as (Matzain et al., 2001)
The empirical relations follow as:
C1
1 
1  C oil / 100 (5)
C3
 2  C 2 N SR (6)
The diffusion constant is given by Wilke and Chang (1955) correlation, but this diffusion
correlation is not sufficient to represent the proportionality constant that drives the diffusion
process. C1, C2, and C3 are three empirical constants which are correlated from the single phase
and two phase flow data. Their values were found to be: C1 = 15.0, C2 = 0.055 and C3 = 1.4.
Additionally Coil is the percentage of oil trapped in wax deposit (%) and is expressed as
0.15
 N RE ,f

C oil 
 100 1  
 8 
  (7)
 v sl 
o  d w
N RE , f   E
o (8)
Where ρo is the oil density (kg/m3), vsl is the liquid superficial velocity (m/s), E is the liquid hold
up , dw is the inside diameter as a result of wax build-up (m) and μo is the oil viscosity (kg/(m.s,
NSR is a dimensionless variable expressed in the form of a flow regime dependent Reynolds
number. The variable was derived for each flow pattern
51
 o vo
Single phase N SR 
o (9)
 v sl 
m  
Intermittent/Bubbly N SR   E  (10)
o
 v sl 
m o  
Annular N SR  E
o (11)
 v sl 
o  
Stratified Smooth/Wavy N SR  E
o (12)
Where vo is the oil velocity (m/s),  is the thickness of the wax layer, (m) and ρm is the average
density of the gas-oil mixture (kg/m3) (Matzain et al., 2001).
The molar deposition rate for a wax forming component i is given by (Hoveden, 2004).
 1 Dwi (Cib  Ciw ) dT

ndep  .
1 2 Tb  Tw dr (13)
The thermal gradient of the laminar sub-layer is found by (Hoveden, 2004)
dT Tb  Tw 
 .hwall
dr o (14)
Where o is the oil thermal conductivity (W/(mK)). Tb is the bulk average flow temperature (K),
Tw is the inner wall surface temperature (K) and hwall is the inner wall surface heat transfer
coefficient (W/(m2K)).
52
The weakness with the Matzain’s model is that it does not include much physics and there are
several constants. Also all experiments in which the model is based were carried out by only
using South Pelto oil, hence there is need to test the model with other crude oils. Also nobody
knows how well the Matzain model works with pipes of different diameter, e.g a full size
production pipeline.
2.6.3: Hydro Model
The Hydro in-house wax deposition model was developed at Norsk Hydro by Sontvedt early in
the nineties. Previously this deposition model was implemented in an Excel spreadsheet, solved
analytically and based on a single phase homogeneous flow. Today a Fortran language solves the
governing equation numerically. The model is implemented in an in house multi phase flow and
heat transfer model(Gjermundsen, 2003).
The model estimates the amount of deposited wax over a time period inside a pipe, and it takes
into account that deposition during time shows an asymptotic growth due to the stressfrom fluid
flow. The model is based on theoryof asymptotic fouling, which is known from the heat
exchange theory.Previous experiments performed at the wax rig in Porsgrunn used three
different crude oils. The Hydro model is based on these Porsgrunn experiments and additional
experiments performed at the Joint Industry Programme (JIP) at the University of Tulsa. This
makes the model more robust when extrapolating to other crude oils.(Gjermundsen, 2003).
Based on theseand other experiments, Sontvedt suggested that wax deposition builds up to an
asymptotic level with time. It was stated that the growth of the wax deposit will stop, even with a
significant temperature gradient, and even though the temperature of the bulk oil is below WAT.
This asymptotic growth cannot be described based on molecular diffusion. It was also observed
53
that shear stripping only occurred at extreme shear and therefore the mechanism is not included
in the model. (Gjermundsen, 2006).
When deposit accumulates the flow area isreduced and the shear forces are increased. The
asymptote is reached when the deposit thickness reaches a value at which the strength of the
outer deposit layer can not withstand the increased shear forces. At this point the deposit build-
up is balanced by the removal.When the temperature at the deposition-bulk flow interface is too
low for the wax crystals to have any strength to withstand the shear stress (that is when
deposition reaches an asymptotic height). Such temperature is referred to as gelation
temperature. Sontvedt suggested that deposition stops when the deposition-bulk flow interface
temperature equals gelation temperature, which is below WAT.
The governing equation for growth of the deposition layer is expressed as

 deposition rate  rate of removal of deposition (15)
t
which means
d 1 dC dT f 2
 Dwo  
dt 1   dT dr f1 (16)
The first term in the equation describes the deposition, ϕ is the porosity of the deposit, and it
modeled as constant over time and is influenced by wall shear stress (τ).
  78.5. 0.13 (17)
54
The second term of the equation describes the removal of deposit by two functions. f1 describes
the relation between the deposit-bulk oil interface temperature (T) and the strength of the wax
filament (wax crystals) in the wax deposit.
f1  2 *10 7 T 4.6868 (18)
f2 describes the stress on the same wax filament on the deposit surface, and it is a function of the
wall shear from the bulk flow. The asymptotic height is reached when f 2  f 1 and at that point
there will not be more growth of the deposit.
    pa   0.93 
  wall    wall  3Pa 
f 2    3 pa  
 
1   wall  3Pa 
  (19)
Where Pa is the pressure in Pascal and τwall is the shear stress at the pipe wall
2.6.4: University of Michigan Model
The University of Michigan also has a JIP. This JIP is smaller than the wax JIP at the University
of Tulsa, but it has a more scientific approach regarding modeling. Experimental works
performed at the University of Michigan confirmed that the particulate deposition such as shear
dispersion, Brownian diffusion and gravity settling are not significant for the flow conditions in a
pipeline. The research group chose to put their further focus on studying gelation along with the
molecular diffusion mechanism (Singh et al., 2001).
First step of the deposition process considered the formation of a deposit at the wall. Then the
model considered the internal diffusion mechanisms of waxy gel deposits. Beside considering
55
the inward diffusion of wax forming components towards the pipe wall,the counter-diffusion of
oil molecules out of the wax deposits was also considered. The wax fraction in the deposit
increases with time, and this phenomenon is called aging. The aging was enhanced due to higher
flow rate, which resulted in wax fraction of 60-70% in the deposit (Singh et al., 2001).
The Michigan model is the only model including the aging of the deposit. Most models state that
the wax-oil deposit has constant wax content, and the wax porosity is used as an adjustable
parameter. However experiment by Singh et al., (2001) proves that the assumption stating that
the composition of the deposit is time invariant to be invalid. (Singh et al., 2001)
A coupled system of differential equations is needed to describe the growth and the aging of the
deposit under laminar/ low shear conditions.
dC
Dow 1   x 
d dr i

dt x (20)
dC
Dow  x 2R   
dx dr i
 (21)
dt  2 R   
Where  is the deposit thickness, x is the wax content of the deposit, R is the radius of the clean
pipe and Dow is the diffusion coefficient for the wax in oil, ϕ is the coefficient describing
diffusion in a porous network and can be found by Cussler (Cussler, 1975) as:
1
 x  
 x2 
1   2  
 1  x  (22)
56
Where α is the average aspect ratio of the wax crystals. The model was developed by Singhet al.
(2000). They observed from experiments that an increase in the wall temperature resulted in the
decrease in gel thickness, an increase wax content in the wall layer and a decrease in wax content
of the interface layer. In addition, increasing flow rate lead toincreased wax content of both
layers. The model proved to be excellent agreement with the experiment.
Venkatesan (2004) extended the model for turbulent flow by adding a shear flux term as.
J s  m. n
(23)
dC
Dow 1   x   J s
d dr i

dt x (24)
Both parameter m and n were fitted from experiments with the model fluid in a test rig,but these
values were only validated for specific fluid and test geometry. and ρ is the density of wax
formed. This is the only model which takes precipitation kinetics into consideration. But the
model has not been tested for different type of crude oils.It is also not available as software for
the industry.
Advantages and disadvantages of the four classes of deposition models are tabulated in Table 2.
57
Table 2: Analyses of different models advantages and dis-advantages
S/N Model Advantages Disadvantages

1 RRR 1. Deposition is based on: 1. Not applicable for laminar
Model a. Molecular diffusion flow.
b. Shear dispersion. 2. The viscosity model neglects
2. Pure textbook approach as shear effects assuming
diffusion constant are given by Newtonian behavior
Hayduk-Minhas correlation 3. No removal mechanisms
(1982). discussed.
3. Implemented in OLGA 4. Lack of sufficient data basis
to estimate the wax porosity
and roughness accurately.
5. Assuming tuning parameters:
a. Wax porosity.
b. Roughness.
6. Need more data for effect of
multi-phase regime.
2 Matzain 1. Semi empirical kinetic model. 1. Based only on South Pelto
model 2. Predicted wax thickness with oil.
an acceptable accuracy, 2. The empirical constants
especially at high oil flow rate. incorporated was taken as
3. Fully compositional. constants:
4. Diffusion constant given by a. C1 = 15.0 (Enhance
the Wilke-Chang correlation the wax deposition)
(1955). b. C2 = 0.055
5. Deposition is based on: c. C3 = 1.4
a. Molecular diffusion Whereas in real situation C2 and C3
and shear dispersion. varies with time.
b. Shear stripping. 3. Does not incorporate the
6. Through the use of effect of shearing of wax
dimensionless variable and deposition, seen in low rate
empirical constant from annular flow.
experiments, It accounts for 4. It is a fitted model that does
what Fick’s law does not not include much physics and
account for, like influences in has many artificial tuning
deposition rate as: constants.
a. Reduced rate due to 5. Uncertain if the models
shear stripping. works for pipe of different
b. Rate enhancement due diameters i.e. production
to entrapment of oil pipelines
and other mechanisms.
7. The effect of flow patterns was
taken into consideration.
8. Implemented in OLGA.
58
S/N Model Advantages Disadvantages
3 Hydro 1. Takes into account the 1. Closure relationship, φ, f1, f2
Model asymptotic growth of the determined from
deposit experiments.
2. Based on results from: 2. Not fully compositional.
a. Several experiments 3. Effect of shear stripping not
(Porsgram and the incorporated.
Univ. of Tulsa) 4. Effect of shear dispersion not
b. Different types of oil incorporated.
(This makes the model
to be more robust when
extrapolating to other
oils.
3. Not based on tuning the shear
coefficient for shear removal
as auto-retardation
mechanisms are described
mechanistically.
4. Deposition porosity was
described mechanistically and
is, therefore scalable.
4 Univ. of 1. The only model that includes 1. Effect of shear stripping not
Michigan aging (the fact that wax incorporated.
fraction in the deposit 2. Effects of shear dispersion
increases with time). and stripping not
2. For laminar flow (by Singh incorporated.
(2000), but extended for 3. The Model is not
turbulent flow (by Venkatesan implemented in any software,
(2004)) hence not fully accessible for
3. The only model taking industry use.
precipitation kinetics into
account.
2.7: Wax Deposit Thickness Determination The pressure drop method was used to determine
the thickness of the deposited wax (Chen et al., 1997).
Once the frictional pressure drop across a pipe section is measured and the flow rate, density and
viscosity of the crude oil in the pipe section are determined, the wax thickness present in the pipe
wall can be calculated accurately from the following equation:
59
n 2n
5 n 2cL     4Q 
(d i  2 w )     
Pf      
(25)
Where Pf is the pressure drop
L is the length of pipe section, d is the hydraulic diameter or effective inside diameter, δw is the
wax thickness, Q is the volumetric flow rate, ρ is the fluid density, Where  is the apparent
viscosity of the crude oil. c = 16, n = 1 for laminar flow and c = 0.046, n = 0.2 for turbulent
flow. Laminar flow exists when N Re  2000 (Chen et al. ,1997).
2.8: Evaluation of Wax InhibitorPerformance
The percentage wax inhibition, WI (%), is defined by the ratio of the difference of wax
deposition rate with and without wax inhibitor to that of the blank oil at a specific temperature.
It is expressed as:
Wb  Wwi
Percentage Wax inhibition ,WI (%)  100 (26)
Wb
Where Wb= Wax deposit by volume of the blank oil (cm3) and
Wwi = Wax deposit by volume during the same period of time of inhibitor/chemically
treated oil (cm3).
A positive value of percentage wax inhibition means the chemical tested inhibits wax
deposition. A negative value means the chemical increases wax deposition. An ideal percentage
wax inhibition is 100% wax inhibition in the operating temperature range. However, obtaining
100% wax inhibition by chemical treatment is difficult and may not be economical due to cost
and required dosage.
The wax inhibition rate varies with the sea water temperature (4 – 32oC in Gulf of Guinea),
60
therefore an approximate integration method for calculating the average wax inhibition
performance should be based upon overall wax inhibition in the subsea environment.
The proposed overall wax inhibition percent is defined as the area integrated from the curve of
wax inhibition against temperature divided by the covered temperature range and from
trapezoidal rule; this is expressed (Hsu, 1995) as:
WI % overall 

50 WI T o

 2 WI T
1
 WI T 2
 .......... .WI T n 1
 WI  
Tn
( 27 )
Tn  T0
Where: (WI)Ti = Percentage Wax Inhibition (%) at temperature Ti
Ti = Coolant temperature, oC at period i.
61
CHAPTER THREE
3.0: EXPERIMENTAL METHODOLOGY
In order to achieve the objectives and the scope of the research, experimental and analytical
procedures used in this study are presented and discussed with more details in the following
sections.
In this study, seven crude oil samples from different floating, production ,storage and
overloading (FPSO) storage tank in different subsea fields in the crude oil producing region of
Nigeria characterized by their high wax content and high gel point temperature were selected as
the experimental samples. Different mechanisms leading to wax deposition such as molecular
diffusion, shear dispersion and shear removal were physically modeled in the flow loop
equipment, and the effects of key influence factors on wax deposition in straight pipelines: inlet
oil temperature, inlet coolant temperature, oil flow rate, percentage of wax inhibitors and the oil
wax content at condition operating in the Nigeria subsea environment (Part of Gulf of Guinea)
were determined using the fabricated flow loop equipment.
3.1: General Chemicals
The chemicals required for the experimental study were obtained from various suppliers such as
methyl methacrylate (95% purity), butyl acrylate (97% purity) and acrylic acid (94% purity) ,
Xylene (98% purity) and different types of acrylate ester co-polymers (97% purity).
62
3.1.1: Chemical Demulsifiers
The demulsifiers used for the study: methyl methacrylate, butyl acrylate and acrylic acid were
selected on the basis of their solubility in xylene.
3.1.2: Crude Oil Samples
Seven crude oil samples (labeled A-G) from different subsea fields in the Niger Delta of Nigeria
were used for the study. The crude oils were taken from the stock tank/FPSO in which the crude
oils were stored. Samples A was used for the demulsifier screening. This is because when
agitated, it forms one of the most stable emulsions with the used wax inhibitors. The other oil
samples were used to confirm the effectiveness of each of the demulsifier used. All crude oil
samples were provided by Total Elf Incorporated. A 5-liter volume of crude oil was used for
each of the experiment.
3.2: Experimental Facility (Rig)
The experimental setup consisted of a flow loop shown in Figure 8 and schematic diagram as
shown in Figure 9. This flow loop was used to perform the wax deposition experiments in
straight pipeline under the single oil phase conditions.
This flow loop is made of mild steel pipe of length 30 cm with an internal diameter of 2.0cm.
The experimental setup has two sections; the test section/tube (A) and the reference section/tube
(B). The crude oil temperature was regulated with a temperature regulator, (R), the oil is pumped
through the test section (A) and the then through reference section (B) after passing through the
liquid mass flow-meter along the flow lines. The test section was jacketed with a steel jacket in
which cold water pumped from a cooling bath was circulated. The purpose of the test section is
to maintain the inner pipe wall at a lower temperature than both the bulk oil temperature and
63
Wax Appearance Temperature (WAT) so as to generate the wax deposit on the inner pipe wall
(just like what would be encountered in actual pipelines). The configuration of the reference
section is completely identical with the test section. However, contrary to the test section, the
inner pipe wall temperature in the reference section was maintained at a higher temperature than
the bulk oil temperature to prevent wax deposition by circulating the heated water through the
jacket of reference section. Thermocouples (T) were placed both at the inlet and outlet of the test
tube and the reference tube to determine the temperatures at each point. Thermocouples were
also attached to the cooling water tank and crude oil tank to take temperature reading. The wax
deposit in the previous experimental run was ensured to be removed by flowing hot oil for few
hours before the commencement of subsequent experimental run.
The optimum conditions of wax deposit inhibition process were determined by using Factorial
Design Optimization. The optimum conditions were confirmed with experimental runs using the
flow loop apparatus.
3.3: Experimental Section
A dynamic flow loop rig (Figure 8) whose schematic diagram is as shown in Figure 9was used to
evaluate the effectiveness of the pour point depressants (PPDs) to inhibit wax deposition during
the flow of the crude oils samples (A-G) in cooled pipeline at coolant temperatures of 4oC to
32oC. The bulk crude-oil temperature was kept at 55oC for all the experimental runs. The
deposition tests were carried out in two phases:
(i) Wax inhibition evaluation with different quantities (0 – 35ml) of pour point depressants
(PPDs) at fixed coolant condition (15oC).
64
(ii) Wax inhibition evaluation with fixed quantity (5ml, 12ml) of pour point depressants
(PPDs) at varying coolant conditions (8 – 22oC).
3.4: Analytical Techniques
Before the start of each wax deposition process (cooling process), the oil was circulated at about
2 liters/min for 25 minutes to stabilize the flow conditions. The temperature range of the coolant
used in the experiments was between 4oC and 32oC.
An identical wax deposition flow processes were then conducted for the chemical-mixed oil.
After completion of the wax deposition test with each PPD, the test tubes were purged with
xylene, then the wax deposit was pigged and recovered for volume measurement. However, wax
deposition rate and wax thickness are calculated using the techniques proposed by Hsu et al.,
(1994).
In summary, Seven crude oil samples (five light grades (APIo> 30) and two medium grades (20 <
APIo< 30)) collected from different subsea fields in Niger delta area of Nigeria were used.
Selection of effective wax inhibitors (WI) were restricted to mixtures of different chain length of
acrylate ester polymers (AEP) diluted with xylene. Demulsifier was used as additive to enhance
the inhibitive capability of the wax inhibitors. The three tested additives are oil soluble
demulsifiers: butyl-acrylate, acrylic acid, and methyl methacrylate. Pour point depression ability
of the mixtures was used as the determining factors in selecting effective wax inhibitors.
Pressure drop method was used to determine the amount of wax deposit. Regression models
developed are based on the results of wax deposition experiments performed in the fabricated
physical model (Flow loop equipment).
65
Figure8:: Laboratory wax deposition test flow loop rig.
66
Figure9: Schematic diagram of wax deposition test flow loop
3.5: Pour PointsDetermination
The ASTM D2500 standard test method was used to determine the pour points of the following:
(a) The seven crude oil samples (A – G)
(b) Each of the crude oil samples diluted with 20, 40 and 60% by volume of aliphatic
solvents: Kerosene and Premium Motor Spirit (PMS)
67
(c) Samples of each of the crude-oil diluted with 1, 3 and 5% by volume of aromatic
solvents
The Pour point is defined as the temperature at which the crude oil sample will no longer flow
when held in horizontal position in a test jar for about 5 seconds, due to formation of a wax gel
network. Pour point is an indicator of the gelling potential of the crude oil sample. Its
measurement is believed to be affected by factors such as wax content, wax crystal size and
number, pressure, and solution gas, as well as the thermal history of the crude oil (Alboudwarej
et al., 2006).
The pour point equipment used in this work is basically a 3-bath bench model 334 from Lawler
Manufacturing Corporation. The equipment has been designed for performing manual cloud
point and pour point tests in conformance with ASTM D2500 and ASTM D97 standard test
methods. Each bath has four (4) test ports machined to accept four (4) test jars. The baths are
preset at different temperatures: 0oC (32oF), -18oC (-0.4oF), and -33oC (-27.4oF).
When the equipment is powered, the baths preset; the actual temperatures are displayed on
digital temperature indicator for each bath. Each bath will then cool to the preset temperature via
the refrigerating system. The pour points were measured by the following procedures.
The three baths are utilized during the test as the test jar is moved around, starting from the
highest preset temperature bath. The test jar is placed first in the 0oC (32oF) bath, and the test
thermometer in the test jar is monitored. If the oil still flows when the test thermometer reads 9oC
(48.2oF), it is transferred to the -18oC (-0.4oF) bath. The test jar is finally transferred to the -33oC
(-27.4oF) bath if the oil still moves when the test thermometer reads -6oC (21.2oF). the following
68
outlined the procedure for accurate pour point measurement using Lawler Model 334 procedure
using ASTM D97 test procedures guidelines.
3.6: Wax Content Determination
The mass of wax in the crude oil samples were determined using (Universal Oil Products LLC)
solvent extraction method UOP46 . The wax content is calculated as follows:
100W
Wax content  (28)
S
S = sample mass, g; W = wax mass, g
3.7: Wax Deposition Experimental Methods
The schematic diagram of the rig used for the wax deposition experiment is shown in Figure7.
The waxy oil sample was made to enter the test section at a relatively higher temperature (55 –
60oC) than the wall/coolant temperature (4 – 22oC) in order to generate wax deposit in the inner
section of the flow-line. The pressure differential method was used in evaluating the volume of
wax deposited in the simulated pipeline (Chen et al., 1997)
After each wax deposition test with the blank oil, the waxy oil was heated up to 65oC, and passed
through the pipe at a rate of 4 liters/min. to remove wax deposit and oil thermal history. Then
different quantities (1-5%) of pour point depressants (PPDs) sample were injected into the oil
reservoir at 65oC. The oil/pour point depressants (PPDs) mixtures were circulated at a volumetric
rate of 7 liters/min. for at least 20 minutes at 65oC. Then the oil/PPD mixtures were cooled to
45oC for this study. Before the start of wax deposition process (cooling process) the oil was
circulated at about 2 liters/min for 25 minutes to stabilize the flow conditions. The temperature
69
range of the coolant used in the experiments ranges from 4oC to 32oC to simulate the conditions
operating in the Gulf of Guinea’s subsea environments.Identical wax deposition flow processes
were further conducted for the chemical-mixed oil. After completion of the wax deposition test
with each PPD, the pipe test sections were purged with xylene, then the wax deposit was pigged
and recovered for volume measurement. However, wax deposition rate and wax thickness were
calculated using the pressure differential techniques proposed by Hsu et al., (1994).
3.8: Sensitivity Analyses of Different Wax Deposition Effluent Variables
Experiments were performed using the flow loop device (Figure6) on blank crude oil A, under
the following conditions:
Table 3: Values of wax deposition effluent factors used in sensitivity analyses
S/N Flow rate Coolant Bulk Oil temperature Flow Period

(liter/min) Temperature (oC) (oC) (hours)
1 0.5 30, 35, 40 55 11
2 0.5 37 48, 52, 56 7.5
3 0.5 28 34,37,40 7.5
4 1.0 20 55 2, 4, 6, 8
5 0.3, 0.5, 0.7, 1.2 37 56 5.5
6 0.5 37 37 5.5
3.9: Wax Appearance Temperature (WAT)
Wax Appearance Temperature (WAT) was evaluated using viscometer techniques, density was
determined using Mettler-Paar digital density meter (DMA-45) with connected Brookfield
70
heating/refrigerating bath for temperature control, while the viscosities were determined using
Brookfield Viscometer model LVDV-II+.
The flowchart/summary of the research study is as shown in Figure 10:
Characterization of each of the seven crude Oil Samples
Sensitivity analyses of the effect of wax deposition effluent variables: coolant

temperature, bulk oil temperature, flow rate and crude oils wax content on wax
deposition rate using the flow loop apparatus
Analyses of chemical compositions of each

71 of the crude oil samples and the
resulting wax deposit scooped from the test section of the flow loop apparatus
Experimental studies of the effectiveness of some petroleum fractions
and polymeric compounds as pour point depressants
Experimental investigation of the effectiveness of the above solvents as wax inhibitor
Experimental studies of the capability of different demulsifiers as wax inhibitors enhancer
Optimization of the wax inhibitor

Optimization of the wax
formulation for each of the crude oil
inhibitor/demulsifier formulation for each
sample using RSMCCD and
of the crude oil sample using RSMCCD
validation with experimental data
and validation with experimental data
Formulation of the optimal wax inhibitor/demulsifier formulation for each of the crude oil
sample for minimum wax deposition.
For each crude oil sample, the RSMCCD method of variables optimization was used to
model the dependence of the operating parameters (temperature differential, wax
inhibitors/demulsifier fraction, flow rate) on wax deposition rate, considering the laminar
and turbulent flow regimes.
The RSMCCD method was used to model and optimize the variation of the wax deposition
rate with the operating parameters (temperatures, wax inhibitors/demulsifier fraction, flow
rate and wax content) and validation of the results with experimental data was done.
Figure10: The flow chart of the procedure for the research study
The research in this dissertation elucidates the fundamental understanding of the processes and
problems in the transportation of waxy crude oil from the production platforms to the export
terminals, improve on the current efforts in the formulation of wax inhibitors, and also response
surface methodology central composite design (RSMCCD), a form of statistical and
72
mathematical method was used to determine the optimum conditions for wax deposition
processes using experimental wax deposition data from flow loop experiments. Seven crude oil
samples from different fields in the Gulf of Guinea were used in the study.
3.10: Process Optimization
Since the study involved the optimization of a process, using more than one variable required the
performance of so many experiments which could be of high cost and time consuming.A
software “Minitab 16’ was used in this study in the design of experiments (DOE). The software
designs experiments in such a way to obtain the greatest amount of information with minimum
number of experiments, thereby minimizes cost of materials and time required. Response surface
methodology central composite design (RSMCCD) software was used in this study to:
i. Optimize wax inhibitors formulation using two chemicals: group of acrylate ester
polymer (AEP) and xylene (two components).
ii. Design an improved optimum wax inhibitors formulation using three different chemicals
(introduction of a demulsifier (butyl acrylate) as the third component).
3.11: Response Surface Methodology (RSM)
Response surface method provides a means of selecting few experiments in a design space that
efficiently represent all possible experiments.
Response surface methodology (RSM) is a combination of mathematical and statistical
techniques used to relate the independent variables (input) to the response of interest (output).
Such a relationship though unknown but can be approximated by a low degree polynomial model
(Khuri and Mukhopadhyay, 2010) of the form:
73
y  f '  x    (29)
Where x   x1 , x2 ,................xk ' , f  x  i s a vector function of m elements that consists of powers
of x1 , x2 ,.............xk up to a certain degree denoted by d  1 ,  is a vector of m unknown
constant coefficient referred to as parameters,  is a random experimental errorassumed to have a
zero mean
Two important models are commonly used in RSM. These include the first-degree model (d = 1),
k
y   0    i xi  
i 1 (30)
And the second-degree model (d = 2)
k k
y   0    i xi  i  j   ij xi x j    ii xi2  
i 1 i 1 (31)
The RSM enables one to: (Khuri and Mukhopadhyay, 2010)
1. Establish a relationship between y and x1 , x2 ,................., xk ) that can be used to predict
response values for given settling of the control variables.
2. Determine, through hypothesis testing, significance of the factors whose levels are
represented by x1 , x2 ,...................xk .
3. Determine the optimum settlings of x1 , x2 ,........................xn that result in the maximum
(or minimum) response over a certain region of interest
3.12: The Central Composite Design (CCD)
74
The central composite design is the most widely used of all the second order design. The CCD
consists of the following three properties:
1. A complete 2k factorial design, whose factors levels are coded as -2, -1, 1, 2. This is
called the factorial portions.
2. An axial portion consisting of 2k points arranged so that two points are chosen on the
axis of each control variable at a distance of  from the design center (chosen as the
point at the origin of the coordinates system).
3. no, center points.
Thus, the total number of design points in a CCD is n = 2k + 2k + no
3.13: Statistical analysis
Each response of the wax deposit inhibition process was used to develop a mathematical model
that correlates the volume of wax deposit to the wax deposit variables studied through first,
second order and interaction terms according to the following second order polynomial equation:
V   0   1Tc   2Tb   3 I w   4 Q   11Tc2   22T b2 33 I w2 (32)

  44 Q 2   12TcTb   13Tc I w   14Tb Q   23Tb I w   24Tb Q   34 I w Q
Where V represents the predicted response (volume of deposited wax, which equals the wax
deposition rate multiplied by time), Tc is the cooling water temperature in Kelvin, Tb is the bulk
oil temperature in Kelvin, Iw is the percentage of wax inhibitor, WI (group of AEP/xylene/butyl
acrylate mixture) in the crude oil sample, and Q the oil/WI flow rate: α12, α13, α14,, α23, α24 and α34
are interaction coefficients; and α11, α22, α33 and α44 are quadratic coefficients. The results and the
75
second-order polynomial were analyzed using analysis of variance (ANOVA) by Minitab 16
software. The model terms were selected or rejected based on the probability (P) value with 95%
confidence level. The quality of the fit of the polynomial model equation was expressed by the
coefficient of determination (R-square) and adjusted R-square and ‘adequate precision’. Three-
dimensional (3D) plots and their respective contour plots were obtained based on the effect of
the level of the two factors. From these 3D plots the simultaneous interaction of the two factors
on the responses was studied. The ‘point optimization’ process was employed to optimize the
level of each factor for maximum response (Salam et al., 2014).
3.14: Determination of Optimum Conditions
Part of the objectives of Response surface methodology is the determination of the optimum
combinations of the independent variables that produce maximum (or minimum) response over
the design space. Hence, for the optimum variables combination to be accepted, there is need for
the developed model to represent the relationship between the independent variables and the
response to an acceptable degree. Such models are used after some set of experiments have been
performed that enable the identification of a region that contain the location of optimum
response.
3.14.1: Optimization of Second-Degree Model
A typical second-degree model is given as:
76
y  X   (33)
Where y  ( y1 , y2 ,.............. yn )' , X is a matrix of order n x p whose uth row is f '  xu  , similarly
y   o  x'  *  x' Bx   (34)
Where x   x1 , x2 ,..............., xk ' ,  *  1 ,  2 ,..............,  k ' and B symmetric matrix of order k x
k whose ith diagonal element is  ii (i  1,2,.................k ) and its (i,j)th off diagonal element is
1
 ij (i, j  1,2,............, k : i  j ) . (35)
2
where the parameter vector  consists of  o and the elements of  * and B. Assuming that

E    0 and Var     2 I n , the least squares estimate of  is  . Where,

   X ' X 1 X ' y (36)
The predicted response at a point x in the region  is then of the form.
   
y  x     x'  *  x ' B x (37)
Using the method of Langrange multipliers, consider the function
  
 
H   o  x'  *  x' B x   x' x  r 2 (38)
Where  is a Lagrange multiplier.
Differentiating H with respect to x and equating the derivative to zero, we have:
77
 
 *  2 B x  x   0
  (39)
Solving for x, we obtain:
1
1  
x    B  I n   * (40)
2 

The solution of equation 40 represent the stationary point of y  x  . A minimum is achieved at
  H 
this point if the Hessian matrix, that is, the matrix of second-order partial derivatives
x  x' 
of H with respect to x is positive define, from equation36 , this matrix is given by:
  H   
 2 B  I n  (41)
x  x'   

The optimal value of y  x  is computed by substituting x from equation 40 into the right hand

side of equation 37. This computation provides plot of y and xi against r (i = 1, 2, …..,k). These
plots provide avenues for the determination at any given distance r from the origin, the value of
the optimum as well as its location (Box and Draper, 2007; Myers and Montgomery, 1995).
3.14.2: Analyzing the Stationary Points
The existence of optimum value for the response indicates the point is a stationary point. The
stationary point is either maximum, minimum or point of inflexion (saddle point).
The stationary point can be found by using matrix algebra, as the second order model in matrix
form is:
78
 
y   o  x' b  x' Bx (42)

The differential of y with respect to the elements of the vector x is

y
 b  2 Bx  0
x (43)
Therefore, the solution to stationary point is
1
x s   B 1b
2 (44)

 1 
 
 2 
. 
 
Where b  .  (45)
. 
 
. 
 
 q 
 
And
79

   
 11 , 12 1q 
.................................................................
 2, 2 
   
  , 
  2q 
 12  22 , 23
2
.................................................
2 
   
   
  13  23  33 ,  34 2 ,......................................... 3q 2 
 
. 
. 
 
B  . .
. 
 
. 
. 
 
. 
 
 
  
 
 1q ,  2 q ,  3q  4 q ...................................................  qq 
 
 
  (46)
The b is a (q x 1) vector of the first-order regression coefficients and B is a (q x q) symmetric

matrix whose main diagonal elements are the quadratic coefficient (  ii ) and whose off-diagonal
elements are one-half the mixed quadratic coefficients (  ij i  j  (Montgomery, 2005). The
estimated response value at the turning (stationary) point is then given as:
 
y s   o  1 x' s b . (47)
2
3.15: The Test for Significance of Regression
An acceptable regression model should explain the variation of the dependent variable to the
response. There are particular tests of hypotheses about the model variables that can help in
measuring the validity of the model. These include:
80
3.15.1: Residual plots
These tests require for the error term (ei’s) to be normally and independently distributed with
zero mean and variance s2.
The model is assumed to be normally distributed if the plot of residual (error term) against the
percent is approximately along a straight line. The error term is the difference between the
experimental and predicted value. Also, the histogram of frequency against the residual should
show a normal distribution.
Other conditions for the model validity is for the residual against the observation order and the
fitted value to be scattered between both positive and negative residual.
3.15.2: The F-test
For the model to be accepted, the statisticalF is also compared to the critical Fa,q, N-q-1, the model
is rejected if the observed F-value is greater than the critical F.
3.15.3: The P-value
Similarly, the model is rejected when P-value for the statistic F is less than the level of
significance (usually taken as a = 0.05).
3.15.4: The R2 and Adjusted-R2
The exactness of how well the regressed model fit the experimental data can be determined by
the value of R2. The value of R2 is between 0 and 1. The closer the value of R2 to 1 the better the
predicted value to the experimental data. In general the R2 is a measure of the deviation of the
responses from their mean value. However, the addition of a variable (whether significant or
81
insignificant) always increases the value of R2. Hence, some researchers prefer the use of
 
adjusted- R 2 , when the insignificant variable is added the value of adjusted- R 2 will often
decrease (Bradley, 2007).
SS R SS
R2   1  E (48)
SST SST
Where
SS E  SS T  SS R (49)
While,
SS E

2
R  1
n  q  1
SS T
n  1 (50)
Where,
SSE = Sum of square of error
SSR = sum of square of regression model
SST = Total sum of square

In this study both the R2 and adjusted- R 2 were used in this study.
82
3.15.5: The sum of square error (SSE)
The sum of square of error is a measure of the level of variation of the predicted value from the
experimental data. The smaller the SSE the more accurate the ability of the regression model to
predict the experimental data.
83
CHAPTER FOUR
4.0: EXPERIMENTAL RESULTS AND DISCUSSION.
4.1: Characterization of Crude Oils and Solvents
Preliminary investigation was conducted on all the oil samples to determine the WAT using a
programmable Rheometer that measured the viscosity of the crude at different temperatures
under different shear rates (Adewusi, 1997; Ajienka, 1981), The results for crude oil sample A
are given in Figure 11, where the wax appearance temperature (WAT) of the sample oil was
determined to be equal to 43oC. Similar plots were obtained for other oil samples, and their
individual WAT were determined, the results of which were tabulated in Table 4. The properties
of the used petroleum fractions are as shown in Table 5.
3
2 1/s
2.5
20 1/s
40 1/s
Viscosity (pa.s)
2
60 1/s
1.5
80 1/s
1 140 1/s
2001/s
0.5
0
20 30 40 50 60 70
Temperature (oC)
Figure11: Viscosity profile of the oil sample at different shear rate and temperature
84
Table 4: Crude Oil Characteristics
Oil Characteristics SAMPLE

A B C D E F G
Crude Oil Grade Light Light Light Light Medium Medium Light
Gravity at 60oF, APIo 42.1 40.5 39.7 36.2 27.2 25.8 34.8
Viscosity at 60oF, cp 8.5 6.7 7.2 8.1 9.3 8.9 8.7
Wax Appearance Temp. ,oC 43 32 35 21 11 9 19
Pour point, oC 27 24 28 16 6 5 13
Wax Content % 40 35 37 32 20 22 34
Table 5: Kerosene and PMS Properties
Properties Kerosene Premium Motor Spirit (PMS)

Density at 60oF, (kg-m-3) 820 737
Viscosity at 60oF, cP 1.64 1.53
4.2: Wax Deposition Effluent Variables
From the result of the sensitivity analyses on the wax deposition effluent factors: inlet coolant
temperature, Bulk oil temperature, oil flow rate and the wax content in the crude oil.
It was observed in all the experimental runs using the wax deposition flow loop apparatus that
the wax thickness initially increases gradually reaching a peak and then starts diminishing
(Figures 12 -14). This is due to wax-oil gel formed initially consisting of oil entrapped in the wax
deposit until the oil start diffusing out of the gel deposit at the later period leading to the
hardening (aging) of the wax-gel deposit and also due to the effect of shear stripping (shear
removal) which increases as the wax deposit thickness increases leading to the erosion of the
wax deposit. These effect start having effect as the flow rate increases as a result flow channel
reduction.
85
4.2.1: Effect of Inlet Coolant Temperature
The inlet coolant temperatures studied in the experiment were selected at 40oC, 35oC and
30oCrespectively, while the inlet oil temperature was kept constant at 55oC. The Figure11, shows
that the oil viscosity at temperature of 35oC is almost 50% higher than that of 40oC.
Figure12shows that at the same inlet oil temperature the thickness of wax deposit decreases with
the decreasing inlet coolant temperature around the gel point. The reasonable explanation for the
phenomenon is the fact that when the inlet coolant temperature is lowered around the gel point of
the oil, the viscosity of the oil at the liquid-deposit interface will increase sharply as the inlet
coolant temperature decreases. Consequently, the higher oil viscosity close to pipe wall can lead
to the following significant influences:
a. Under the same oil velocity conditions, the shear stress at the liquid-deposit interface will
increase sharply with increasing oil viscosity at the liquid-deposit interface caused by the
decreasing inlet coolant temperature.This will enhance the effect of shear stripping,
leading to decrease in the amount of wax deposit.
b. In terms of the Fick’s mass fusion law, both the radial temperature gradient and the
concentration gradient with respect to temperature will increase as the inlet coolant
temperature decreases, which will have a positive effect on wax deposition.However, the
increasing oil viscosity at the liquid-deposit interface caused by the decreasing inlet
coolant temperature will diminish the molecular diffusion coefficient, which will have a
negative effect on wax deposition.
The combination of the two aspects mentioned above ultimately leads to decrease in the
thickness of wax deposit with decreasing inlet coolant temperature, both the shear stress
and the oil viscosity at the liquid-deposit interface are dominant in the process of wax
86
deposition for the inlet coolant temperature being around the gel point temperature of the
oil.
0.16
0.14 T(wall) =40 Deg.C

Deposited Wax Thickness (Ratio)
0.12 T(wall) =35 Deg.C
T(wall) =30 Deg.C

0.1
0.08
0.06
0.04
0.02
0
0 2 4 6 8 10 12 14
Time (Hours)
Figure12: Wax deposit versus time at different wall temperatures for oil A (flow rate of
1.4ltr/min., Temp. 55oC).
4.2.2: Effect of Bulk Oil Temperature
In the previous studies (Dwivedi, 2012; Semenov, 2012; Noville and Naveira, 2012), the
selection of oil temperature which were studied in the experiment was differentiated in term of
WAT of the oil sample. However, whether the bulk oil temperature was below its WAT or not
did play an important role in the precipitation of wax molecules. Therefore, in this study, the
selection of oil temperatures was differentiated in term of wax appearance temperature of 43oC.
87
4.2.2.1: Case 1: The BulkOil Temperature Above its WAT
The inlet oil temperatures studied in the experiment were selected at 48oC, 52oC and 56oC,
respectively. The inlet coolant temperature was kept constant at 37oC.Figure13shows the
observed dimensionless thickness of the wax deposits as a function of inlet oil temperature at oil
flow rate of 1.4 liter/min. The results indicate that the thickness of wax deposits decreases with
the increasing inlet oil temperature which is above the WAT under the fixed inlet coolant
temperature and oil velocity conditions. The reasons may be due to the fact that when the bulk
oil temperature is above its WAT, there are no wax molecules precipitating out of the bulk oil,
which results in the decrease in wax deposition with increasing oil temperature. In addition,
under the fixed inlet coolant temperature conditions, the higher oil temperature will generate
higher temperature at the liquid-deposit interface which can increase the solubility of wax
molecules, ultimately leading to the fewer amounts of wax deposits.
0.2
0.18 T (oil) = 65 Deg.C
0.16 T (oil) = 55 Deg.C

0.14 T (oil) = 45 Deg.C
0.12
0.1
0.08
0.06
0.04
0.02
0
0 2 4 6 8 10
Time (Hours)
Figure13: Wax deposit versus time at different inlet oil temperatures above WAT at inlet coolant
temperature of 37oC and oil flow rate of 1.4 liter/min.
88
4.2.2.2: Case 2: The BulkOil Temperature Below its WAT
The inlet oil temperatures were selected at 34oC, 37oC and 40oC, respectively. The inlet coolant
temperature was kept constant at 28oC.Figure 14 shows the dimensionless thickness of wax
deposit as a function of inlet oil temperature for different oil velocities. The results of case 2
indicate that the thickness of wax deposits increases with increasing inlet oil temperature which
is below the WAT under the fixed inlet coolant temperature and oil velocity conditions. The
reasons may be due to:
i. When the coolant temperature is below its WAT, the zone of wax precipitation
will be enlarged with the increasing oil temperature. The enlarged zone of wax
precipitation caused by the increase in oil temperature will be prone to make more
amount of wax deposit.
ii. For the fixed inlet coolant temperature, the higher inlet oil temperature can exert
much bigger thermal driving force which enhances its potential to generate more
amounts of wax deposits.
89
0.1
T (oil) = 32 Deg.C.
0.09
T (oil) = 37 Deg.C.
0.08
T (oil) = 41 Deg.C.
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
0 2 4 6 8 10
Time (Hours)
Figure14: Dimensionless thickness of wax deposit versus time at inlet coolant temperature of
28oC (below WAT) at oil flow rate of 1.4 liter/min.
4.2.3: Effect of Oil Flow Rate
Flow regimes were generally believe to have great impacts on wax deposition.The higher oil
flow rate could generate more wax deposit for the laminar flow regime, and less wax deposit for
the turbulent flow regime due to the effect of shear stripping (Todi and Deo, 2006). In this study,
the oil flow rate rather than flow regime is considered the dominant factor affecting the thickness
of wax deposit (Figure 15). In the laminar flow conditions the wax deposit was observed to be
governed by the following theory;
a. The increased oil flow rate increases the shear stress at the liquid-deposit interface, which
will reinforce the intensity of shear stripping, consequently reducing the wax deposit
thickness.
90
b. The internal heat transfer coefficient increases with the increasing oil flow rate for
laminar flow, which increases the radial temperature gradient, consequently leading to
increase in deposit wax thickness.
4.2.3.1: The Flow Rate in Laminar FlowRegime
The results show that as the velocity increases from laminar flow regime (flow rates of 1.00,
1.40, and 1.80 liter/min) to laminar-turbulent transition flow regime (flow rate of 2.20 liter/min)
where the effect of shear stripping started to become dominant as shown in Figure15. The
depositedwax increases when the radial temperature gradient is dominant (during the laminar
flow regime) and deposit thickness decreases when the shear stripping is dominant (during the
laminar-turbulent transition flow regime). The wax that deposited at a higher flow rate is harder
showing by the increase in density of the deposited wax during the turbulent flow regime.
0.18
0.16 Q = 1.00 L/min,Re= 1373
0.14 Q = 1.40 L/min, Re= 1627
0.12 Q = 1.80 L/min, Re= 1881
0.1 Q = 2.20 L/min, Re= 2135
0.08
0.06
0.04
0.02
0
0 2 4 6 8 10 12 14
Time (Hours)
Figure 15: Effect of Flow Rate (Laminar/Laminar-turbulent flow region) on Dimensionless Wax
Thickness.
91
4.2.3.2: The Flow Rate in Turbulent FlowRegime
In turbulent flow regime, the opposite of what was observed in the laminar-turbulent transition
flow regime occurred. The wax deposition rate decreases as the flow rate increase (Figure 16)
confirming what was observed in the literatures that the effect of shear removal becoming more
relevant in the turbulent flow region (Hsu and Brubaker, 1995).
Only those wax crystals and crystal cluster capable of firm attachment to the surface, with good
cohesion among themselves will not be removed from the wax surface (Kelechukwu et al., 2010)
0.05 Q = 1.70 L/min,Re= 3058

Q = 2.00 L/min, Re= 3558
Wax deposit thickness
0.045
Q = 2.40 L/min, Re= 4270
0.04
(mm)
0.035
0.03
0.025
0 1 2 3 4 5 6 7 8
Time (Hours)
Figure 16: Effect of Flow Rate (Turbulent flow region) on Dimensionless Wax Thickness.
These results confirmed the general observations in the literature that molecular diffusion and
shear dispersion are major mechanisms responsible for wax deposition and the effect of other
92
methods: Brownian diffusion, gravity settling/segregation are minor (Dwivediet al., 2012;
Semenov, 2012; Noville and Naveira, 2012; Leontaritis and Geroulis, 2011; Al-Yaari, 2011).
4.3: Comparison of Shear dispersion and Molecular Diffusion Contribution to Wax
Deposition
For all flow conditions, oil is in laminar flow either throughout the pipe or at least in a thin
laminar sub-layer adjacent to the pipe wall. When the oil is being cooled, there will be a
temperaturegradient across the laminar sub-layer. If the temperatures are below the level where
the solid wax crystals can be precipitated, then the flowing elements of oil will contain
precipitated solid particles, and the liquid phase will be in equilibrium with the solid phase; that
is, the liquid will be saturated with dissolved wax crystals. The temperature profile near the wall
will lead to a concentration gradient of dissolved wax, and this dissolved material will be
transported towards the wall by molecular diffusion. When this diffusing material reaches the
solid/liquid interface, it will be precipitated out of solution.
Figure 17 shows the comparison between the deposited wax due to shear dispersion and those
due to shear dispersion and molecular diffusion. Comparing the wax deposition from the two
processes it can be deduced that more wax is deposited through the molecular diffusion
process.The effect of shear dispersion on wax deposition cannot be assumed to be
inconsequential as assumed by most of the current wax deposition models (Fredenslund et al.,
1988; Banki et al., 2008a and b). Hence the wax deposition is caused by temperature gradient
rather than velocity gradient due to higher value of molecular diffusion coefficient compared to
shear dispersion coefficient.
93
0.08
Dimensionless wax deposit (Ratio)
0.07
0.06
0.05
0.04
Shear dispersion
0.03
Molecular diffusion
0.02
0.01
0
0 2 4 6 8 10 12
Time (Hours)
Figure
17 Plot of Dimensionless wax thickness due shear dispersion and molecular diffusion effect
against time.
4.4: Sensitivity Analyses of Shear Dispersion Influent Factorsto Wax Deposition
The effects of various factors influencing the wax deposition tendency during shear
dispersion/gravity settling processes, were experimentally investigated in term of dimensionless

wax deposit defined as the ratio of the wax thickness to the initial inner radius of the pipe, .
R
4.4.1: Effect of Bulk oil and Wall temperature differentialBelow the WAT
Keeping the wax content at 20% and oil flow rate at 1.4 liters/min. and varying the bulk oil and
wall temperature showed that the amount of wax deposition increases with decrease in bulk oil
and wall temperature (Figure18). This is due to the fact that more wax precipitates as the
temperature decreases further below theWAT, leading to bigger wax crystal formation in the
bulk oil. More wax crystals indicate higher shear dispersion coefficient, this is in agreement
with the Bhattacharya (1991) equation.
94
0.013
T=40 Deg.C.
0.012
Dimensionless wax thickness (Ratio)
T=35 Deg.C.
0.011 T = 30 Deg.C.
0.01
0.009
0.008
0.007
0.006
0.005
0.004
0 2 4 6 8 10 12 14 16
Time (Hours)
Figure18: Plot of Dimensionless wax thickness due shear dispersion effect against time at
different temperatures
4.4.2: Effect of Flow Rate
Decreasing the flow rates in the experimental run and keeping the bulk oil and wall temperature
at 40oC and wax content at 20%, resulted in increase in the wax deposit thickness after about five
hours of experimental run. This is partly due to the increase in velocity gradient in the radial
direction as a result of decrease in flow rate and longer residence time of the waxy oil in the
pipe-line, (Figure19).
95
0.013
Q=1.40L/min
0.012
Dimensionless Wax thickness (Ratio)
Q= 1.00L/min
0.011
Q = 0.70 L/min
0.01
0.009
0.008
0.007
0.006
0.005
0.004
0 2 4 6 8 10 12 14 16
Time (Hours)
different flow rates
4.4.3: Effect of Wax content
Increasing the wax content of the sample oil to 32%, by adding the wax scrapped from earlier
experimental run to a new oil sample results in the increase in the wax deposit (Figure 20).The
wax thickness increases with the wax content from 20% to 40%.
96
0.014
WC=20%
0.013
Dimensionless wax thickness (Ratio)
WC=32%
0.012
WC =40%
0.011
0.01
0.009
0.008
0.007
0.006
0.005
0.004
0 2 4 6 8
Time (Hours) 10 12 14 16
different wax contents.
4.5: Variation of Wax Content with Time.
The change in wax fraction in the deposited wax was confirmed by observing the changes
in the density of deposited wax with time, the resultsareas shown in Figure 21. This agrees with the
theoretical concept of wax continuously diffusing into the deposit and the oil seeping out of the
deposit ultimately leading to the net wax density to increase with time. Expectedly, the variation is
not linear due to the circulation of depleted oil repeatedly through the flow loop, leading to
reduction in oil content of subsequent fluid through the flow loop.
97
885
880
875
Oil Density (kg/cubic-m)
870
865
860
855
850
845
840
835
830
0 1 2 3 4 5 6 7 8 9
Time (hours)
Figure21: The change in density of the deposited wax deposit with time at oil flow rate of 1.00
liters/min
4.6: Variation of Outlet Temperature of Test Section with Time
At initial oil temperature of 55oC and coolant temperature of 35oC, the change in outlet
bulk oil and coolant temperatures was observed at different periods. The result is as shown in
Figure 22, where the outlet bulk oil temperature increases with time, while the outlet coolant
temperature decreases with time. This indicates that the wax deposit acts as an insulator, thereby
reducing the heat lost by the bulk oil to the pipe wall as the wax deposit grows. Hence as the wax
deposits grow the loss of heat by the oil to the pipe wall reduces.
98
55
Outlet temperature (Deg. C.)
50
45
40
35 Bulk oil temp.

Water/wall temp.
30
0 5 10 15 20 25 30
Time (hours)
Figure22: Outlet temperatures of the bulk oil and the water/wall temperature at different time.
4.7: Bulk Crude Oil and Wax deposit composition analyses
The need for the PPD to be of the similar length as the deposited wax necessitated the need to
analyze the deposited wax and the crude oil in order to determine the critical carbon number
(CCN) of the wax deposit. The analyses were conducted using high temperature gas
chromatography (HTGC) for three of the oil samples (A, D, and E). There was a noticeable
difference in the physical appearance of the three samples. Observation of the change in mass
percentage of n-alkane composition of the deposited wax in relation to the n-alkane in the crude
oil for each of the three crude oils showed that there was significant addition of C15-C43 alkanes
within the wax deposit and reduction in the same C15-C43 alkanes within the used oil. These
compositional changes for samples A, D and E are as shown in Figures23-25 respectively.
99
0.7
0.6
Change in % mass of n-alkane
0.5
0.4
0.3
0.2
0.1
0
10111213141516171819202122232425262728293031323334353637383940414243444546
-0.1
-0.2
Carbon Number
Figure 23: Mass percentage change of n-alkanes in crude-oil A and its wax deposit
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
-0.1 10 11 12 13 14 15 1617 18 19 20 2122 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 4546
-0.2
Carbon Number
Figure24: Mass percentage change of n-alkanes in crude-oil D and its wax deposit
100
1
0.8
0.6
0.4
0.2
0
10 11 12 13 14 15 1617 18 19 20 2122 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 4546
-0.2
Carbon Number
Figure25: Mass percentage change of n-alkanes in crude-oil E and its wax deposit
It was observed that the n-alkanes percentage mass change transition which corresponds to the
critical carbon number (CNN) of the respective oil (Paso, 2004), for oil sample A is within C15 -
C17 (Figure23), for sample D within C19 - C21 (Figure24), and for oil sample E within C23 -
C25 (Figure25).
These observations originated the used of a group of acrylate ester polymer of alkyl side chain
length between C15 – C17 (solvent P1) as pour point depressant for samples A, B and C, the
alkyl side chain length between C19 – C21 (solvent P2) as pour point depressant for samples D
and G, and alkyl side chain length between C23 – C25 (solvent P3) as pour point depressant for
samples E and F.
4.8: Analysis of Pour Point Experimental Results
The effects of Aliphatic and aromatic compounds in the depression of crude oil pour points are
discussed below:
101
4.8.1: Effect of Petroleum fractions as pour point depressants
Petroleum fractions (Kerosene, Premium motor spirit) were added to crude oil samples in order
to study the effectson pour point depression.
Experimental results showed that relatively high volume of aliphatic solvents (up to 60%) was
needed to depress the pour point of light crude by 8oC and medium crude by as low as 2oC
(Figures26 and 27). These results showed that the use of these aliphatic solvents is not
economically viable. The applicability of alternative solvents (aromatic solvents) as pour point
depressants is therefore highly desirable.
The ineffectiveness of the aliphatic solvents (Kerosene and Petrol (PMS)) in depressing the
crude pour point by appreciable range is due to the fact that the aliphatic solvents are straight or
ring form of hydrocarbons that modify the thermodynamic properties of the crudes by shifting
the wax precipitation equilibrium curve in order to depress the pour point. High dosages of
aliphatic solvents are needed to achieve this equilibrium curve shift.
12
20% composition
10 40% composition
Pour point depression C
60% composition
8
0
A B C D E F G
Oil Sample
Figure26: Pour point depressions of Kerosene on oil samples.
102
14
20% composition
12
C
40% composition
10
Pour point depression
60% composition
8
0
A B C D E F G
Oil Samples
Figure 27: Pour point depressions of PMS (Petrol) on oil samples.
The corresponding effect of the petroleum fractions on wax deposit is as shown in Figure 28, it
shows that as much as 60 % composition by volume of petroleum fractions (PMS, kerosene) are
needed for just about 44 % reduction in the wax thickness, this makes the need for more efficient
wax inhibitors a matter of necessity.
1.2 No solvent
Deposited Wax Thickness (mm)
20 % Composition
1
40 % Composition
0.8 60 % Composition
0.6
0.4
0.2
0 2 4 6 8 10
Time (Hours)
Figure 28: Effect of petroleum fractions (Kerosene) on Wax deposition
103
These inadequacies necessitated the study on the use of aromatic solvents (polymeric
compounds): group of pour point depressants known as acrylate ester copolymer dissolved in
liquid xylene.
4.8.2: Effect of Acrylate Ester Polymers asPour PointDepressants
Acrylate ester co-polymers were added to the crude oil to study their effect in depressing the
pour points.
The results as shown in Figure29 were promising as aromatic solvent as low as 5% in
composition produced remarkable results, reducing the pour point by as much as 5oC. These are
chemicals that contain a carbon ring with delocalized electron double bonding. They modified
the kinetics of wax deposition by embedding into the wax crystal and binding into the wax of
similar size. These interactions thereby decrease the rate of wax formation, leading to a more soft
wax that can be easily transported (Jang et al., 2007).
12
1% composition
Pour point depression ( C)
10
3% composition
8 5% composition
0
P1 on Oil A P1 on Oil B P1 on Oil C P on Oil D P3 on Oil E P3 on Oil F P2 on Oil G
Figure29: Pour point depressions on oil samples by different percentage of polymers.
104
The effect of Acrylate ester polymer (AEP) is as shown in Figure 30, relatively lower volumes of
AEP compared to petroleum fraction inhibit the wax deposit volume by as much as 66%.
1.2
No solvent 1 % Composition
Deposited wax thickness (mm)
1
3 % Composition 5 % Composition
0.8
0.6
0.4
0.2
0
0 2 4 6 8 10 12
Time (Hours)
Figure30: Effect of Acrylate Ester Polymer on Wax deposition
4.8.3: Verification of Different Acrylate Ester Co-polymers, AEP (solvent)in Xylene
Solution as Wax Inhibitor.
The performance of AEP as wax inhibitors in term of wax deposit thickness versus time is as
shown in Table 6.
A total of twenty experiments were performed to investigate the effects of different solvents (P1,
P2 and P3) on crude oil samples(A, D and E). Other experimental results, keeping the crude oil
bulk temperature at 55oC, and coolant temperature at 15oC in term of the volume of oil scrapped
from the sub-cooled section of the flow line are as shown in Table 6.
105
Table 6: Experimental Wax Deposition flow loop results.
Exp. Oil Polymers Solvent Wax Deposited , Wax scooped

No. Sample (ml) (ml) (Experimental) from flow rig
(ml)
1 A P1 0 48 43
2 A P1 5 32 25
3 A P1 10 29 23
4 A P1 30 17.5 13
5 A P2 8 45 40
6 A P2 12 53 46
7 A P2 34 61 55
8 D P2 0 29 22
9 D P2 10 22 18
10 D P2 16 17 15
11 D P2 32 10 6
12 E P3 0 19 14
13 E P3 8 17.5 12
14 E P3 14 15 11
15 E P3 35 10.5 6
The value of the average absolute deviation (AAD) of 5.1 between the experimentally
determined volume and the volume of wax scrapped from the flow loop confirmed the
effectiveness of the wax deposits evaluation method.
Table 6 shows the summary of the physical conditions and the results of the experiments
performed on different crude oils using the three produced polymer solvents (P1 - P3). These
consist of the determined volume of wax deposited from the flow loop experiment and the
amount of wax scooped from the cooled pipe after each of the experiment. It can be seen that
different pour point depressants (PPDs) have varying effect on different crude samples. Also the
volume of pour point depressants (PPDs) needed for optimal performance in inhibiting wax
varies from oil to oil and from solvent to solvent.
106
4.9: Wax Inhibition Evaluation with Fixed Quantity of (PPDs) at Varying Ambient
Conditions.
These experiments were performed with a new sample of the same oil keeping the bulk
temperature at 55oC, and varying the coolant temperature to model the effect of different ambient
conditions on wax deposition. The thermal history of the oil was eradicated after each
experiment by circulating the oil without the coolant for about one (1) hour. This also dissolved
the deposited wax into the oil before the next experimental run. Further analyses were done, to
investigate the cumulative effect of varying solvent (PPDs) and ambient conditions on wax
deposition. Results of these analyses are as shown in Table 7.
Table 7: Experimental flow loop results for sample A at varying ambient temperatures.
Exp. Oil Polymer Solvent Coolant Wax Deposited (ml) Wax scooped
No. Sample (ml) temperature (Experimental) from flow rig
o
C (ml)
16 A P1 5 8 44 37
2 A P1 5 15 32 25
17 A P1 5 18 28 23
18 A P1 5 27 23 18
6 A P2 12 15 53 46
19 A P2 12 12 61 53
20 A P2 12 8 72 62
It was observed from Table 7 that the pressure drop method used to determined the amount of
wax deposition in these experiments was effective, considering the closeness of the results of the
experimental runs to the amount of wax scooped from the flow loop taken into consideration
some amount of wax lost due to shearing at the end of each experiment.
The first four experiments show that the solvent P1 does have effect on the inhibition of wax
deposit for sample A, as the wax deposition was reduced from 48ml when no solvent was used to
107
17.5 when 30ml of the solvent was used (Figure 31). This confirmed the efficacy of solvent with
an average molecular weight around the molecular weight of corresponding alkane of carbon
number equal to the CCN of the deposited wax in inhibiting the wax deposition.
60
No solvent (Exp. 1)
50
5ml of C15-17 (Exp. 2)
Wax Deposition (ml)
40 10ml of C15-17 (Exp. 3)

30ml of C15-17 (Exp. 4)
30
20
10
0
0 2 4 6 8 10
Time ( Hours.)
Figure31: Effect of different quantities of solvent P1 on wax deposition
This hypothesis was further confirmed in experiments 5 - 7 by the failure of solvent P2 to inhibit
wax deposition on sample A (Figure 32), even with addition of as much as 34ml of the solvent to
5 liters of the oil: the wax deposition increased from 48ml when no solvent was used to 61ml
when 34ml of solvent was used; the more the solvent was added, the more wax deposited. This
was due to the fact that the structure of the solvent (P2) is relatively too long to obstruct wax
deposition, instead it increased wax deposition by combining with other similar polymer to form
wax-like structure.
108
70
No solvent (Exp. 1)
60 8ml of C19-21 (Exp. 5)
12ml of C19-21 (Exp. 6)
Wax Deposition (ml)
50
34ml of C19-21 (Exp. 7)
40
30
20
10
0
0 2 4 6 8 10
Time (Hours)
Figure 32: Effect of different quantities of solvent P2 on wax deposition
While using this same solvent P2 on heavier crude whose CCN is comparable with the solvent’s
average CCN, a remarkable result was observed as shown in experiments 8 – 11 inhibiting the
wax deposit volume from 29ml when no solvent was used to 10ml when 32 ml of the solvent
was added (Figure 33).
35 No solvent (Exp. 8)
30 10ml of C19-21 (Exp. 9)
16ml of C19-21 (Exp. 10)
Wax Deposition (ml)
25
32ml of C19-21 (Exp. 11)
20
15
10
0
0 2 4 6 8 10
Time (Hours.)
Figure 33: Further effect of different quantities of solvent P2 on wax deposition
109
The theory of “the efficiency of the solvent with a similar molecular weight as average carbon
number of the deposited wax” necessitated the used of solvent P3 on sample C in experiments
12-15, as the CCN of wax deposit with sample C without the additive was determined to be 24
(high temperature gas chromatography analysis). Experiments 12 - 15 showed that the amount of
wax deposited reduced from 19ml when no solvent was used to 10.5ml when 35ml of the solvent
P3 was used (Figure 34).
20 No Solvent (Exp. 12)

18 8ml of C23-25 (Exp. 13)
16 14ml of C23-25 (Exp. 14)
Wax Deposition (ml)
14 35ml of C23-25 Exp. 15)

12
10
8
6
4
2
0
0 2 4 6 8 10
Time (Hours)
Figure 34: Effect of different quantities of solvent P3 on wax deposition
The effect of varying ambient temperatures showed that as the temperature of the coolant was
reduced below the WAT, the amount of deposited wax increased as shown in experiments 16-18
from 23ml at the coolant temperature of 22oC to 44ml at the coolant temperature of 8oC when
5ml of solvent P1 of similar structure as the oil, was added to sample A (Figure 35).
110
50 coolant temp. = 8 C (Exp. 16)
45
coolant temp. = 15 C (Exp. 2)
40
35
Wax Deposition (ml)

30
25
20
15
10
5
0
0 2 4 6 8 10
Time (Hours)
Figure 35: Effect of different coolant temperature on wax deposition
The increase in amount of deposited wax was relatively higher when dissimilar polymer P2 was
added to sample A as shown in Exp. 6, 19 and 20, increasing from 53ml when the coolant
temperature was 15oC, to 72ml when the temperature was reduced to 8oC (Figure 36).


Wax Deposition (ml)
50
40
30
20
10
0
0 2 4 6 8 10
Time (Hours)
Figure 36: Further effect of different coolant temperatures on wax deposition

111
Figure 37 shows the results when 35ml of P1 solvent was injected after 12 hours of sample-A
passing through the flow loop rig. This solvent earlier confirmed to be effective on this sample A
only succeeded in obstructing the wax deposition, keeping it at 41ml, but was not able to remove
the already deposited wax.
60
50
Wax deposition (ml)
40
30
No solvent
20
35 ml of P1 inj. after 12 hrs.

10
0
0 2 4 6 8 10
Time (Hours)
Figure37: Wax deposition profiles at different injection time using sample A
It can be observed that the pour point depressant (PPD) used to prevent paraffin deposition
exhibited varying degrees of effectiveness, depending on the crude oil being tested, amount of
solvent added and the ambient temperature.
4.10: Evaluation of Wax Inhibition Performance
The Figure38 shows the plot of cumulative wax deposition rate at different coolant temperatures
when 15ml of solvent P1 was added to sample A. The results showed that the cumulative wax
deposition rate increases from 0.07mm/hr at coolant temperature of 32oC to 0.31mm/hr when the
coolant temperature was reduced to 4oC. The ineffectiveness of the solvent on the oil was
112
observed at a coolant temperature of 8oC, which is below the pour point temperature of the oil.
This is due to the formation of the emulsion by the solvent in the oil at these temperatures. This
makes the formed wax gel network to trap relatively bigger emulsified solvent thereby increasing
the amount of wax formed. The wax inhibition percentage reduces from 28% at coolant
temperature of 32oC to -13% when the coolant temperature was 4oC.
0.4
Cummulative wax deposition rate
0.35
with inhibitor
0.3
without inhibitor
0.25
(mm/hr)
0.2
0.15
0.1
0.05
0
0 5 10 15 20 25 30 35
Coolant Temperature (oC)
Figure 38: Cumulative wax deposition rate with and without the P1 inhibitors using sample A.
Figure39 shows the cumulative wax deposition rate profiles with and without the solvent, when
solvent P2 used for sample A. This confirmed the earlier results that solvent P2 is ineffective on
sample A. The wax inhibition percentage varied from -300% at 4oC to -17% at 32oC.
113
0.3
0.25 with inhibitors
0.2 without inhibitors

(mm/hr)
0.15
0.1
0.05
0
0 5 10 15 20 25 30 35
Coolant temperature (oC)
Figure39: Cumulative wax deposition rate with and without the P2 inhibitors using sample A.
Figures 40 and 41, also confirmed the effectiveness of Solvent P2 on sample B, as the wax
inhibition was reduced from 54% at coolant temperature of 32oC to -17% at coolant temperature
of 4oC, and on Solvent P3 on sample C, as the wax inhibition was reduced from 66% at coolant
temperature of 32oC to -13% at coolant temperature of 4oC.
0.25 with inhibitors

without inhibitors
0.2
0.15
(mm/hr)
0.1
0.05
0
0 5 10 15temperature
Coolant 20 (oC) 25 30 35
Figure 40: Cumulative wax deposition rate with and without the P2 inhibitors on sample B.
114
0.18
0.16
with inhibitors
0.14
without inhibitors
0.12
(mm/hr)
0.1
0.08
0.06
0.04
0.02
0
0 5 10 15 20 25 30 35
Coolant temperature (oC)
Figure41: Cumulative wax deposition rate with and without the polymer, P3 on sample C.
From these results, it was generally observed that the wax inhibitors/solvents tested showed good
wax inhibition performance at high temperatures, but many of the inhibitors/additives exhibited
relatively poor performance at low temperatures (below the pour points of the oils). This is due
to inactivity of wax inhibitors at these temperatures, as crude oils below their pour points form
emulsions with the inhibitors due to higher surface tension and high viscosity exhibited by the oil
This prevented the inhibitors from dissolving in the oil and, thus, from inhibiting wax deposition.
4.11: Demulsifier Screening Process
The three identified demulsifiers: methyl methacrylate, butyl acrylate and acrylic acid were
screened by comparing their effect in inhibiting wax deposits. Figure42 showed the observed
result when equal percentage of each of these demulsifiers was mixed with oil sample A in the
flow loop Rig.
115
1.2
No demulsifier
1
Acrylic acid
0.8 Butyl acrylate
0.6 methyl methacrylate
0.4
0.2
0
0 50 100 150 200 250 300 350 400
Time (Minutes)
Figure 42: Effect of different demulsifiers on wax deposit using crude-oil sample A
The results were similar using crude oil samples D and E the results of which were shown
Figures43 and 44 respectively.
1
0.9
Butyl acrylate
0.8
Acrylic acid
0.7
methyl methacrylate
0.6
0.5 No demulsifier
0.4
0.3
0.2
0.1
0 50 100 150 200 250 300
Time (Minutes)
Figure 43: Effect of different demulsifiers on wax deposit using crude-oil sample D
116
0.75
0.7
Deposited Wax thickness (mm)
Acrylic acid
0.65
Butyl acrylate
0.6
methyl methacylate
0.55
No demulsifier
0.5
0.45
0.4
0.35
0.3
0 50 100 150 200 250 300 350
Time (Minutes)
Figure 44: Effect of different demulsifiers on wax deposit using crude-oil sample E
In Figures 42-44, butyl acrylate was seen to provide the best inhibition ability among the three
tested demulsifiers. This necessitated the used of butyl acrylate as the demulsifier in the crude oil
and acrylate ester polymer mixture.
4.12: Effect of Demulsifier on Overall Wax Inhibition
Figure 45 displays the results from a series of wax deposition tests comparing the wax inhibition
with and without polymer based demulsifier on the inhibitor. It shows the wax inhibition profile
of 5000ppm of the demulsifier in a mixture of P1 and sample A and wax inhibition profile of
mixture of sample A and P1 without the de-emulsifier at temperatures between 4oC and 32oC.
The solution of P1, sample A and de-emulsifier returned an average wax inhibition of 15.6% as
against 7.3% obtained by the solution of P1 and sample A without the demulsifier. This is due to
the de-emulsification of the emulsion formed by the inhibitor in the oil especially at low
temperatures due to the reduction of the surface active agents surrounding the emulsified
117
inhibitor making it to form a perfect solution with the oil thereby enhancing its wax inhibiting
efficiency.
0.35
0.3
0.25
Wax Inhibition (fraction)
0.2
0.15
0.1 with demulsifier
without demulsifier
0.05
0
-0.05 0 5 10 15 20 25 30 35
-0.1
-0.15
Figure45: Effect of demulsifier on inhibition of polymer P1 on sample A
Figure 46 confirmed the ability of the polymer based demulsifier in enhancing the efficiency of
wax inhibitors particularly at temperatures below the pour points, when the increased surface
active agent and viscosity of the oil make the formation of emulsion with the added wax
inhibitors feasible. The overall wax inhibition at low coolant temperatures (4 – 10oC) was
calculated to be -8%. Figure 46 further confirmed the ability of a demulsifier to increase the
overall wax inhibition percentage of solvent P2 on sample B from 35% to 42% when the
demulsifier was added.
118
0.6
0.5
0.4
Wax inhibition (fraction)
0.3
0.2 with demulsifier

0.1 without demulsifier
0
0 5 10 15 20 25 30 35
-0.1
-0.2
-0.3
Figure44: Effect of demulsifier on inhibition percentage of polymer, P2 on oil sample B
4.13: EffectsofDemulsifier on Crude Oil Wax Deposits
The effects of demulsifier (butyl acrylate) on wax deposition as crude oils flow in sub-cooled
pipeline were investigated using the Minitab 16 software. The software was used to design the
experiments, analyze the experimental data and interpret the results. This was done by
comparing the optimal wax deposits when the mixture of crude oil, pour point depressant and
demulsifier flow in a sub-cooled pipeline compared to the optimal wax deposits when the
mixtures flow without the demulsifier.
4.13.1: Crude oil mixed with pour point depressant
In each of the experiments, three different samples of acrylate ester co-polymers of varying alkyl
side chain lengths were used as pour point depressants. Each mixture used as pour point
depressant contains equal percentage of each of the three synthesized acrylate ester copolymers.
Each sample was prepared by dissolving them in xylene; as each polymer products are solid
119
material at room temperature. The samples were made by dissolving the 25% of these active
chemicals in xylene. The samples were then put in an oven at a temperature of 60oC, in order to
dissolve the sample.
4.13.2: Crude oil mixed with pour point depressant and a demulsifier.
The wax deposition measurements were then performed on each of the three crude oil samples
(A,D and E) using these pour point depressant mixtures with and without the demulsifier. The
comparisons are then made between the two processes for each crude oil sample to evaluate the
effect of the demulsifier in the enhancement of each of the pour point depressant performance.
4.14: Modeling of Wax inhibitor (AEP/xylene) Mixture Formulation
The design of experiment (DOE) for optimal formulation of pour point depressants (PPD) was
developed using Minitab 16 software.
For group of AEP/xylene mixture formulation the general regression equation is of the form
V   o  1 x1   2 x2  11 x12   22 x22  12 x1 x2 (51)
The wax deposition measurements were performed using the dynamic flow loop Rig shown in
Figure 8 on each of the seven (7) different crude oils of which properties are shown in Table 1.
The two chemical studied are Acrylate ester polymers (15 – 27 centiliters) and xylene (50 – 70
centiliters). Conducting a total of fourteen (14) experiments on each of the seven (7) crude oils,
with nine (9) experiments formulated in a factorial design (including four (4) factorial points,
four (4) axial points and 1 centre point) the remaining five (5) experiments involving the
repetition of the central point to obtain the good estimate of the experimental error. The response
selected for analysis was the average overall wax inhibition percentage defined in equation (27).
120
The same software (Minitab 16) was used for the regression analysis of the experimental data to
fit the second order polynomial equation and also for the evaluation of the statistical significance
of the developed equation.
From the earlier observations (section 4.8), solvent of similar structure as each oil’s deposited
wax was used in order to have maximum performance: a group of Acrylate ester polymer of
alkyl chain C-15, C-16, and C-17, (solvent P1) was applied on oil A, B and C. solvent of alkyl
chain C-19, C-20, and C-21, (solvent P2) was applied on Oil D and G. While, solvent of alkyl
chain C-23, C-24, and C-25 was applied on Oil E and F.
The levels of the variable factors in the experiment are shown in Table 8. While Table 9 shows
the experimental design with the parameter in coded and actual form.
Table 8: Levels of factors selected for the central composite design for AEP/xylene mixture.
Levels
Factors -2 -1 0 1 2
Acrylate Ester Polymer, AEP (cl) X1 15 18 21 24 27
Xylene (cl) X2 50 55 60 65 70
The results of the different experiments in the AEP/xylene mixture formulation as designed by
the minitab 16 software for crude oil samples A to G are tabulated in Table 9.
121
Table 9: Design matrix of the central composite design and corresponding responses using
AEP/xylene mixture
Run X1 X2 AEP Xylene Oil-A Oil-B Oil- Oil- Oil-E Oil-F Oil-
(cl) (cl) C D G
1 -1 -1 18 55 18 17 14 23 10 12 19
2 2 0 27 60 53 59 57 53 43 39 53
3 -1 1 18 65 40 39 37 35 31 27 39
4 1 1 24 65 31 30 31 34 20 21 34
5 1 -1 24 55 40 45 38 33 32 26 40
6 0 2 21 70 23 10 16 21 14 13 25
7 0 -2 21 50 10 4 2 7 5 4 10
8 -2 0 15 60 38 40 43 45 31 33 39
9 0 0 21 60 33 37 29 31 25 21 37
10 0 0 21 60 35 34 27 30 27 23 36
11 0 0 21 60 32 35 30 33 28 20 38
12 0 0 21 60 34 33 28 33 26 22 37
13 0 0 21 60 35 36 30 31 27 23 37
14 0 0 21 60 34 32 29 32 25 23 39
Each response of the average overall wax inhibition using AEP/xylene mixture was used to
develop a mathematical model that correlates the average overall wax inhibition to the volume of
chemicals studied (AEP and Xylene) through first, second order and interaction terms according
to the following second order polynomial equation.
2 2 2
Y  a0   a j x j   aij xi x j   a jj x 2j (52)
j 1 ij 1 j 1
Where Y is the predicted average wax inhibition, W.Iav (%), xi and xj represent the variables
(volume of AEP and Xylene (centiliters)), ao is the offset term, aj is the linear effect, aij is the first
order interaction effect and ajj is the squared effect.
The same software (Minitab 16) was used for the regression analysis of the experimental data to
fit the second order polynomial equation and also for the evaluation of the statistical significance
122
of the developed equation. The analyses of variance (ANOVA) results for the two components
mixture formulation for each of the Crude oil A-G is given in Appendix 1.
The Analysis of variance (ANOVA) results i.e. P-values of the models (regression analysis)
and individual term during AEP/xylene mixture formulations for the crude oil samples are
summarized in Tables 10 and 11 respectively.
Table 10: P values for Crude oil A-G in AEP/Xylene mixture formulations
Term OIL-A OIL-B OIL-C OIL-D OIL-E OIL-F OIL-G
Regression 0.007 0.012 0.018 0.010 0.016 0.012 0.015

AEP 0.024 0.018 0.025 0.028 0.024 0.037 0.000
Xylene 0.020 0.019 0.025 0.037 0.023 0.028 0.019
AEP*AEP 0.013 0.028 0.021 0.019 0.037 0.025 0.012
Xylene*Xylene 0.024 0.028 0.034 0.027 0.029 0.032 0.024
AEP*Xylene 0.037 0.031 0.035 0.023 0.038 0.043 0.011
Lack of fit 0.289 0.840 0.604 0.451 0.222 0.992 0.397
In coded units,
For group of AEP/xylene mixture formulation, the variables coefficients for oil B to G are
as given in Table 11.
Table 11: individual term coefficients for Crude oil A-G in coded units in AEP/xylene
mixture formulations
Constant 467.5 569.5 680.6 486.4 432.4 447.3 378.8

AEP (cl) 1259.7 1561.8 2026.4 1526.39 1152.1 1335.4 995.1
Xylene (cl) -543.6 -599.1 -490.9 -55.46 -602.8 -326.6 -395.7
AEP*AEP 811.8 1012.3 1476.3 1206.9 714.2 983.6 618.4
Xylene*Xylene -432.8 -698.1 -493.5 -440.52 -430.4 -333.4 -489.9
AEP*Xylene -1291.7 -1541.7 -1250.0 -458.33 -1375.0 -833.3 -1083.3
For oil A, the model equation in coded units is given as:
123
W .I .av  467.5  1259.7 x1  543.6 x2  811.8x12  432.8x22  1291.7 x1 x2 (53)
In un-coded units,
For group of AEP/xylene mixture formulation, the variables coefficients for oil A to G in
un-coded units are as given in Table 12.
Table 12: individual term coefficients for Crude oil A-G in AEP/xylene mixture
Formulations
Constant -1161.3 -1620.6 -1122.1 -676.5 -1208.1 -744.6 -1175.37

AEP (cl) 18.56 21.58 6.476 -8.58 21.97 4.03 16.83
Xylene (cl) 32.27 46.77 34.92 25.67 32.66 23.47 33.34
AEP*AEP 0.325 0.405 0.591 0.4828 0.286 0.393 0.247
Xylene*Xylene -0.173 -0.279 -0.197 -0.176 -0.172 -0.133 -0.196
AEP*Xylene -0.517 -0.617 -0.500 -0.183 -0.550 -0.333 -0.433
W .I .av  1161.3  18.56x1  32.27 x2  0.325x12  0.173x22  0.517x1 x2 (54)
Where W.I. av = overall wax inhibition percentage, x1is the volume of group of acrylate
ester polymers (centiliters) (W.I) mixture, x2 is the volume of the xylene (centiliters) in the
wax inhibitor.
It can be observed that the acrylate ester co-polymer and xylene has a synergistic effect on
the overall wax inhibition percentage (W.Iave) that is they increase with the overall wax
inhibition percentage.
4.15: Interaction plot of Overall Wax Inhibition (WIav) for AEP/Xylene Formulation
For an interaction to be significant at least one parent factors must be significant (Wu,
2000). Since both factors (variables) are significant, there is probability of the interaction
124
between the variables.
The interaction plot of the Acrylate ester co-polymer (AEP)/xylene is as shown in Figure
47.
Interaction Plot of WIav (%) for Oil A

Data Means
50 55 60 65 70
50 AEP
(cl)
40 15
18
30 21
AEP (cl)
24
20 27
10
50 Xy lene
(cl)
40 50
55
30 60
Xylene (cl)
65
20 70
10
15 18 21 24 27
Figure 47: Interaction plot for overall wax inhibition percentage (%) for crude oil A
(AEP/xylene)
Acrylate ester co-polymer (AEP) and xylene: both rows indicate the interaction between the
two factors
Row 1: The lines for level -1 of the acrylate ester polymer (AEP) increases, while the line
for level 1 decreases at different rates and the level 0 of the AEP increases and then
decreases, while the volume of xylene increases.
Row 2: The lines for level -1 of the acrylate ester polymer (AEP) increases, while the line
125
for level 1 decreases at different rates and the level 0 of the AEP decreases and then
increases, while the volume of xylene increases.
4.16: Response Surface Graphs for Acrylate Ester Copolymer/Xylene Interactions
Figures 48 and 49 show the surface and the contour graph respectively for oil sample A, the
interaction between the volume of acrylate ester polymers (AEP) and xylene in the two
mixture formulations. It can be observed that as the volume of each component increases
(xylene; 0.50 – 0.70 liter and volume of acrylate ester co-polymer: 0.15 – 0.27 liter)) the
overall wax inhibitor ranges between -13% to 57%, hence there exist an optimum overall
wax inhibition formulation for each of the oil sample.
S ur fa c e plot of W Ia v (% ) for O il A , Vs . A EP (c l) , Xy le ne ( cl)
55
40
W Iav ( % ) , O il A
25
25
10
20 A EP ( cl)
50
60 15
70
X y le ne ( cl)
Figure 48: Surface plots of WIav for varying AEP and xylene for Oil A in AEP/Xylene
formulation
126
C o n to u r P l o t o f W I a v ( % ) f o r O i l A V s . A E P ( c l ) , X y l e n e ( c l )
70
O W I (2
c o m p , O il
A)
< - 10
65 - 10 – 0
0 – 10
10 – 20
20 – 30
Xylene (cl)
30 – 40
60 40 – 50
> 50
55
50
1 5 .0 1 7 .5 2 0 .0 2 2 .5 2 5 .0
A EP ( c l)
Figure 49: Contour plots of WIav for varying AEP and xylene for Oil A in AEP/Xylene
formulation.
The others contour plots of variables interaction for crude oil B to G in AEP/Xylene
formulation are given in Appendix 1.
4.17: Residual Plot of Overall Wax Inhibition (W.Iav) in AEP/Xylene Formulation
The various residual: Normal probability plot, Residual against Fits, Histogram and residual
against observation order plots for the overall wax inhibition (W.Iav) for Oil A in
AEP/Xylene formulation is as shown in Figure 50.
127
Residual Plots of WIav (%)for Oil A AEP/Xylene formulation
Normal Probability Plot Versus Fits
99 2
90 1
Residual
Percent
50 0
10 -1
1 -2
-2 -1 0 1 2 10 20 30 40 50
Residual Fitted Value
Histogram Versus Order

3 2
1
Frequency
Residual
2
0
1
-1
0 -2
-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Residual Observation Order
Figure 50 : Residual plots of WIav for Oil A AEP/Xylene formulation
It was observed in all the residual plots that the model successively predict the experimental
wax inhibition data
The residual plots for crude oil B to G in AEP/Xylene formulation are given in Appendix 3.
4.18: Optimization of AEP/Xylene Formulations
Based on the statistical analysis, the models were highly significant, judging by the low
probability value (P<0.05) returned by the models from the two formulations (group of
AEP/xylene mixture and group of AEP/xylene/butyl acrylate mixture). Also, when the
models sum of squares were decomposed into several sources, and the significant term
verified based on individual P-values that were less than 0.05. In the formulation of group
of AEP/xylene mixture, x1 , x2 , x1 x 2 , x12 and x 22 are significant models terms while during
the group of AEP/xylene/butyl acrylate mixture formulation, x1 , x2 , x 3 , x1 x 2 , x1 x 3 , x 2 x3 ,
128
x12 , x 22 and x32 are significant. Where x1 is the volume of Xylene in the wax inhibitor
(W.I) mixture (cl), x2 is the volume of the group of acrylate ester polymers (cl) and x 3 is
the volume of butyl acrylate (cl). The significance of x 3 , x1 x 3 , x 2 x 3 , and x32 judging by
their P-value being less than 0.01 (P<0.01) confirmed the viability of the inclusion of a
demulsifier (butyl acrylate) in enhancing the wax inhibitor performance.
Further confirmation of the models adequacy through the lack-of-fit test was discovered to
be statistically significant, as the lack-of-fit tests were statistically insignificant judging by
the values of P-value of between 0.11 and 0.58 returned by the models.
Furthermore, the optimum conditions for maximizing the overall wax inhibition percentage
(OWI), calculated by settling the partial derivatives of the models (regression equations) to
zero with respect to the corresponding variables.
The analysis of variance (ANOVA) results for the statistical analyses of crude oils A to G in
AEP/Xylene formulation is given in Appendix 8.
4.19: Analysis of the stationary point of Acrylate Ester Polymer/Xylene Mixture
Formulation
The model equation for crude-oil A can be represented in matrix form by:
 
y   o  x' b  x' B (55)

The derivatives of y with respect to the element of the vector x is,

y
 b  2 Bx  0
x (56)
Therefore the solution to stationary point is
129
1
x s   B 1b
2 (57)
The calculation is as follows:
For crude-oil A, the model equation in un-coded units is,
W .I .av  1161.3  18.56x1  32.27 x2  0.325x12  0.173x22  0.517 x1 x2 (58)
0.325  0.517   18.56 

2  B  0.325  0.259  b 
B  32.27 
 0.517 0.325    0.517 0.325 
 
1 d b 
Since B 1    (59)
ad  bc   c a 
 
1  0.173 0.259 0.91  1.36 

B 1      
0.19 0.517 0.325  2.72  1.71

 27 
B 1b   
 105.66
1
The stationary point using the equation x s   B 1b , gives:
2
13.50 
xs   
52.80
This means that for crude oil A, the optimum volumes of acrylate ester polymer and xylene
for maximum overall wax inhibitions are 13.50 cl and 52.80 cl respectively.
Similarly, the optimum conditions were determined by using the same computational
process for crude oil B to G.
The results of which were tabulated in Table 13.
130
Table 13: The PPD Formulation Optimization results using copolymer (AEP)/Xylene
Oil AEP Mixture Xylene (g) Maximum overall Overall wax

(g) wax inhibition (W.I. inhibition (W.I.av)
av) from software (Experiment) (%)
(%)
A 13.50 52.80 61 58
B 15.65 58.16 52 54
C 14.35 63.12 51 55
D 15.47 61.34 62 65
E 16.23 52.89 48 46
F 17.23 51.79 51 48
G 15.36 64.07 56 49
The closeness of the optimum overall average wax inhibition from the software (48-62%)
and the experimental optimum overall wax inhibition using the optimal AEP/xylene mixture
formulation (46-65%) as shown in Table 16 and percentage AAD value of 3.40% validated
the effectiveness of the proposed wax inhibition evaluation method. Though, the
experimental optimum overall wax inhibition values were in general much lower than
alternate optimum overall wax inhibition values from the Minitab 16 software.
4.20: Modeling of AEP/Xylene/Butyl acrylate (Demulsifier) Mixture Formulation
In an efforts to enhance the wax inhibition process of the pour point depressants, butyl acrylate
which acts as a demulsifier was added as one of the component of the wax inhibitors/pour point
depressant. The above procedures were then repeated with three chemicals: group of acrylate
ester polymers (AEP), xylene and butyl acrylate (demulsifier) mixture on each of the seven crude
oil samples. A total of twenty-one (21) experiments were conducted on each of the seven (7)
crude oils, with sixteen (16) experiments formulated in a factorial design, including nine (9)
131
factorial points, six (6) axial points and 1 centre point, the remaining five (5) experiments
involving the repetition of the central point to obtain the good estimate of the experimental error.
The response selected for analysis was the average overall wax inhibition percentage.
Applying solvent of similar structure as each deposited wax to each of the crude oil in order to
have maximum performance: a group of Acrylate ester polymer of alkyl chain C-15, C-16, and
C-17, P1 was applied on oil A, B and C. solvent of alkyl chain C-19, C-20, and C-21, P2 was
applied on Oil D and G. While, solvent of alkyl chain C-23, C-24, and C-25, P3 was applied on
Oil E and F.
The levels of the variable factors in the experiment are shown in Table 14. While Table 15
shows the experimental design with the parameter in coded and actual form.
Table 14: Levels of factors selected for the central composite design for AEP/xylene/butyl
acrylate mixture.
Levels
Factors -2 -1 0 1 2
Acrylate Ester Polymer, AEP (cl) X1 15 18 21 24 27
Xylene (cl) X2 50 55 60 65 70
Butyl acrylate (cl) X3 1 2 3 4 5
The results of the different experiments in the AEP/xylene/butyl acrylate mixture formulation as
designed by the minitab 16 software for crude oil samples A to G are tabulated in Table 15.
132
Table 15: Design matrix of the central composite design and corresponding responses using
AEP/xylene/butyl acrylate mixture
Run X1 X2 X3 AEP Xylene Butyl- Oil- Oil- Oil- Oil- Oil- Oil- Oil-
(cl) (cl) Acrylate A B C D E F G
(cl)
1 -1 -1 1 18 55 4 25 27 29 23 18 29 25
2 -1 -1 -1 18 55 2 38 45 39 37 24 37 45
3 2 0 0 27 60 3 57 60 62 50 41 53 64
4 -1 1 -1 18 65 2 37 41 47 32 31 33 38
5 1 1 -1 24 65 2 40 43 42 31 34 42 43
6 1 -1 -1 24 55 2 26 28 26 15 19 30 30
7 1 -1 1 24 55 4 37 43 42 33 23 35 40
8 1 1 1 24 65 4 64 67 66 58 51 68 70
9 0 0 2 21 60 3 45 46 50 48 36 50 48
10 0 2 0 21 70 3 58 61 62 50 48 59 53
11 0 -2 0 21 50 3 30 39 29 27 18 34 32
12 -2 0 0 15 60 3 43 43 51 39 37 36 47
13 -1 1 1 18 65 4 38 34 46 32 40 40 38
14 0 0 -2 21 60 1 32 36 35 30 26 38 39
15 2 -2 0 27 50 3 34 41 36 29 20 34 33
16 0 0 0 21 60 3 20 23 27 12 15 22 24
17 0 0 0 21 60 3 21 21 25 15 17 25 25
18 0 0 0 21 60 3 18 19 24 13 18 26 27
19 0 0 0 21 60 3 20 21 26 14 17 22 22
20 0 0 0 21 60 3 19 22 22 15 17 23 24
21 0 0 0 21 60 3 22 23 25 15 15 24 27
Each response of the average overall wax inhibition using AEP/xylene/butyl acrylate mixture
was used to develop an empirical model that correlates the average overall wax inhibition to the
mass of chemicals studied (AEP, Xylene and butyl acrylate) through first, second order and
interaction terms according to the following second order polynomial equation.
For group of AEP/xylene/butyl acrylate mixture the model equation result is given as:
133
W .I .av   o  1 x1   2 x 2   3 x3  11 x12   22 x22 (60)
2
   12 x1 x 2  13 x1 x3   23 x 2 x3
33
Where W.I. av = overall wax inhibition percentage, x1 is the volume of the group of acrylate
ester polymers (centiliters) (W.I) mixture, x2 is the volumeof xylene (centiliters) and x 3 is the
volume of butyl acrylate (centiliters) in the wax inhibitor, 1,  2, and  3 are the linear
coefficients; 12, 13, and  23 are the interaction coefficients, while the 11 ,  22, and  33 are the
quadratic coefficients.
For group of AEP/xylene/butyl acrylate mixture the model equation is of the form
W .I .av   o  1 x1   2 x 2   3 x3  11 x12   22 x 22   332  12 x1 x 2  13 x1 x3   23 x 2 x3 (56)
Where x1 is the mass fraction of Xylene in the wax inhibitor (W.I) mixture, x2 is the volume of
the group of acrylate ester polymers and x 3 is the mass fraction of butyl acrylate., 1,  2, and  3
are the linear coefficients; 12, 13, and  23 are the interaction coefficients, while the
11 ,  22 , and  33 are the quadratic coefficients.
The Analysis of variance (ANOVA) results of individual term during AEP/xylene/butyl acrylate
mixture formulations for the crude oil samples in un-coded units and P-values of the models
(regression analysis) are summarized in Table 16 and 17 respectively. The analyses of variance
(ANOVA) results for the three components mixture formulation for each of the Crude oil A-G is
given in Appendix 9.
134
Table 16: Individual term coefficients for Crude oil A-G in AEP/xylene/butyl acrylate mixture
formulations in un-coded units.
Constant 1195.08 1977.71 1391.63 1852.78 1130.46 1665.17 1632.98

AEP -56.75 59.19 49.85 59.27 35.24 46.43 58.89
Xylene -15.44 40.53 25.61 36.21 22.90 35.45 28.20
Butyl-acrylate -98.14 112.62 92.69 117.56 77.00 101.12 131.27
AEP*AEP 0.87 0.82 0.87 0.81 0.63 0.59 0.84
Xylene*Xylene 0.07 0.28 0.20 0.24 0.17 0.23 0.17
Butyl-acry.* Butyl-acry. 4.70 4.79 4.38 5.94 3.67 3.96 4.58
AEP*Xylene 0.26 0.30 0.13 0.31 0.12 0.32 0.30
AEP* Butyl-acrylate 1.96 2.67 2.12 2.46 0.75 1.33 2.33
Xylene* Butyl-acrylate 0.53 0.50 0.43 0.57 0.70 0.90 0.95
The P-values for different crude oil sample are as given as:
Table 17: P-values for Crude oil A-G in AEP/xylene/butyl acrylate mixture formulations
Regression 0.023 0.031 0.042 0.023 0.035 0.024 0.026

AEP 0.015 0.034 0.027 0.024 0.042 0.014 0.017
Xylene 0.036 0.025 0.017 0.024 0.032 0.018 0.029
Butyl-acrylate 0.022 0.035 0.018 0.033 0.025 0.034 0.028
AEP*AEP 0.023 0.015 0.042 0.038 0.021 0.014 0.033
Xylene*Xylene 0.045 0.042 0.044 0.039 0.042 0.039 0.040
Butyl-acry.* Butyl-acrylate 0.034 0.047 0.036 0.029 0.031 0.048 0.032
AEP*Xylene 0.017 0.018 0.025 0.036 0.029 0.017 0.018
AEP* Butyl-acrylate 0.007 0.006 0.011 0.017 0.012 0.013 0.010
Xylene* Butyl-acrylate 0.034 0.027 0.026 0.041 0.038 0.028 0.012
Lack of fit 0.735 0.845 0.719 0.471 0.346 0.405 0.983
For Oil A, the model equation in un-coded units is given as:
135
W .I av  1195.08  56.75 x1  15.44 x2  98.14 x3  0.87 x12 (61)
 0.07 x22  4.70 x32  0.26 x1 x2  1.96 x1 x3  0.53x2 x3
Where W.I. av = overall wax inhibition percentage, x1is the volume of group of acrylate ester
polymers (centiliters) (W.I) mixture, x2 is the volume of the xylene (centiliters) and x3 is the
volume of butyl acrylate (centiliters) in the wax inhibitor.
It can be observed that each of the variables has synergistic effect on the overall wax inhibition
percentage. They increase the overall wax inhibition percentage.
4.21: Interaction Plot of Overall wax inhibition (WIav) for AEP/Xylene/Butyl-acrylate
Formulation.
Any significant interaction must have at least one significant parent factors (Wu, 2000). Since all
the three factors (variables) are significant, there is probability of the interactions between the
variables.
The interaction plot of the Acrylate ester co-polymer (AEP)/xylene/butyl-acrylateis as shown in
Figure51.
136
Inte r a ction P lot of W Ia v (% ) for O il A
Da ta M e a ns
50 55 60 65 80
60
A EP
(cl)
15
40 18
A E P ( c l)
21
24
20 27
60
Xy le ne
(cl)
50
40 55
X y le n e ( c l)
60
65
20 80
60
B uty l-a cry la te
(cl)
1
40 2
B u t y l- a c r y la t e ( c l)
3
4
20 5
15 18 21 24 27 1 2 3 4 5
Figure 51: Interaction plot for overall wax inhibition percentage W.I.av(%) for crude oil A
(AEP/xylene/butyl-acrylate)
Each pair of factors provides the following summary
a. Acrylate ester co-polymer (AEP) and xylene: both rows indicate the interaction
between the two factors
Row 1: The lines for level -1 and 1 of the AEP increase but at different rates and the level
0 of the AEP decreases and then increases, while the volume of xylene increases.
Row 2: The lines for level 1 of the xylene increase, while the line for level -1 decrease at
different rates and the level 0 of the xylene decreases and then increases, while the
volume of AEP increases.
137
b. Acrylate ester co-polymer (AEP) and butyl acrylate: both rows indicate the
interaction between the two factors
Row 1: The lines for level 1 of the acrylate ester co-polymer increase, while the line for
level -1 decrease at different rates and the level 0 acrylate ester co-polymer of the
decreases and then increases, while the volume of butyl-acrylate increases.
Row 2: The lines for level 1 of the butyl-acrylate increase, while the line for level -1
decrease at different rates and the level 0 of the butyl-acrylate decreases and then
increases, while the volume of acrylate ester co-polymer increases.
c. Xylene and butyl acrylate: both rows indicate the interaction between the two factors
Row 1: The lines for level -1 and 1 of the xylene increase but at different rates and the
level 0 of the xylene decreases and then increases, while the volume of butyl-acrylate
increases.
Row 2: The lines for level -1 and 1 of the butyl-acrylate increase but at different rates and
the level 0 of the butyl-acrylate decreases and then increases, while the volume of xylene
increases.
4.22: Response Surface Graphs for Acrylate Ester Copolymer/Xylene/Butyl-acrylate
Interactions
The Figures 52 and 53 show the surface and the contour plots for the AEP and butyl acrylate
interactions in AEP/ Xylene/butyl acrylate mixture formulation. It was observed that as the
volume of AEP varies between 15 to 27centiliter and volume of butyl acrylate varies between 1
to 5 centiliter the overall wax inhibition percentage of the AEP/xylene/butyl acrylate mixture
varies between 36% to 75%, indicating the effectiveness of the AEP and butyl acrylate in the
mixture formulation.
138
S ur fa ce pl ots of W Ia v ( % ) , O il A V s A EP (c l ) V s B uty l a cr y la te ( c l)
60
W I a v ( % ) , O il A
40
25
20
20 A EP ( cl)
1.0
2.5 15
4.0
5.5
B uty l A cr y la te ( cl)
Figure 52: Surface plot of WIav at varying AEP and butyl-acrylate for Oil A in AEP/Xylene/butyl
acrylate formulation
C o nto ur p l o t f o r W .I.a v ( % ) f o r O i l A V s . A E P ( c l ) , B uty l a c r y l a te ( c l )

5
O W I (O il
A)
< 40
40 – 60
4 60 – 80
80 – 100
Butyl acrylate (cl)
> 100
H o ld V alu es
X y len e 60
3
1
1 5 .0 1 7 .5 2 0 .0 2 2 .5 2 5 .0
A EP ( c l)
Figure53: Contour plots of WIav for varying AEP and butyl-acrylate for Oil A in
139
Similarly, Figures54 and 55 show the surface and the contour plots for the xylene and butyl
acrylate interactions in AEP/ Xylene/butyl acrylate mixture formulation. It was observed that as
the mass fraction of xylene varies between 50 to 70 cl and mass fraction of butyl acrylate varies
between 1 to 5 cl, the overall wax inhibition percentage of the AEP/xylene/butyl acrylate mixture
varies from between 47% to 92%, indicating the effectiveness of the AEP and butyl acrylate in
the mixture formulation.
S urface plot of W Iav (% ) (O il A) Vs Xy lene (cl), Buty l acry late (cl)
60
W Iav ( % ) , O il A
40
80
20 70
60 X y le ne ( cl)
1.0
2.5 50
4.0
5.5
B uty l A cr y late ( cl)
Figure54: Surface plots of WIav for varying xylene and butyl acrylate for Oil A in
140
C o nto ur p lo t o f W .I.a v ( % ) f o r O il A V s .X y l e ne ( c l) , B uty l a c r y la te ( c l )
5
O W I ( O il
A)
< 40
40 – 60
60 – 80
4
80 – 100
Butyl acrylate (cl)
> 100
H o ld V alu es
A E P 21
3
1
50 55 60 65 70 75 80
X y le n e (c l)
Figure 55: Contour plots of WIav for varying xylene and butyl acrylate for Oil A in
In the same argument, Figures56 and 57 show the interaction of AEP and xylene in overall wax
inhibition for oil sample A, as the volume of AEP value increases from 15 to 27 cl and that of
xylene changes from 50 to 70 cl, the overall wax inhibition percent increases from 39% to 90%.
Hence the inclusion of each of the three chemicals: AEP, Xylene and butyl acrylate in the wax
inhibitors mixture formulation is justifiable.
141
S urface plot of WIav (% ) , O il A Vs AEP (cl), Xylene (cl)
60
WIav ( % ), O il A
40
25
20
20 A EP (cl)
50
60 15
70
80
Xylene ( cl)
Figure56: Surface plots of WIav for varying AEP and xylene for Oil A in AEP/Xylene/butyl
C o n to u r p lo t o f W .Ia v ( % ) f o r O il A V s . A E P ( c l) , X y le n e ( c l)
80
O W I ( O il A )
< 40
40 – 60
75 60 – 80
80 – 100
100 – 120
70 > 120
Xylene (cl)
H o ld V a lu e s
B u ty l a c r y la t e 3
65
60
55
50
1 5 .0 1 7 .5 2 0 .0 2 2 .5 2 5 .0
A E P ( c l)
Figure57: Contour plots of WIav for varying AEP and xylene for Oil A in AEP/Xylene/butyl
142
The others contour plots of variables interaction for crude oil B to G in AEP/Xylene/Butyl-
acrylate formulation are given in Appendix 2.
4.23: Residual Plot of Overall Wax Inhibition (W.Iav) in AEP/Xylene/Butyl-acrylate
Formulation
against observation order plots for the overall wax inhibition (W.Iav) for Oil A in
AEP/Xylene/Butyl acrylate formulation is as shown in Figure 58.
Residual Plots of WIav for Oil A AEP/Xylene/Butyl-acrylate formulation

99 10
90
5
Residual
Percent
50
0
10
1 -5
-5 0 5 10 20 30 40 50 60

8 10
6
Frequency
5
Residual
4
0
2
0 -5
-2.5 0.0 2.5 5.0 7.5 2 4 6 8 10 12 14 16 18 20
Figure 58: Residual plots of WIav for Oil A AEP/Xylene formulation
143
It was observed in all the residual plots that the model successively predict the experimental wax
inhibition data to acceptable degree of accuracy.
The residual plots for crude oil B to G in AEP/Xylene/Butyl-acrylate are given in Appendix 4.
4.24: Analysis of the stationary point of Acrylate Ester Polymer/Xylene/Butyl Acrylate
Mixture Formulation
For Crude-Oil A, the model equation in un-coded units is,
W .I .overall  1161.3  18.56 x1  32.27 x2  0.325x12  0.173x22  0.517x1 x2
(63)
0.87 0.26 1.96 

 2 2 0.87 0.13 0.98  56.75
  
0.53  B  0.26
 b   15.44 
B  0.26 0.07 0.07 0.27 
2    98.14
 
1.96 0.53 4.70  1.96 0.53 4.70 
 
0.1859  0.0916  0.0335

 
1   0.6928
B 1  2.1682 0.0199 
0.07226  
0.0006  2.063 0.0271 

2.573  1.268  0.464

 
B 1   9.588 30.007 0.275 
 
0.008  2.855 0.375 
144
 80.931
B b   53.783
1
 6.804 
1
The stationary point using the equation x s   B 1b , gives:
2
40.466
x s  26.891
3.402 
This means that for crude oil A, the optimum volumes of acrylate ester polymer, xylene and
butyl acrylate required for maximum value of overall wax inhibitions are40.466 cl,26.891 cl and
3.402 cl respectively. Similarly, the optimum conditions were determined for maximum overall
wax inhibition by using the same computational process for crude oil B to G. The optimum wax
inhibition results from the AEP/xylene/butyl acrylate mixture experiments determined to be
between 67-82% as against the values of between 63-86 %using the Minitab 16 software with
percentage AAD value of 3.95% further confirmed the viability of the proposed evaluation
method (Table 18).
Table 18: The Optimization results of PPD Formulation using copolymer/xylene/Butyl acrylate
Oil AEP Xylene (cl) Butyl acrylate Optimal overall Optimal overall
Mixture (cl) (cl) wax inhibition wax inhibition
from software (Experiment) (%)
(%)
A 40.46 26.89 3.40 84 81
B 42.67 28.87 3.17 76 71
C 41.19 26.91 2.5 83 78
D 46.27 22.24 3.3 82 79
E 41.13 24.58 4.1 63 67
F 46.15 25.81 4.5 66 68
G 40.04 28.77 3.7 86 82
145
4.25: Comparison of Optimum AEP/Xylene to Optimum AEP/Xylene/Butyl acrylate
Formulation
Comparing the optimum overall wax inhibition of the AEP/Xylene formulation to that of the
AEP/Xylene/Butyl acrylate formulation, it was observed that the addition of the butyl acrylate
was able to enhance the performance of the wax inhibitors judging by the increase in the overall
wax inhibition (WIav) percentage from between 46 – 65% when AEP and Xylene was used in the
formulation compare to WIav percentage of between 67 – 82% in the formulation involving the
use of AEP, xylene and butyl acrylate. This confirmed the viability of the inclusion of butyl
acrylate as an additive in the wax inhibitors formulation. This is as shown in Figure 59.
90 AEP/Xylene AEP/Xylene/butyl acrylate

Overall wax inhibition (Optimum), %
80
70
60
50
40
30
20
10
0
A B C D E F G
Oil Samples
Figure59: Optimal overall wax inhibition for AEP/Xylene and AEP/Xylene/Butyl acrylate.
4.26: Modeling of Wax Deposition Process for Each Crude Oil Sample.
The modeling of wax deposit inhibition process during the flow of single phase fluid in sub-
cooled pipeline were determined through the use of response surface methodology central
146
composite design (RSMCCD) using four identified wax deposit inhibition variables: Bulk oil
temperature, coolant temperature, flow rate and the fraction of wax inhibitor in the oil.
Optimal wax inhibitor formulation for each crude oil from the previous chapter was used as
effective wax inhibitor among other factors.
A total of thirty-one (31) experiments were conducted on each of the seven (7) crude oils, with
Twenty-five (25) experiments formulated in a factorial design (including sixteen (16) factorial
points, eight (8) axial points and 1 centre point) the remaining six (6) experiments involving the
repetition of the central point to obtain the good estimate of the experimental error. The response
selected for analysis was the volume of deposited wax in cubic-centimeters.
The close range of the overall wax inhibition values when the experiments were repeated
validate the effectiveness of the proposed wax inhibition evaluation method (Overall wax
inhibition). When the resulting optimal wax inhibitor (WI) formulations shown in Table 10 were
used as the wax inhibitors in the wax deposition inhibition processes, and the wax deposition
inhibition process optimized for each of the seven crude samples using RSMCCD. The study
was in two stages: the laminar regime (Re < 2000) and laminar- turbulent transition regime
(2000 < Re < 3000).
4.26.1: Modeling of Wax depositionin the Laminar Flow Regime (Re <2000)
Constraining the four wax deposition process variables within the following values: Oil bulk
temperature, Tb (321 - 329 K), coolant/pipeline temperature, Tc (277 - 298 K), oil flow rate, Q
(0.2 - 0.7 liter/min (Re < 2000)) and the percentage of wax inhibitors added to the oil, Iw (1 -
5%). The software was first used to determine the minimum number of variable combinations
required to evaluate the optimum condition for wax deposition inhibition for each of the oil.
147
Then, experiments were performed with the designed variables combinations. The volume of
wax deposit from the experimental runs for each of the crude oil sample was then entered into
the software so as to develop mathematical models that correlates the volume of wax deposit to
the studied variables through first, second term and interaction term.
The levels of the variable factors in the experiment are shown in Table 19. While Table 20 shows
the experimental design with the parameter in coded and actual form.
Table 19: Levels of factors selected for the central composite design for wax deposition
inhibition.
Levels
Factors -2 -1 0 1 2
Cooling water temperature (K), A X1 277 282 287 292 297
Bulk oil temperature (K), B X2 321 323 325 327 329
Wax inhibitor percentage (%), C X3 1 2 3 4 5
Flow rate (Litres/min), D X4 0.3 0.4 0.5 0.6 0.7
The wax deposition data for fluid flow in laminar flow regime is given as:
148
Table 20: Design matrix of the central composite design and corresponding responses of wax
deposits inhibition process (Laminar flow, Re < 2000).
Run X1 X2 X3 X4 A B C D Oil- Oil- Oil- Oil- Oil- Oil- Oil-

A B C D E F G
1 -1 -1 1 -1 282 323 4 0.4 111 117 129 106 94 98 104
2 0 0 0 0 287 325 3 0.5 112 102 120 95 78 92 102
3 -1 -1 1 1 282 323 4 0.6 130 139 143 132 108 106 125
4 0 0 0 0 287 325 3 0.5 112 104 121 97 78 94 100
5 0 0 0 0 287 325 3 0.5 113 108 119 99 80 96 103
6 -1 -1 -1 -1 282 323 2 0.4 95 105 125 87 79 100 87
7 2 0 0 0 297 325 3 0.5 172 184 189 170 117 154 166
8 0 0 0 -2 287 325 3 0.3 103 106 134 97 90 101 98
9 -1 1 -1 1 282 327 2 0.6 101 97 105 92 86 93 99
10 -1 -1 -1 1 282 323 2 0.6 115 126 142 111 93 111 108
11 1 1 -1 1 292 327 2 0.6 161 172 182 160 122 132 149
12 1 -1 -1 1 292 323 2 0.6 100 105 114 90 83 93 95
13 -1 1 -1 -1 282 327 2 0.4 83 76 90 70 75 83 77
14 1 -1 1 1 292 323 4 0.6 143 157 161 143 122 128 150
15 1 -1 1 -1 292 323 4 0.4 127 135 146 119 109 116 128
16 1 1 1 -1 292 327 4 0.4 225 238 253 228 167 193 217
17 0 0 2 0 287 325 5 0.5 188 207 199 189 154 152 190
18 0 2 0 0 287 329 3 0.5 178 167 191 177 145 162 165
19 0 0 0 0 287 325 3 0.5 114 104 120 98 79 94 99
20 1 -1 -1 -1 292 323 2 0.4 81 82 95 66 70 85 72
21 0 -2 0 0 287 321 3 0.5 93 97 120 84 92 97 88
22 -2 0 0 0 277 325 3 0.5 97 95 97 90 64 100 86
23 1 1 -1 -1 292 327 2 0.4 142 150 164 139 110 121 127
24 -1 1 1 1 282 327 4 0.6 155 147 145 145 124 134 145
25 1 1 1 1 292 327 4 0.6 241 261 263 252 181 201 233
26 0 0 0 0 287 325 3 0.5 115 107 118 94 80 93 100
27 -1 1 1 -1 282 327 4 0.4 134 124 132 125 107 124 122
28 0 0 0 2 287 325 3 0.7 144 150 160 144 121 122 145
29 0 0 -2 0 287 325 1 0.5 91 107 110 81 85 80 87
30 0 0 0 0 287 325 3 0.5 114 105 119 95 78 95 102
31 0 0 0 0 287 325 3 0.5 116 108 121 98 81 92 105
The close range of the overall wax inhibition values when the experiments were repeated
validate the effectiveness of the proposed wax inhibition evaluation method (Overall wax
inhibition). When the resulting optimal wax inhibitor (WI) formulations shown in Table 8 were
used as the wax inhibitors in the wax deposition inhibition processes, and the wax
149
depositioninhibition process optimized for each of the seven crude samples using the RSMCCD.
The study was of two stages: laminar regime (Re < 2000) and the turbulent flow regime (Re
>3000).
4.26.2: Modeling of Wax deposition in the Turbulent Flow Regime (Re > 3000)
Table 21: Design matrix of the central composite design and corresponding responses of wax
deposits inhibition process (Turbulent, Re >3000).
Run X1 X2 X3 X4 A B C D Oil- Oil- Oil- Oil- Oil- Oil- Oil-

A B C D E F G
1 -1 -1 1 -1 282 323 4 1.1 122 129 140 111 113 120 123
2 0 0 0 0 287 325 3 1.2 111 100 115 100 73 93 108
3 -1 -1 1 1 282 323 4 1.3 95 100 119 95 80 96 99
4 0 0 0 0 287 325 3 1.2 110 131 147 111 90 105 104
5 0 0 0 0 287 325 3 1.2 98 92 110 99 84 97 93
6 -1 -1 -1 -1 282 323 2 1.1 93 110 120 86 73 105 95
7 2 0 0 0 297 325 3 1.2 114 103 120 105 73 90 112
8 0 0 0 -2 287 325 3 1.0 116 107 126 100 79 85 107
9 -1 1 -1 1 282 327 2 1.3 150 142 150 146 120 138 140
10 -1 -1 -1 1 282 323 2 1.3 225 238 253 228 167 193 217
11 1 1 -1 1 292 327 2 1.3 80 76 99 62 65 89 70
12 1 -1 -1 1 292 323 2 1.3 82 70 95 74 73 86 79
13 -1 1 -1 -1 282 327 2 1.1 138 153 169 134 105 116 120
14 1 -1 1 1 292 323 4 1.3 106 111 124 112 100 102 104
15 1 -1 1 -1 292 323 4 1.1 140 153 154 149 127 132 150
16 1 1 1 -1 292 327 4 1.1 112 102 120 95 78 92 102
17 0 0 2 0 287 325 5 1.2 125 134 140 127 112 111 129
18 0 2 0 0 287 329 3 1.2 134 124 132 125 107 124 122
19 0 0 0 0 287 325 3 1.2 155 164 177 155 128 137 154
20 1 -1 -1 -1 292 323 2 1.1 235 257 269 258 178 208 225
21 0 -2 0 0 287 321 3 1.2 114 105 119 95 78 95 102
22 -2 0 0 0 277 325 3 1.2 144 150 163 147 117 127 153
23 1 1 -1 -1 292 327 2 1.1 93 97 120 84 92 97 88
24 -1 1 1 1 282 327 4 1.3 103 101 138 100 88 107 98
25 1 1 1 1 292 327 4 1.3 97 95 97 90 64 100 86
26 0 0 0 0 287 325 3 1.2 93 94 100 87 63 105 90
27 -1 1 1 -1 282 327 4 1.1 115 107 118 90 80 91 102
28 0 0 0 2 287 325 3 1.4 175 170 185 182 153 168 170
29 0 0 -2 0 287 325 1 1.2 178 190 193 174 112 150 160
30 0 0 0 0 287 325 3 1.2 185 200 193 193 158 156 194
31 0 0 0 0 287 325 3 1.2 122 129 140 111 113 120 123
150
The above procedures were repeated in the turbulent flow regimes by investigating the volume of
wax deposited during specific period at oil flow rates of between 1.1 - 1.5 liter/min (Re > 3000).
The response selected for analysis was the volume of deposited wax in cubic-centimeters for
each of the crude oil sample. The results of the different experiments in the wax deposition
effluent factors sensitivity analyses as designed by the minitab -16 software for crude oil samples
A to G with the parameter in coded and actual form are tabulated in Table 21 above.
4.27: Analyses of Variance (ANOVA) and Model Fitting (Re < 2000)
The ANOVA results of the equations for all the crude oil samples (A-G) in the laminar flow
region (Re < 2000) when the flow rates are between 0.3 to 0.7 liters per minute were analyzed
for the oil samples. Based on the statistical analysis, the models for all oils were highly
significant, with very low probability values (P<0.05). The significance meant that there were
only small chances that incorrect predictions could occur because of experimental error or noise
factors. The models terms in the equations are those that are left after the insignificant variables
and interactions have been eliminated. Decomposing the models sum of squares into several
sources, significant terms were verified based on individual P-values of 0.05. The results from
this study showed that, Tc, Tb, Iw, TcTb, TcIw, TbIw, Tc2, Tb2,and Iw2, are significant judging by
their individual P-value less than 0.05, while Q, TcQ, TbQ, IwQ and Q2, are insignificant as their
P-values are more than 0.05.
This confirms the assumptions in most of the available wax deposition models that the shear
dispersion and removal effects which depend on flow rate are inconsequential in the laminar
flow regime. The lack-of-fit test F-tests were used to test the model’s (regression equations)
151
adequacies. The lack-of-fit results were not statistically significant as the P-values ranges from
0.106 to 0.365. Adequate precision is a measure of the range in predicted response relative to its
associated error i.e. , a signal-to noise ratio and the desire value is equal or more than 4 (Mason
et al., 2003).
The analyses of variance (ANOVA) results for the four variables for each of the Crude oil A-G
in laminar flow regime in both coded and un-coded units are given in Appendix 10.
The model equation for oil A in coded units is then given as (equation 64):
V  113.825  18.583Tc  21.250Tb  24.25I w  18.50TcTb (62)

 7.250Tc I w  9.25Tb I w  5.25Tc2  5.50Tb2  6.50 I w2
The model coefficients of each of the terms for Oil B to G are given in Table 22.
Table 22: The Models equations coefficients for oil A-G in un-coded units for laminar flow
regime.
Coefficients
Term Oil A Oil B Oil C Oil D Oil E Oil F Oil G
Constant, α0 305055 46495 480255 454042 425763 40402 314206
1 3
Coolant temp., (Tc), α1 -695.0 -1054 -972.29 -894.51 -443.08 -594.7 -655.46
Oil temp., (Tb), α2 -1260 -19241 -2089 -1999.1 -2229.5 -1944 -13506
W.I. Percent, (Iw), α3 -2036 -2419 -2482 -2133.9 -728.22 -2109 -1770
Flow rate, (Q), α4 1198.0 2019.6 1973.2 -350.04 -927.08 -4658 162.41
Cool. temp.* Cool. temp., α11 0.1983 0.3758 0.2766 0.3125 0.1183 0.3395 0.1725
Oil temp. * Oil temp., α22 1.1458 1.8177 2.1041 2.1093 2.9270 2.4349 1.3281
W.I. Perc. * W.I. Perc., α33 6.3333 10.770 6.5416 10.562 8.8333 8.6145 8.0625
Flow rate * Flow rate, α44 195.83 527.08 791.66 606.25 670.83 586.45 443.75
Cool. temp. * Oil temp., α12 1.7875 2.5750 2.5000 2.1875 1.1500 1.2187 1.7000
Cool. temp. * W.I. Perc., α13 1.5250 2.2000 2.2250 2.0000 1.1000 1.6625 1.8500
Cool.t temp. * Flow rate, α14 -1.000 -0.750 -2.2500 4.5000 1.2500 2.6250 4.0000
Oil temp. * W.I. perc., α23 4.8750 5.3750 5.6250 4.6875 1.1875 4.9062 3.7500
152
Oil temp. * Flow rate, α24 -3.125 -6.875 -6.2500 -4.3750 0.000 10.312 -5.000
W.I. perc. * Flow rate, α34 -1.250 3.750 -3.750 0.000 -10.000 6.8750 -5.000
The results of the optimum values for each variable and the resulting minimum response (wax
deposit volume) for all the crude oils samples were obtained.
4.28: Interaction Plot of Wax Deposition Process (Laminar Flow) for Crude-Oil A
Since three of the four factors (variables) are significant, there is probability of the interactions
The interaction plot of the wax deposition process (four variables) in subsea pipeline is as shown
in Figure60.
Interaction Plot of Wax Deposit for Oil A (4 Components)

Data Means
321 323 325 327 329 0.3 0.4 0.5 0.6 0.7
200
Coolant
temp.
277
150 282
Coolant temp.
287
292
100 297
200
Oil temp.
321
323
150
325
Oil temp.
327
329
100
200
W.I.
percent
1
150 2
W.I. percent
3
4
100 5
200
Flow
rate
0.3
150 0.4
Flow rate
0.5
0.6
100 0.7
277 282 287 292 297 1 2 3 4 5
Figure60: Interaction plot for wax deposition process (cm3)
153
a. Coolant temperature (Tc) and Bulk oil temperature (Tb): both rows indicate some level
of interaction between the two variables:
Row 1: The lines for level -1, 0 and 1 of the coolant temperature increase but at different
rates, while the value of bulk oil temperature increases.
Row 2: The lines for level 0 and 1 of the bulk oil temperature increase but at different
rates, and that of level -1 of bulk oil temperature has no effect (insignificant), while the
value of coolant temperature increases.
b. Coolant temperature (Tc) and the percentage wax inhibitors (Iw): both rows indicate
some level of interaction between the two variables:
Row 1: The lines for level -1, 0 and 1 of the coolant temperature increase but at
different rates, while the value of wax inhibitor percent increases.
Row 2: The lines for level -1, 0 and 1 of the percentage wax inhibitor increase but at
different rates, while the value of coolant temperature increases.
c. Coolant temperature (Tc) and the flow rate (Q): both rows indicate some level of
interaction between the two variables:
Row 1: The lines for level -1, 0 and 1 of the coolant temperature increase but at
different rates, while the value of flow rate increases.
Row 2: The lines for level -1, 0 and 1 of the flow rate increase but at different rates,
while the value of coolant temperature increases.
d. Bulk oil temperature (Tc ) and the percentage wax inhibitor (Iw): both rows indicate
Row 1: The lines for level -1, 0 and 1 of the bulk oil temperature increase but at
different rates, while the value of percentage wax inhibitor increases.
154
different rates, while the value of bulk oil temperature increases.
e. Bulk oil temperature (Tc ) and the flow rate (Q): both rows indicate some level of
Row 1: The lines for level -1, 0 and 1 of the bulk oil temperature increase but at
while the value of bulk oil temperature increases.
f. Percentage wax inhibitor (Iw ) and the flow rate (Q): both rows indicate some level of
while the value of percentage wax inhibitors increases.
The lines for level 1 of the xylene increase, while the line for level -1 decrease at different rates
and the level 0 of the xylene decreases and then increases, while the volume of AEP increases.
4.29: Response Surface Plots (Laminar Flow Regime)
In the laminar flow regime (Re < 2000), the surface and the contour plots from the experimental
data analyses are as shown in Figures61 –72.
The 3D response surfaces and two dimensional contour lines were based on the individual crude
oil model with two variables kept constant at their coded zero levels, while varying the other two
variables within the experimental range. The figures show the interactive effect of two of the
155
variables: Coolant temperature, oil temperature, Wax inhibitors percentage and the flow rates on
each other keeping the other two variables at their coded zero level.
Figures 61 and 62 show the interactive effect of coolant temperature and oil temperature on each
other keeping the other two variables; Wax inhibitor percentage and flow rate at their zero level
of 3% and 0.5 liters respectively the results show that the volume of deposited wax decreases to
the 96 cm3 as the coolant and oil temperature increases then begins to increase again to a wax
volume of 312 cm3. This is due to higher temperature difference between the two temperatures
leading to more pronounced effect of molecular diffusion of wax particles towards the pipe wall.
Also though the effect of coolant temperature on wax deposition was insignificant at low bulk oil
temperature the effect was more prominent as the bulk oil temperature increases, this is due to
the higher rate of molecular diffusion between the bulk oil and the pipe wall at higher radial
temperature gradient leading to more wax deposit on the pipe wall.
S ur f a c e P l o t o f W a x D e po s i t ( c m3 ) f o r O i l A V s .C o o l a nt te m p., O il T e m p.
H o ld V alu es
W .I . p er c en t 3
F lo w r ate 0.5
300
W a x D e p . ( cm 3 ) 200
100 3 2 7 .5
3 2 5 .0
O i l T e m p .( K )
3 2 2 .5
280
285
290 3 2 0 .0
295
C o o l a n t T e m p .( K )
Figure61: Surface Plot for Oil A (Coolant Temp. Vs Oil temp.) in laminar flow regime
156
Contour Plot of Wax Deposit (cm3 ) for O il A Vs.Coolant Temp., O il Temp.
329
W .D (O Il A )
< 100
328 100 – 150
150 – 200
327 200 – 250
250 – 300
Oil Temperature (K)
> 300
326
H o ld Valu es
W ax Inh ib ito r P erce. 3
325 F lo w rate 0.5
324
323
322
321
280 284 288 292 296
Coolant Temperat ure (K)
Figure62: Contour plot for Oil A (coolant Temp. vs Oil temp.) during laminar flow regime
Figures63 and 64 show the interactive effect of coolant temperature and wax inhibitor percentage
on each other keeping the other two variables; oil temperature and flow rate at their zero level of
325K and 0.5 liters/min respectively.The results show that the volume of deposited wax
decreases to the 87 cm3 as the coolant and wax inhibitor percentage increases then begins to
increase again up to a wax volume of 264 cm3 as the increase in the two variable continues. This
shows that there exist optimum values of wax inhibitor percentage and coolant temperature that
gives a minimum volume of wax deposit and increasing or decreasing the values of these
variables beyond their optimum value will result in increase in volume of wax deposited.
157
S urface Plot of W ax Deposit (cm3 ) for O il A Vs.Coolant Temp., W ax Inhib. %
H o ld Valu es
O il T emp eratu re 325
F lo w rate 0.5
250
W.D ( cm 3 ) 200
150 5.5
100 4.0
2.5 W a x I n h . P e r ce . ( % )
280
285 290 1.0
295
C o o la n t T e m p e r a tu r e ( K )
Figure63: Surface plot for oil A (coolant temp. Vs wax inhib. Percent) in laminar flow regime
C o nto ur P lo t o f W a x D e po s it ( c m3 ) fo r O il A V s .C o o la nt T e mp.,W a x Inhib.%

5
W .D (O I l A )
< 100
100 – 150
150 – 200
Wax Inhibitor Perce. (%)
4 200 – 250
> 250
H o ld V alu es
O il T em p er atu r e 325
F lo w r ate 0.5
3
1
280 2 84 2 88 2 92 2 96
Co o la nt Te mp e r a t u r e ( K)
Figure 64: Contour plots for oil A (coolant temp. vs wax inhib. Percent) in laminar flow regime
158
Figures 65 and 66 show the interactive effect of coolant temperature and flow rate on each other
keeping the other two variables; Wax inhibitor percentage and oil temperature at their zero level
of 3% and 325K respectively. Though the effect of flow rate on wax deposit is insignificant (P-
value > 0.05) the results show that the volume of deposited wax increases to the 122 cm3 as the
flow rate increases from 0.3 to 0.7 liter/min at different coolant temperatures. Indicating that the
effect of shear dispersion (when increase in flow rate allow more wax particles to be sheared
toward the pipe due to radial velocity gradient created in the fluid flow) is not negligible in the
laminar flow region.
Surface Plot of Wax Deposit (cm3) for Oil A Vs. Coolant Temp., Flow Rate
Ho ld Values
O il Temperature 325
W ax Inhibitor Perce. 3
200
W.D ( cm 3 ) 150
0.75
100
0.60
0.45 F low r a te ( L it./m in.)
280 285
290 0.30
295
C oola nt T e m pe r a tur e ( K )
Figure65: Surface Plot for oil A (Coolant Temp. Vs. Flow Rate) in Laminar flow regime
159
C o n to u r P l o t o f W a x D e p o s i t ( c m 3 ) f o r O i l A V s . C o o l a n t T e m p . , F l o w R a te )
0 .7
W .D ( O I l A )
< 100
100 – 120
120 – 140
0 .6 140 – 160
Flow rate (Lit./min.)
160 – 180
> 180
H o ld V a lu e s
O il T em p e r a tu r e 325
0 .5 W a x I n h ib ito r P e r c e . 3
0 .4
0 .3
280 284 288 292 296
Co o la n t Te mp e r a t u r e ( K)
Figure 66: Contour Plot for oil A (Coolant Temp. Vs Flow Rate) inLaminar flow regime
Figures67 and 68 show the interactive effect of Wax inhibitor percentage and oil temperature on
each other keeping the other two variables; coolant and flow rate at their zero level of 287K and
0.5 liters respectively.The results show that the volume of deposited wax decreases to the 87 cm3
as the wax inhibitor percentage and oil temperature increases up to a wax volume of 268 cm3.
This is due to higher temperature difference between the two temperatures leading to more
pronounced effect of shear diffusion of wax particles toward the pipe’s wall. This effect
overwhelmed the wax deposit reducing capability of the wax inhibitors. Simultaneous variation
of the wax inhibition percentage between 1 – 5% and the bulk oil temperature shows that the
wax deposit reduces from 137cm3 to about 87cm3 before it increases to about 273cm3,
confirming that the optimum percentage of wax inhibitor occur for minimum wax deposition.
160
S urface Plot of W ax Deposit (cm3 ) for O il A Vs. O il Temp., W ax Inhib. P erc.
H o ld Valu es
C o o lan t T emp er atu re 287
F lo w rate 0.5
250
W .D ( c m 3 ) 200
150 5.5
10 0 4.0
2.5 W a x I n h ib ito r P e r c e .( % )
32 0.0
3 22.5 1.0
325.0
3 27.5
O il T e m p e r a tu r e ( K )
Figure67: Surface plot for Oil A (Oil Temp. vs Wax Inhib. Percent) in laminar flow regime
C o nto ur P l o t of W a x D e pos i t ( c m3 ) fo r O i l A V s . O il Te mp., W a x Inhib. % .

5
W .D (O I l A )
< 100
100 – 150
150 – 200
Wax Inhibitor Perce.(%)
200 – 250
4
> 250
H o ld V alu es
C o o lan t T em p er atu r e 287
F lo w r ate 0.5
3
1
321 322 3 23 3 24 3 25 3 26 32 7 32 8 32 9
Oil Te mp e r a t u r e (K)
Figure68: Contour plot for Oil A (Oil Temp. vs Wax Inhib. Percent) during laminar flow regime
161
Figures 69 and 70 show the interactive effect of flow rate and oil temperature on each other
keeping the other two variables; coolant temperature and Wax inhibitor percentage at their zero
level of 287K and 3% respectively.Though insignificant (P-value > 0.05) the results show that
the volume of deposited wax increases from 72 cm3 as the flow rate and oil temperature
increases up to a wax volume of 296 cm3. This is due to dual effect of higher temperature
difference between the coolant and oil temperatures and the shear dispersion effect leading to
more pronounced effect of shear diffusion/dispersion of wax particles toward the pipe wall. The
effect of flow rate on wax deposition was also confirmed to be insignificant in the laminar flow
regime (0.3 – 0.7 liter/min)
S ur f a c e P l o t o f W a x D e p o s i t ( c m 3 ) f o r O i l A V s .O i l T e m p . , F l o w R a te
H o ld V a lu es
C o o lan t T em p er atu r e 287
W a x I n h ib ito r P e r c e . 3
200
W .D ( c m 3 ) 1 5 0
0.7 5
100
0.60
0.45 F lo w r a te ( L it ./m in )
320 .0
3 22.5 0 .30
325 .0
327 .5
O il T e m p e r a tu r e ( K )
Figure69: Surface plot for Oil A (Oil Temp. vs Flow rate) in laminar flow regime
162
C o n to u r P l o t o f W a x D e p o s i t ( c m 3 ) f o r O i l A V s . O i l T e m p . , F l o w r a te
0 .7
W .D ( O I l A )
< 80
80 – 100
100 – 120
0 .6 120 – 140
140 – 160
160 – 180
> 180
H o ld V alu es
0 .5 C o o lan t T e m p er atu r e 287
W a x I n h ib ito r P e r c e. 3
0 .4
0 .3
321 322 323 324 325 326 327 328 329
O il Te m p e r a t u r e ( K)
Figure70: Contour plot for Oil A (Oil Temp. Vs Flow rate) during laminar flow regime
Figures71 and 72 (P-value >0.05) show the interactive effect of Wax inhibitor percentage and
flow rate on each other keeping the other two variables; coolant and oil temperature at their zero
level of 287K and 325K respectively.The results show that the volume of deposited wax
increases from 95 cm3 as the flow rate and Wax inhibitor percentage increases up to a wax
volume of 296 cm3. This is due to overwhelming effect of shear dispersion at this temperature
range. Also the effect of flow rate was observed to be insignificant in the laminar flow region.
163
Surface Plot of Wax Deposit (cm3) for Oil A Vs. Wax Inhib. Perc., Flow Rate
Ho ld Values
C oo lant Temperature 287
O il Temperature 325
200
W.D ( cm 3 ) 150
0.75
100
0.60
0.45 F low r a te ( L it./m in.)
1.0
2.5 0.30
4.0
5.5
Wa x I nhibitor P e r ce .( % )
Figure71: Surface plot for Oil A (Wax Inhib. Perc. Vs. Flow Rate) in Laminar flow regime
C onto ur P lot of W a x D e pos it (c m3 ) fo r O il A V s .W a x Inhi b. P e r c ,. F lo w r a te )

0.7
W .D (O Il A )
< 100
100 – 120
120 – 140
0.6 140 – 160
160 – 180
180 – 200
> 200
H o ld V alu es
0.5 C o o lan t T em p er atu r e 287
O il T em p er atu r e 325
0.4
0.3
1 2 3 4 5
W a x Inhibit o r Pe r c e . (%)
Figure72: Contour plot for Oil A (Wax Inhib. Perc. Vs. Flow Rate) during laminar flow regime
164
In these experiments in laminar flow regime (Re < 2000) the interactions between the variables
for other oil samples (B to G) were observed to be similar to that of oil A. While the Tc, Tb, Iw,
TcTb, TcIw, TbIw, Tc2, Tb2and Iw2, are significant judging by their individual P-value less than
0.05, the effect of Q, TcQ, TbQ, Q2 and IwQ, are insignificant as their P-values are more than
0.05.The others contour plots of variables interaction of wax deposition for crude oils B to G in
laminar flow regime are given in Appendix 5.
4.30: Residual Plot of Wax Deposition for Crude Oil A in Laminar Flow Regime
against observation order plots for the wax deposition for Oil A in crude oil flow in sub-sea
straight pipeline is as shown in Figure 73.
Residual Plots for W.D (cm3) OIL A

99
4
90
2
Residual
Percent
50 0
10 -2
1 -4
-5.0 -2.5 0.0 2.5 5.0 100 150 200 250
Hist ogram Versus Order

6.0 4
4.5 2
Frequency
Residual
3.0 0
1.5 -2
0.0 -4
-4 -2 0 2 4 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Figure 73 : Residual plots of Wax Deposition for Oil A in Laminar flow regime
165
deposition (cm3) of crude oil flow in sub-sea straight pipeline to acceptable degree of accuracy.
The residual plots of other wax deposition for crude oil B to G are given in Appendix 6.
4.31: Models optimization (Laminar flow)
The Wax deposition models (regression equations) for the oil samples built with codified
unitsare as given in Table 23.
Table 23: The regression equations (Models) coefficients for oil A-G (Laminar flow).
Coefficients
Constant 305055 464951 480255 454042 425763 404023 314206
A: Coolant temp. -695.00 -1054.2 -972.29 -894.51 -443.08 -594.7 -655.46
B: Oil temp. -1260.2 -1924.0 -2089.7 -1999.1 -2229.5 -1944.4 -1350.7
C: W.I. Percent -2036.0 -2419.6 -2482.4 -2133.9 -728.22 -2109.8 -1770.0
D: Flow rate 1198.0 2019.6 1973.2 -350.04 -927.08 -4658.3 162.41
A2: Coolant temp.* Coolant 0.1983 0.3758 0.2767 0.3125 0.1183 0.3395 0.1725
temp.
B2: Oil temp. * Oil temp. 1.1458 1.8177 2.1041 2.1093 2.9270 2.4349 1.3281
C2: W.I. Percent * W.I. 6.3333 10.770 6.5416 10.562 8.8333 8.6145 8.0625
Percent
D2: Flow rate * Flow rate 195.83 527.08 791.66 606.25 670.83 586.45 443.75
AB: Coolant temp. * Oil 1.7875 2.5750 2.5000 2.1875 1.1500 1.2187 1.7000
temp.
AC: Coolant temp. * W.I. 1.5250 2.2000 2.2250 2.0000 1.1000 1.6625 1.8500
Percent
AD: Coolant temp. * Flow -1.0000 -0.7500 -2.2500 4.5000 1.2500 2.6250 4.0000
rate
BC: Oil temp. * W.I. 4.8750 5.3750 5.6250 4.6875 1.1875 4.9062 3.7500
percent
BD: Oil temp. * Flow rate -3.1250 -6.8750 -6.2500 -4.3750 -4.8E-15 10.312 -5.000
CD: W.I. percent * Flow -1.2500 3.7500 -3.7500 -3.4E-14 -10.000 6.8750 -5.000
rate
166
These equations (models) were optimized by settling the partial derivatives of the equations to
zero with respect to the corresponding variables to find the stationary variables.
And since the model equation in coded unit is given as:
V  113.825  18.583Tc  21.250Tb  24.25I w  8.50TcTb

 7.250Tc I w  9.25Tb I w  5.25Tc2  5.50Tb2  6.50 I w2 (63)
To determined the optimum parameters for minimum wax deposition, the following
computations were performed:
5.25 18.50 7.25  0.50 

 2 2 2
 
18.50 5.50 9.20 0.00
B 2 2 
and
7.25 9.20 6.50  0.50 
 2 
 
 0.50 0.00  2.50 2.50 
5.25 9.25 3.63  0.25 

18.58   
21.25  
   18.50 5.50 4.60 0.00 
b B
24.25  
  7.25 9.20 6.50  1.25 
9.59   
 0.50 0.00  2.50 2.50
167
0.059496 0.077495  0.10615 0.04712
 
0.287301  0.01626  0.17044  0.05649
B 
1 
 
 0.58279  0.07483 0.63069 0.257066 
 
 0.57089  0.05933 0.60946 0.647641
The B-1b is then given by
 0.273336
0.318001 
B 1b   
5.338611 
 
9.115944. 
1
The stationary point using the equation x s   B 1b , was solved as:
2
0.13665 
 0.159 
1
x s   B 1b   
2  2.66931
 
 4.55797
and un-coded into the various transformation
Tc  287
0.13665 
5
Tb  325
 0.159 
2
 2.66931  I w  3
Q  0. 5
 4.55797 
0.1
Then the stationary point in term of the natural variables: coolant temperature, bulk oil
temperature, percentage of wax inhibitor in the oil, and flow rate were then determined.
168
The optimum values of each variable in turbulent flow regime for Crude oil A was evaluated
as, Tc = 287.68 K , Tb = 324.68 K , Iw = 0.331 % , Q = 0.44203 lit./min depositing wax at a
rate of 7.40 cm3/hour on the pipe wall.
The results of the optimum values for each variable and the resulting minimum response (wax
deposit volume) for all the crude oils samples are as shown in Table 24.
Experimental verification of the predicted optimum conditions gave an actual volume of wax
deposit of between 60 and 91 cm3 as against wax deposit volume of between 65 and 94 cm3 from
the software. The small error between the predicted and actual volume of wax deposit: with
AAD of 4.80 % indicated that the software models were valid and accurate in representing the
actual experimental values and also in predicting the inhibition of wax deposit within the range
studied.
Table 24: Optimum wax inhibition variables for Oil A-G in the laminar flow regime.
Oil Coolant Bulk Oil Perc. of Flow Rate Opt. wax dep. Wax dep.
Sample Temp. (oC) Temp. WI (%) (Liter/min) vol. software vol. (Exp.),
(oC) (cm )3
cm3
A 287 324 0.3 0.44 78 74
B 280 325 1.4 0.43 94 91
C 291 322 0.7 0.52 90 79
D 289 326 1.5 0.36 83 75
E 282 327 2.1 0.41 73 67
F 288 324 0.9 0.63 65 60
G 282 325 0.4 0.54 75 69
169
4.32: Analyses of Variance (ANOVA) and Model fitting (Turbulent Flow Regime (Re >
3000))
Similarly, the ANOVA results of the model equations for all the crude oil samples (A-G) in the
turbulent flow region ( Re > 3000) when the flow rates were between 1.1 to 1.5 liters per minute
in coded and un-coded units are summarized in Table 25 and 26 respectively. Based on the
statistical analysis, the models for all oils are highly significant, with very low probability values
(P < 0.0001). The analyses of variance (ANOVA) results for the four variables for each of the
Crude oils A-G in turbulent flow regime is given in Appendix 4.
Table 25: Regression equations (Models) coefficients for Oil A – G. (turbulent transition flow),
in coded units.
Coefficients
Constant 114.11 104.18 119.44 97.60 76.71 91.61 105.78
A: Coolant temp. 38.50 45.67 45.67 40.50 25.17 26.03 36.00
B: Oil temp. 42.50 36.83 36.83 46.83 27.00 31.75 36.67
C: W.I. Percent 46.83 50.17 43.17 53.83 42.83 34.08 51.00
D: Flow rate 17.50 21.67 15.33 25.17 14.50 10.75 22.83
temp.
C2: W.I. Percent * W.I. 25.33 43.08 26.17 42.25 35.33 34.46 32.25
Percent
temp.
Percent
AD: Coolant temp. * Flow -2.00 -1.50 -4.50 9.00 2.50 5.25 8.00
rate
BC: Oil temp. * W.I. 39.00 43.00 45.00 37.50 9.50 39.25 30.00
percent
BD: Oil temp. * Flow rate -2.50 -5.50 -5.00 -3.50 0.00 8.25 -4.00
CD: W.I. percent * Flow -0.50 1.5 -1.50 0.00 -4.00 2.75 -2.00
rate
170
And in un-coded unit they are given as:
Table 26: Regression equations (Models) coefficients for Oil A – G. (turbulent flow), in un-
coded units.
Coefficients
Constant 305055 464951 480255 454042 425763 404023 314206
A: Coolant temp. -695.01 -1054.3 -972.29 -894.51 -443.08 -594.71 -655.47
B: Oil temp. -1260.2 -1924.0 -2089.8 -1999.1 -2229.5 -1944.4 1350.8
C: W.I. Percent -2036.0 -2419.7 -2482.5 -2133.9 -728.2 -2109.8 1770.1
D: Flow rate 1198.0 2019.6 1973.3 -350.0 -927.08 -4658.3 162.42
temp.
C2: W.I. Percent * W.I. 6.333 10.77 6.541 10.563 8.833 8.615 8.063
Percent
temp.
Percent
AD: Coolant temp. * Flow -1.00 -0.75 2.25 4.50 1.25 2.63 4.00
rate
BC: Oil temp. * W.I. 4.875 5.38 5.63 4.69 1.18 4.91 3.75
percent
BD: Oil temp. * Flow rate -3.125 -6.875 -6.25 -4.38 0.00 10.31 -5.00
CD: W.I. percent * Flow -1.250 3.750 -3.75 0.00 -10.00 6.88 -5.00
rate
Decomposing the models sum of squares into several sources, and verifying the significant terms
based on individual P-values of 0.05,the P-values of different variables and their interactions for
the crude oil samples in both coded and un-coded units are shown in Table 27.
171
Table 27: P-values in the turbulent flow regime for Oil A – G.
P-values
Constant 0.000 0.000 0.000 0.000 0.000 0.000 0.000
A: Coolant temp. 0.049 0.000 0.000 0.000 0.000 0.000 0.000
B: Oil temp. 0.000 0.000 0.000 0.000 0.000 0.000 0.000
C: W.I. Percent 0.000 0.000 0.000 0.000 0.000 0.000 0.000
D: Flow rate 0.000 0.000 0.000 0.000 0.000 0.000 0.000
temp.
C2: W.I. Percent * W.I. 0.000 0.000 0.000 0.000 0.000 0.000 0.000
Percent
temp.
Percent
AD: Coolant temp. * Flow 0.538 0.674 0.343 0.125 0.502 0.253 0.214
rate
BC: Oil temp. * W.I. 0.000 0.000 0.000 0.000 0.020 0.000 0.000
percent
BD: Oil temp. * Flow rate 0.444 0.138 0.294 0.536 1.000 0.082 0.526
CD: W.I. percent * Flow 0.877 0.674 0.748 1.000 0.289 0.543 0.750
rate
Lack of fit 0.131 0.256 0.365 0.216 0.290 0.269 0.106
The results from this study showed that, Tc, Tb, Iw, Q, TcTb, TcIw, TbIw, Tc2, Tb2, and Iw2 are
significant judging by their individual P-value less than 0.05, while TcQ, TbQ, IwQ and Q2, are
insignificant as their P-values are more than 0.05. Hence the flow rate term (Q) that was shown
to be insignificant during the laminar flow regimes, was now significant (P-value < 0.05) in the
turbulent flow regime.This confirms the importance of shear dispersion and removal in the
turbulent flow regime, when the flow rate is high (Re > 3000).
Hence the model equation for oil A, in un-coded unit is given as:
172
V  305057  694.006Tc  1262Tb  2037.01I w  1198.04Q  1.7875TcTb (64)
2 2 2
 1.525Tc I w  4.875Tb I w  0.198T  1.156T  6.343I
c b w
The analyses of variance (ANOVA) results for the four variables for each of the Crude oil A-G
in turbulent flow regime in both coded and un-coded units are given in Appendix 11.
4.33: Interaction Plot of Wax Deposition Process (TurbulentFlow) for Crude-Oil A
Since all the four factors (variables) are significant, there is probability of the interactions
The interaction plot of the wax deposition process (four variables) in subsea pipeline is as shown
in Figure 74.
Interaction Plot of Wax Deposit (cm3) for Oil A

Data Means
321 323 325 327 329 1.0 1.1 1.2 1.3 1.4
200
Coolant
temp. (K)
277
150 282
Coolant temp. (K)
287
292
100 297
200
Oil temp.
(K)
321
150 323
Oil temp. (K)
325
327
100 329
200
Wax
Inhib.
Perc.
150
(%)
Wax Inhib. Perc. (%)
1
2
100 3
200
Flow4rate
5
(Lit./min.)
1.0
150 1.1
1.2
1.3
100 1.4
277 282 287 292 297 1 2 3 4 5
Figure 74: Interaction plot of wax deposition process (cm3) (Four variables)
173
The same interactions variation between the variables that was observed in the laminar flow
regime was also observed during the turbulent flow regime but with different rate of deposition
relatively lower to that obtain in laminar flow regime).
4.34: Response Surface Plots (Turbulent flow regime)
The contour plots for different variables interactions for crude oil sample-A during the turbulent
flow regime are as shown in Figures 75 – 80. The contour graphs show that there are interactions
between the effluent variables.
Contour Plot of Wax Deposit Vs. Coolant , Bulk Oil Temperature

329
W.D.
(cm3), Oil
328 A
< 100
327 100 – 125
125 – 150
150 – 175
326
Oil temp. (K)
175 – 200
200 – 225
> 225
325
324
323
322
321
280 284 288 292 296
Coolant temp. (K)
Figure75: Contour plot for Oil A (Coolant temp. Vs Bulk oil temp) in turbulent flow regime
174
C ontour P lot of W ax De posit V s. C oola nt Te mper a tur e , W a x Inhib. P e r c.
5
W .D .
(c m3), O il
A
< 100
4 100 – 125
125 – 150
150 – 175
175 – 200
200 – 225
> 225
3
1
280 284 288 292 296
Co o la nt t e mp. (K)
Figure 76: Contour plot for Oil A (Coolant Temp. Vs. Wax inh,b. perc.) in turbulent flow regime
Contour P lot of W ax Deposit Vs.Bulk O il Temper ature, W ax Inhib. P erc.

5
W .D .
(c m3), O il
A
< 100
4 100 – 125
125 – 150
150 – 175
175 – 200
200 – 225
> 225
3
1
321 322 323 324 325 326 327 328 329
Oil t e mp. (K)
Figure77: Contour plots for Oil A (Bulk oil Temp. Vs. Wax inhib. Perc.) in turbulent flow
regime
175
Contour Plot of Wax Deposit Vs.Wax Inhib. Perc., Flow Rate
1.4
W.D.
(cm3), Oil
A
< 100
1.3 100 – 125
125 – 150
150 – 175
175 – 200
200 – 225
> 225
1.2
1.1
1.0
1 2 3 4 5
Figure78: Contour plot for Oil A. (Flow rate Vs. Wax inhib, Perc.) in turbulent flow regime
C ontour P lot of W ax De posit V s. C oola nt Te mper a tur e , F low R a te

1.4
W .D .
(c m3), O il
A
< 100
1.3 100 – 125
125 – 150
150 – 175
175 – 200
200 – 225
> 225
1.2
1.1
1.0
280 284 288 292 296
Co ola nt t e mp. (K)
Figure79: Contour plot for Oil A. (Flow rate Vs. Coolant temp.) in turbulent flow regime
176
Contour Plot of Wax Deposit Vs.Bulk Oil Temperature, Flow Rate
1.4
W.D.
(cm3), Oil
A
< 100
1.3 100 – 125
125 – 150
150 – 175
175 – 200
200 – 225
> 225
1.2
1.1
1.0
321 322 323 324 325 326 327 328 329
Oil temp. (K)
Figure80: Contour plot for Oil A. (Flow rate Vs. Bulk oil temp.) in turbulent flow regime
In the turbulent flow regime ( Re > 3000) the interactions between the bulk oil temperature and
the coolant temperature, coolant temperature and percentage of wax inhibitors in the crude oil,
and the crude oil bulk temperature and percentage of wax inhibitor in the crude oil for sample A
were similar to what were observed in the laminar flow regime.
4.35: Residual Plot of Wax Deposition for Crude Oil A in Turbulent Flow Regime
The various residual: Normal probability plot, Residual against Fits, Histogram and Residual
against observation order plots for the wax deposition for Oil A in crude oil turbulent flow in
sub-sea straight pipeline is as shown in Figure 81.
177
Residual Plots for Oil A (4 Components, Turb.)
99
4
90
2
Residual
Percent
50 0
10 -2
1 -4
-5.0 -2.5 0.0 2.5 5.0 100 150 200 250

6.0 4
4.5 2
Frequency
Residual
3.0 0
1.5 -2
0.0 -4
-4 -2 0 2 4 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
Figure 81: Residual plots of Wax Deposition for Oil A in Turbulent flow regime
deposition (cm3) of crude oil flow in sub-sea straight pipeline during turbulent flow regime to
acceptable degree of accuracy.
The residual plots of other wax deposition for crude oil B to G flow in subsea straight pipeline
during turbulent flow regime are given in Appendix 7.
4.36: Models Optimization (Turbulent Flow Regime)
Optimization of the Wax deposits models (regression equations) for the crude oil samples built
with codified factors in the Table 30 by settling the partial derivatives of the equations to zero
with respect to the corresponding variables.
178
The optimum values for each variable and the corresponding minimum response (wax deposit
volume) for all the crude oils samples were determined by equating the partial derivatives to
zero.
And in coded unit its given as:
V  114.108  38.50Tc  42.50Tb  46.833I w  17.50Q  71.50TcTb

 30.50Tc I w  39.00Tb I w  19.833Tc2  18.333Tb2  25.33I w2
(65)
To determined the optimum parameters for minimum wax deposition, the following
computations were performed:
19.833 71.50 30.50  2.00 

 2 2 2 
 
71.50 18.833 39.00  2.50 
B 2 2 
and
30.50 39.00 25.333  0.50 
 2 
 
 2.00 11.438  11.438 40.925
19.833 35.75 15.25  1.00 

38.500  
42.500  
71.50 18.833 19.50  1.25 
b   B
46.833  
  30.50 39.00 25.333  0.50
17.500   
 2.00  2.50  0.50 7.833 
0.019254 0.020374  0.02718 0.004842 

 
0.080643  0.00223  0.04666 0.008451 
B 
1 
 
 0.14712  0.02107 0.143911  0.01755
 
0.021263 0.003146  0.01265 0.130478 
179
0.419096 
0.972553 
B 1b   
 0.12703
 
2.643439 
1
2
 0.20955
 0.48628
1
x s   B 1b   
2 0.063516 
 
 1.32172 
Tc  287
 0.20955 
5
Tb  325
 0.48628 
2
0.0663516  I w  3
Q  1.3
 1.32172 
0.1
temperature, percentage of wax inhibitor in the oil, and flow rate were then determined.
180
The optimum values of each variable in turbulent flow regime for Crude oil A was evaluated as,
Tc = 285.95 K , Tb = 324.03 K , Iw = 3.06635 % , Q = 1.1678 lit/min depositing wax at a rate of
6.50 cm3/hour on the pipe wall.
Experimental verification of the predicted optimum conditions for the crude oil A to G gave an
actual volume of wax deposit of between 47 and 78 cm3 as against wax deposit volume of
between 50 and 82 cm3 from the software (Table 28).
Table 28: Optimum wax inhibition variables for Oil A-G in turbulent flow regime.
Oil Coolant Bulk Oil Perc. Flow Rate Min.. wax Wax AAD
Sample Temp. Temp. of WI (Liter/min) dep. vol. dep. vol. (%)
(oC) (oC) (%) software (Exp.),
(cm3) cm3
A 286 324 3.0 1.17 68 65 5.41
B 281 326 3.7 1.36 82 78 5.13
C 289 322 4.1 1.18 74 71 4.23
D 293 327 4.3 1.27 77 74 4.05
E 296 322 2.7 1.33 50 47 6.38
F 290 323 2.2 1.46 55 53 3.77
G 288 328 3.2 1.29 67 63 6.35
The small error between the predicted and actual volume of wax deposit: with overall AAD of
5.05% indicated that the software models were valid and accurate in representing the actual
experimental values and also in predicting the inhibition of wax deposit within the range studied.
181
Table 29: Range of minimum wax deposit (cm3)at different flow regimes for different Crude oils
Crude oil grade Laminar flow Turbulent flow

Light (APIo> 30) 69 - 91 63 - 78
Medium (20 < APIo< 30) 60 - 67 47 - 53
It was observed (Table 29) that the wax deposits are more in the light crude oils (APIo> 30),
ranging between 69cm3 to 91cm3 in the laminar flow regime, and 63cm3 to 78cm3 in the
turbulent flow regime, compared to the wax deposit in the medium crudes (20 < APIo< 30)
which is between 60cm3 to 67cm3 in the laminar flow regime, and 47cm3 to 53cm3 in the
turbulent flow regime. Also the shear removal effect in the turbulent flow regime is more in the
light grade crude oil due to the soft nature of wax deposit formed by these crude grades which
are easier to be sheared off, compared to hard wax deposit formed in the medium crude grade
which is harder to be sheared off.
In the turbulent transition regime ( Re > 3000) the interactions between the bulk oil temperature
and the coolant temperature, coolant temperature and percentage of wax inhibitors in the crude
oil, and the crude oil bulk temperature and percentage of wax inhibitor in the crude oil for
sample A were similar to what were observed in the laminar flow regime, except the flow rate
term, Q that is significant in addition to the Tc, Tb, Iw, TcTb, TcIw, TbIw, Tc2, Tb2 and Iw2 terms as
their P-value are lower than 0.05 (P < 0.05), while the TcQ, TbQ, IwQ and Q2 terms are
insignificant judging by their P-value of above 0.05 (P > 0.05).
182
4.37: Modeling of General Wax Deposition Process in Laminar Flow Regime (Five
Variables)
The wax deposit inhibition process during the flow of single phase fluid in sub-cooled pipeline
was modeled by RSMCCD using four identified wax deposit inhibition variables: Bulk oil
temperature, coolant temperature, flow rate and the fraction of wax inhibitor in the oil and the
wax content (Wc) which depends on individual oil as the fifth variable.
Constraining the five wax deposition process variables within the following values:
Coolant/pipeline temperature (277 - 298 K), Oil bulk temperature, (321 - 329 K), oil flow rate,
(0.2 - 0.7 liter/min (Re < 2000)), the percentage of wax inhibitors added to the oil (1 – 5 %) and
the wax content, (28 – 40 %), the software was first used to determine the minimum number of
variable combinations required to evaluate the optimum condition for wax deposit inhibition for
each of the oil. Then, experiments were performed with variable combinations.
A total of fifty-four (54) experiments were conducted , with Forty-nine (49) experiments
formulated in a factorial design (including thirty-six (36) factorial points, twelve (12) axial points
and 1 center point) the remaining six (6) experiments involving the repetition of the central point
to obtain the good estimate of the experimental error. The response selected for analysis was the
volume of deposited wax in cubic-centimeters.
second order and interaction terms.
183
From the P-values the result shows that the terms Tc, Tb, Iw, Wc, Tc2, Tb2, Iw2, Wc2, TcTb, TcIw,
TbIw, and QWc have significant influence on the wax deposition rate. While the effect of Q, TcQ,
TcWc, TbQ, IwQ, Q2 and IwWc are insignificant.
The model equation in un-coded units is given as:
V  485226 565.48Tc  2392.28.Tb  891.81I w  475.74Wc  0.252Tc2 (66)

 2.96Tb2  8.7I w2  1023.13Q2  0.99Wc2  1.33TcTb  2.28Tc I w  0.0469TcWc
 4.57Tb I w  1.237TbWc  19.531QWc
While, the model equation in coded units is given as:
V  87.774 30.50Tc  47.41.Tb  22.851I w 13.80Q  39.534Wc  25.18Tc2  47.38Tb2 (67)

 34.80I w2  40.93Q2  35.70Wc2  53.31TcTb  45.687Tc I w  2.813TcWc
 11.44Tb I w  29.687TbWc  23.44QWc
Where V represents the predicted response (volume of deposited wax), Tc is the cooling water
temperature in Kelvin, Tb is the bulk oil temperature in Kelvin, Iw is the percentage of wax
inhibitor, WI (group of AEP/xylene/butyl acrylate mixture) in the crude oil sample, Q the oil/WI
flow rate and Wc the wax content percentage.
The analyses of variance (ANOVA) results for the five variables for Crude oil flow in sub-sea
straight pipeline in laminar flow regime in both coded and un-coded units are given in Appendix
12.
184
4.38: Analyzing the Stationary Point of Wax Deposition Process in Laminar Flow (Five
Variables)
Since the model equation for crude-oil A can be represented in matrix form by:
 
y   o  x' b  x ' B (68)


y
 b  2 Bx  0
x (69)
1
x s   B 1b
2 (70)
The calculation is as follows:
For crude-oil A since, the model equation in coded units is,
V  87.774 30.50Tc  47.41.Tb  22.851I w  39.534Wc  25.18Tc2  47.38Tb2 (71)

 34.80I w2  40.93Q2  35.70Wc2  53.31TcTb 11.31Tc I w  11.44Tb I w  23.44QWc
185
25.175 53.313  45.687  11.312  2.813 
 2 2 2 2
 
53.313 47.384 36.563 11.438 29.687 
 2 2 2 
B   45.687 36.563 34.80  11.438 11.563 
 2 2  and
 23.437 
 11.312 11.438  11.438 40.925
2
 
 2.813 29.687 11.563 23.437 35.701 
 
25.175 26.657  22.844  5.656  1.407 

 
 30.50 53.313
 47.41 47.384 18.282 5.719 14.844 
 
   
b   22.85 B   45.687 36.563 34.80  5.719 5.782 
   
13.80   11.312 11.438  11.438 40.925 11.718 
39.53   
 2.813 29.687 11.563 23.437 35.701
 
 0.00496 0.012512  0.00936  0.00187  0.00327 

 
0.016343 0.004001 0.012416 0.006357  0.00512 
 
1  
B   0.02681 0.016593 0.005202  0.00032  0.00869
 
 0.01465 0.0011296  0.00117 0.028994  0.0146 
 
0.004322  0.01513 0.01198  0.02436 0.044409 
 
 0.38301
 1.08639 
 
B 1b   0.43585
 
 0.23895
2.278492 
186
1
2
0.19151 
0.5432 
1  
xs  B 1b  0.21793 
2  
0.11948 
 1.139246
Tc  287
0.19151  (72)
5
Tb  325
0.5432  (73)
2
0.21793  I w  3 (74)
Q  0.5
0.11948  (75)
0.1
Wc  34
 1.139246 
3 (76)
temperature, percentage of wax inhibitor in the oil, flow rate, and wax content were then
determined.
187
The optimum values of each variable in laminar flow regime are evaluated as, Tc = 287.95 K , Tb
= 326 K , Iw = 3.21793 % , Q = 0.51195 lit/min , Wc = 30.5823 %, depositing at a rate of 3.4
cm3/hour on the pipe wall.
Similarly, the optimum conditions were determined by using the same computational process for
crude oil B to G in the laminar flow regime. The results are as given in Table 30.
Table 30: Optimum wax inhibition variables for Oil A-G in laminar flow regime. (5 variables)
Oil Coolant Bulk Oil Perc. Flow Rate Wax Wax dep. rate
3
Sample Temp. Temp. of WI (Liter/min) content (software.) cm /hour
(oC) (oC) (%) (%)
A 287.95 326.00 3.21 0.51 30.58 6.9
B 285.62 328.13 3.75 0.36 32.11 6.5
C 289.11 322.46 4.12 0.48 28.46 7.1
D 291.28 327.55 4.30 0.37 29.78 6.7
E 296.47 322.32 2.75 0.43 23.58 5.7
F 294.36 323.95 2.27 0.56 21.61 5.3
G 288.86 328.45 3.29 0.39 30.43 7.5
4.39: Variables Interaction Analyses of Wax Deposition Process
Since four of the five factors (variables) are significant, there is probability of the interactions
The interaction plot of the wax deposition process (five variables) in subsea pipeline is as shown
in Figure 82.
188
Interaction Plot for Wax deposit (cm3)
Data Means
321 325 329 0.8 1.0 1.2 1.4 1.6
Coolant
temp. (K)
300 267
277
200 Coolant temp. (K)
287
297
100 307
Bulk oil
temp. (K)
300 321
325
Bulk oil temp. (K) 200 329
100
Wax
inhib.
300 Perc.
(%)
200 Wax inhib. Perc. (%)
-1
1
100 3
Flow
5 rate
(Lit./min.)
7
300 0.8
1.0
Flow rate (Lit./min.) 200 1.2
1.4
100 1.6
Wax
content
300 (%)
22
200 Wax content (%)
28
34
100 40
267 277 287 297 307 -1 1 3 5 7 22 28 34 40
Figure 82: Interaction plot for wax deposition process (cm3) (Five Variables)
a. Coolant temperature (Tc) and Bulk oil temperature (Tb): both rows indicate some level
Row 1: The lines for level -1,and 0 of the coolant temperature decrease but at different
rates, that of level 1 increases, while the value of bulk oil temperature increases.
Row 2: The lines for level 0 and 1 of the bulk oil temperature decrease and then increase
but at different rates, and that of level -1 of bulk oil temperature decreases, while the
value of coolant temperature increases.
b. Coolant temperature (Tc) and the percentage wax inhibitors (Iw): both rows indicate
189
Row 1: The lines for level -1of the coolant temperature increases, that of level 1
decreases but at different rates, while that at level 0 decreases then increases while the
value of wax inhibitor percent increases.
Row 2: The lines for level -1and 1 of the percentage wax inhibitor decrease but at
different rates, but that at level 0 decreases then increases, while the value of coolant
c. Coolant temperature (Tc ) and the flow rate (Q): both rows indicate some level of
Row 1: The lines for level -1and 1 of the coolant temperature increase but at different
rates, and that of level 0 decreases then increases while the value of flow rate
increases.
Row 2: The lines for level -1 and 1 of the flow rate decrease but at different rates, but
that of level 0 decreases then increases while the value of coolant temperature
increases.
d. Coolant temperature (Tc ) and the wax content (Wc): both rows indicate some level of
Row 1: The lines for level -1and 1 of the coolant temperature increase but at different
rates, and that of level 0 decreases then increases while the value of wax content
increases.
Row 2: The lines for level -1 and 1 of the wax content decrease but at different rates,
but that of level 0 decreases then increases while the value of coolant temperature
increases.
190
e. Bulk oil temperature (Tc ) and the percentage wax inhibitor (Iw): both rows indicate
Row 1: The lines for level 0 and 1 of the bulk oil temperature decrease and then
increase but at different rates, but that at level -1 decrease, while the value of
percentage wax inhibitor increases.
Row 2: The lines for level 0 and 1 of the percentage wax inhibitor decrease, but at
different rates, and that at level -1 has no significant effect, while the value of bulk oil
f. Bulk oil temperature (Tc ) and the flow rate (Q): both rows indicate some level of
Row 1: The lines for level -1, 0 and 1 of the bulk oil temperature decrease and then
increase at different rate, while the value of flow rate increases.
Row 2: The lines for level -1, 0 and 1 of the flow rate decrease but at different rates,
while the value of bulk oil temperature increases.
g. Bulk oil temperature (Tc ) and the Wax content (Wc): both rows indicate some level
Row 1: The lines for level -1, 0 and 1 of the bulk oil temperature decrease and then
increase at different rate, while the value of wax content increases.
Row 2: The lines for level -1, 0 and 1 of the wax content decrease but at different
rates, while the value of bulk oil temperature increases.
h. Percentage wax inhibitor (Iw ) and the flow rate (Q): both rows indicate some level of
191
Row 1: The lines for level -1 and 1 of the percentage wax inhibitor increase but at
different rates and that at level 0 decreases then increases, while the value of flow rate
increases.
Row 2: The lines for level -1and 1 of the flow rate decrease but at different rates, and
that at level 0 decreases then increases, while the value of percentage wax inhibitors
increases.
i. Percentage wax inhibitor (Iw ) and the Wax content (Iw): both rows indicate some
level of interaction between the two variables:
Row 1: The lines for level -1 and 1 of the percentage wax inhibitor increase but at
different rates and that at level 0 decreases then increases, while the value of wax
content increases.
Row 2: The lines for level -1and 1 of the wax content decrease but at different rates,
and that at level 0 decreases then increases, while the value of percentage wax
inhibitors increases.
j. Flow rate (Q ) and the Wax content (Iw): both rows indicate some level of interaction
between the two variables:
Row 1: The lines for level -1 and 1 of the flow rate increase but at different rates and
that at level 0 decreases then increases, while the value of wax content increases.
Row 2: The lines for level 1 of the wax content decreases, that of level -1 has no
significant effect, and that at level 0 decreases then increases, while the value of
percentage wax inhibitors increases.
192
4.40: Modeling of Wax Deposition Process in Turbulent Flow Regime (Five Variables)
Constraining the five wax deposition process variables within the following values:
Coolant/pipeline temperature (277 - 298 K), Bulk oil temperature, (321 - 329 K), oil flow rate,
(1.1 – 1.5 liter/min (2000 < Re < 3000)), the percentage of wax inhibitors added to the oil (1 – 5
%) and the wax content, (28 – 40 %). The software was first used to determine the minimum
number of variable combinations required to evaluate the optimum condition for wax deposit
inhibition for each of the oil. Then, experiments were performed with variable combinations.
A total of fifty-four (54) experiments were conducted , with Forty-nine (49) experiments
formulated in a factorial design (including thirty-six (36) factorial points, twelve (12) axial points
and 1 centre point) the remaining five (5) experiments involving the repetition of the central
point to obtain the good estimate of the experimental error. The response selected for analysis
was the volume of deposited wax in cubic-centimeters.
second order and interaction terms.
From the P-values the result shows that the terms Tc, Tb, Iw, Q, Wc, Tc2, Tb2, Iw2, Q2, Wc2, TcTb,
TcIw, TbIw, and QWc have significant influence on the wax deposition rate. While the effect of
TcQ, TcWc, TbQ, IwQ, and IwWc are insignificant.
The model equation in un-coded units is given as:
193
V  22620.5  29.31Tc  117Tb  153.367I w 15.002Q  25.171Wc (77)
 0.00854Tc2  0.01536Tb2  0.100I w  36.882Q2  0.0633Wc2  0.0757TcTb
 0.1234Tc I w  0.457Tb I w  0.1419TbWc  4.0885QWc
While, the model equation in coded units is given as:
V  144 .38 18.665Tc  3.282Tb  66.26I w  4.994Q  14.846Wc  0.850Tc2 (78)
 2.458Tb2  0.403I w2  1.475Q2  2.279Wc2  3.031TcTb  2.469Tc I w  3.565Tb I w

 3.406TbWc  4.906QWc
Where V represents the predicted response (volume of deposited wax), Tc is the cooling water
temperature in Kelvin, Tb is the bulk oil temperature in Kelvin, Iw is the percentage of wax
inhibitor, WI (group of AEP/xylene/butyl acrylate mixture) in the crude oil sample, Q the oil/WI
flow rate and Wc the wax content percentage.
The analyses of variance (ANOVA) results for the five variables for Crude oil flow in sub-sea
straight pipeline in turbulent flow regime in both coded and un-coded units are given in
Appendix 13.
4.41: Response Surface Analyses of Five Variables Wax Deposition (Turbulent Flow
Regime)
Figure 83 shows the interactive effect of coolant temperature and bulk oil temperature on each
other keeping the other three variables; Wax inhibitor percentage, flow rate and the wax content
at their zero level of 3%, 0.5 liter/min. and 34% respectively the results show that the volume of
194
deposited wax varies between 100cm3 and 400cm3 as the Coolant/pipeline temperature varies
between 277 - 298 K, and Bulk oil temperature, (321 - 329 K) . This is due to higher temperature
difference between the two temperatures leading to more pronounce effect of molecular diffusion
of wax particles toward the pipe’s wall. Also though the effect of coolant temperature on wax
deposition was insignificant at low bulk oil temperature the effect wax more prominent as the
bulk oil temperature increases, this is due to the higher rate of molecular diffusion between the
bulk oil and the pipe wall at higher radial temperature gradient leading to more wax deposit on
the pipe wall.
Contour plot of Wax deposit Vs. Coolant, Bulk Oil Temperature. Surface plot of Wax Deposit Vs. Coolant, Bulk Oil Temperature.
Wax
305
deposit
(cm3)
300 < 100
100 – 200
200 – 300
295
Coolant temp. (K)
300 – 400
> 400
290
450
285
Wax deposit (cm3) 300
280
150
300
275 0 290
320.0
270 322.5
325.0
270
327.5
321 322 323 324 325 326 327 328 329 Bulk oil temp. (K)
Bulk oil temp. (K)
Figure 83: Contour and Surface Plots Of Wax Deposit Vs. Coolant, Bulk Oil Temperature.
The results of the variables interactions indicating the variables, the hold variables (fixed
variables) and the variation in their wax deposit volumes are tabulated in Table 31.
195
Table 31: Wax Deposit at different Variables Interaction (Five Variables)
S/N Interacting variables Hold (Fixed) Variables Volume of Deposited

Wax (cm3)
1 Coolant Temperature, Wax Inhibitor Percentage, 100 - 400
Bulk Oil Temperature. Flow Rate
Wax Content.
2 Coolant Temperature, Bulk Oil Temperature, 100 - 300
Wax Inhibitor Percentage. Flow Rate,
Wax Content.
Flow Rate. Wax Inhibitor Percentage,
Wax Content.
Wax Content. Wax Inhibitor Percentage,
Flow Rate,
5 Bulk Oil Temperature, Coolant Temperature, 200 - 400
Wax Inhibitor Percentage. Flow Rate,
Wax Content.
6 Bulk oil Temperature, Coolant Temperature, 100 - 300
Flow Rate. Wax Inhibitor Percentage,
Wax Content.
7 Bulk Oil Temperature, Bulk Oil Temperature, 100 - 300
Wax Content. Wax Inhibitor Percentage,
Flow Rate.
8 Wax Inhibitor Percentage, Coolant Temperature, 200 - 300
Flow Rate. Bulk Oil Temperature,
Wax Content.
9 Wax Inhibitor Percentage, Coolant Temperature, 100 - 300
Wax Content. Bulk Oil Temperature,
Flow Rate.
10 Flow rate, Coolant Temperature, 100 - 400
Wax Content. Bulk Oil Temperature,
Wax Inhibitor Percentage.
The contour and surface plots of other variables interaction are given in figure 84 to 92.
196
Contour plot of Wax deposit Vs. Coolant Temperature, Wax Inhib. Perc. Surface plot of Wax Deposit Vs. Coolant Temperature, Wax Inhib.Perc.
Wax
305
deposit
(cm3)
300 < 100
100 – 200
200 – 300
295
Coolant temp. (K)
300 – 400
> 400
290
450
285
280
150
300
275
0 290
270 0
2 270
4
6
-1 0 1 2 3 4 5 6 7 Wax inhib. Perc. (%)
Wax inhib. Perc. (%)
Figure 84: Contour and Surface Plots of Wax Deposit Vs. Coolant Temperature, Wax Inhibitor
Percentage.
Contour plot of Wax deposit Vs. Coolant Temperature, FlowRate. Surface plot of WaxDeposition(cm3) Vs. Coolant Temperature, FlowRate.
Wax
305
deposit
(cm3)
300 < 100
100 – 200
200 – 300
295
Coolant temp. (K)
300 – 400
> 400
290
450
285
280
150
300
275 0 290
0.8
270 1.2 270
1.6
0.90 1.05 1.20 1.35 1.50 Flowrate (Lit./min.)
Flowrate (Lit./min.)
Figure 85: Contour and Surface Plots of Wax Deposit Vs. Coolant Temperature, Flow Rate.
197
Contour plot of Wax deposit Vs. Coolant Temperature, Wax content. Surface plot of Wax Deposit Vs. Coolant Temperature, Wax Content.
Wax
305
deposit
(cm3)
300 < 100
100 – 200
200 – 300
295
Coolant temp. (K)
300 – 400
> 400
290
450
285
280
150
300
275
0 290
270 25 270
30
35
40
25 30 35 40 Wax content (%)
Wax content (%)
Figure 86: Contour and Surface Plots of Wax Deposit Vs. Coolant Temperature, Wax Content.
Contour plotof WaxdepositVs. BulkOil Temperature, WaxInhib. Perc. Surfaceplotof WaxDeposit Vs. BulkOil Temperature, WaxInhib. Perc.
329
Wax
deposit
328 (cm3)
< 100
327 100 – 200
200 – 300
Bulk oil temp. (K)
300 – 400
326 > 400
450
325
Waxdeposit (cm3) 300
324
150
323 327.5
0 325.0
Bulk oil temp. (K)
322 0
322.5
2
4 320.0
6
321 Waxinhib. Perc. (%)
-1 0 1 2 3 4 5 6 7
Waxinhib. Perc. (%)
Figure 87: Contour and Surface Plots of Wax Deposit Vs. Bulk Oil Temperature, Wax Inhibitor
Percentage.
198
Contour plot of Waxdeposit Vs. BulkOil Temperature, FlowRate. Surface plot of WaxDeposit Vs. BulkOil Temperature, FlowRate.
329
Wax
deposit
328 (cm3)
< 100
327 100 – 200
200 – 300
Bulk oil temp. (K)
300 – 400
326 > 400
450
325
Waxdeposit (cm3) 300
324
150
323 327.5
0 325.0
Bulk oil temp. (K)
322 0.8 322.5
1.2 320.0
1.6
321 Flowrate (Lit./min.)
0.90 1.05 1.20 1.35 1.50
Figure 88: Contour and Surface Plots of Wax Deposit Vs. Bulk Oil Temperature, Flow Rate.
Contour plot of Wax deposit Vs. Bulk Oil Temperature, Wax content. Surface plot of Wax Deposit Vs. Bulk Oil Temperature, Wax Content.
329
Wax
deposit
328 (cm3)
< 100
327 100 – 200
200 – 300
Bulk oil temp. (K)
300 – 400
326 > 400
450
325
324
150
323 327.5
0 325.0
322 322.5 Bulk oil temp. (K)
25
30 320.0
35
40
321 Wax content (%)
25 30 35 40
Wax content (%)
Figure 89: Contour and Surface Plots of Wax Deposit Vs. Bulk Oil Temperature, Wax Content.
199
Contour plot of Wax deposit Vs. Wax Inhib. Perc. FlowRate. Surface plot of Wax Deposit Vs. Wax Inhib Perc., FlowRate.
7
Wax
deposit
6 (cm3)
< 100
5 100 – 200
200 – 300
300 – 400
4 > 400
450
3
2
150
6
1
4
0
0 0.8 0
1.2
1.6
-1 Flow rate (Lit./min.)
0.90 1.05 1.20 1.35 1.50
Figure 90: Contour and Surface Plots of Wax Deposit Vs. Wax Inhibitor Percentage, Flow Rate.
Contour plot of Waxdeposit Vs. Wax Inhib. Perc. , Wax content. Surface plot of Wax Deposit Vs. Wax InhibPerc., Wax Content.
7
Wax
deposit
6 (cm3)
< 100
5 100 – 200
200 – 300
300 – 400
4 > 400
450
3
2
150
6
1
4
0
0 25 0
30
35
40
-1 Wax content (%)
25 30 35 40
Waxcontent (%)
Figure 91: Contour and Surface Plots of Wax Deposit Vs. Wax Inhibitor Percentage, Wax
Content.
200
Contour plot of Wax deposit Vs. FlowRate, Wax content. Surface plot of Wax Deposit Vs. FlowRate, Wax Content.
Wax
deposit
1.50 (cm3)
< 100
100 – 200
200 – 300
1.35 300 – 400

> 400
450
1.20
1.05 150 1.6
0 1.2
0.90 25
30 0.8
35
40
Wax content (%)
25 30 35 40
Waxcontent (%)
Figure 92: Contour and Surface Plots of Wax Deposit Vs. Flow Rate, Wax Content.
4.42: Analyzing the Stationary Point of Wax Deposition Process in Laminar Flow (Five
Variables)
Since the model equation for crude-oil A can be represented in matrix form by:
 
y   o  x' b  x' B (79)


y
 b  2 Bx  0
x (80)
1
x s   B 1b
2
201
From the Crude oil A model equation
0.850 3.031  2.469  1.469 5.156 

 2 2 2 2 
 
3.031 2.458  3.656 0.594  3.406 
 2 2 2
B   2.469  3.656 0.403 1.344  2.531 
and
 2 2
  4.906 
 1.469 0.594 1.344 1.475
2 
 
5.156  3.406  2.531  4.906 2.279 
 
0.850 1.5155  1.2345  0.7345 2.578 

 
18.665  3.031
 3.282 2.458 1.828 0.297  1.703 
 
   
b   66.26 B   2.469  3.656 0.403 0.672  1.2655
   
 4.994  1.469 0.594 1.344 1.475  2.453 
14.846   
5.156  3.406  2.531  4.906 2.279 
 
 1.48639 0.12022  0.48458  1.72582  0.35544 

 
0.76578  0.03711 0.009428 0.996706 0.184058 
 
B   1.13978
1
 0.28206  0.62481  0.8555  0.18923 
 
 
 1.79862 0.261168  0.30061 2.60766  0.74392 
 
0.63043  0.07848  0.23061  1.16951  0.29357
7.31252 
11.54538 
 
B 1b  2.251475 
 
 12.5315
5.25308 
202
1
2
 3.65626 
 5.772692
1 1  
x s   B b   1.125738 
2  
6.26573 
 2.62654 
Tc  287
 3.65626  (81)
5
Tb  325
 5.772692  (82)
2
 1.125738  I w  3 (83)
Q  1.3
6.26573  (84)
0. 1
Wc  34
 2.62654 
3 (85)
temperature, percentage of wax inhibitor in the oil, flow rate, and wax content were then
determined.
203
The optimum values of each variable in the turbulent flow regime are evaluated as, Tc = 268.72
K , Tb = 313.45 K , Iw = 1.874 % , Q = 1.9265 lit/min , Wc = 26.12 %, depositing at a rate of 3.4
cm3/hour on the pipe wall.
Similarly, the optimum conditions were determined by using the same computational process for
crude oil B to G in the turbulent flow regime. The results are as given in Table 32.
Table 32: Optimum wax inhibition variables for Oil A-G in turbulent flow regime. (5 variables)
Oil Coolant Bulk Oil Perc. Flow Rate Wax Wax dep. rate
3
Sample Temp. Temp. of WI (Liter/min) content (software.) cm /hour
(oC) (oC) (%) (%)
A 268.72 313.45 1.87 1.93 26.12 3.4
B 271.43 314.67 2.03 1.57 29.28 4.7
C 273.72 317.89 1.56 2.03 27.31 7.2
D 269.45 316.31 1.72 1.72 30.04 5.7
E 265.34 309.67 1.15 1.63 25.97 3.6
F 264.12 304.65 1.23 1.86 23.65 3.1
G 271.48 312.59 2.14 2.11 31.97 7.3
The results confirm the observation that the wax deposition is more in a light crude oils (Samples
A, B, C, D and G), which are between 3.4 to 7.3 cm3/hour, compared to between 3.1 to 3.6
cm3/hour in the medium crude oils (Sample E and F)
4.43: Model Validation
The coefficient with two factors and others with second order terms show the interaction the
interaction between the two factors and the quadratic effect respectively. The accuracy of the
model developed can be understood by the value of R2, adjusted R2 and standard deviation. The
R2 indicates the ratio between sum of the square (SSR) with total sum of the square (SST) and it
describes up to what extent the model perfectly estimated experimental data point. Based on the
204
R2, it was observed that that the experimental data fitted well because of the predicted R2 of
0.9751 is in reasonable agreement with the adjusted R-Square of 0.9618. The accuracy between
the predicted and experimental values was demonstrated in Figure 93 within the studied range
and Figure 94 outside the studied range.
4.43.1: Experimental versus Predicted Values
Comparison of the experimental and predicted values of the volume of wax deposit both within
and outside the studied range shows a high level of compliance between the two values. Figures
93 and 94 highlight the bar-chart of the experimental and predicted value within and outside the
studied range.The operating parameters chosen to validate the model within the studied range are
tabulated in the Table 33 below:
Table 33: Operating parameters within the studied range used for the model validation
Experimental Coolant Bulk Oil Percentage Flow Rate Wax Content

runs Temperature Temperature Wax (Liters/min)
(K) (K) Inhibitors (%)
1 279 325 3 1.2 29
2 288 321 2 1.3 34
3 280 327 1.5 1.1 36
4 290 323 4 1.4 30
5 298 324 5 1.3 37
6 296 322 4.5 1.4 29
The above operating parameters in Table 33 above were substituted in the developed model to
obtain the predicted values and experiments were performed using the operating parameters to
obtain the experimental values. The experimental and the predicted values for the six
experimental runs are tabulated in Figure 93.
205
300
Volume of Wax Deposit (cm3) Predicted values Experimental value
250
200
150
100
50
0
1 2 3 4 5 6
Experiments runs
Figure 93: Model validation within studied range.
The operating parameters chosen to validate the model outside the studied range are tabulated in
the Table 34 below:
Table 34: Operating parameters outside the studied range used for the model validation
Experimental Coolant Bulk Oil Percentage Flow Rate Wax Content

runs Temperature Temperature Wax (Liters/min)
(K) (K) Inhibitors (%)
1 275 320 6 1.5 25
2 276 315 7 1.7 23
3 274 331 5.5 1.5 43
4 273 330 6.5 1.6 45
5 302 332 8.5 1.8 24
6 300 330 8 1.7 22
Similarly, the predicted and experiment values were compared as shown in Figure 94.
206
300
Predicted values Experimental value
Volume of wax Deposit (cm3)
250
200
150
100
50
0
1 2 3 4 5 6
Experimental runs
Figure 94: Model validation outside studied range
4.43.2: The Residual Plots for General Wax Deposition Model
against observation order plots are as shown in Figure 95.
R e s id u a l P lo ts fo r W a x D e p o s it, ( c m 3 )
N o r m a l P r o b a b ilit y P lo t V e r s u s F it s
99
20
90
10
Residual
Percent
50 0
-10
10
-20
1
- 20 -10 0 10 20 0 100 200 300
R e s id u a l F it t e d V a lu e
H is t o g r a m V e rsu s O rd e r
20
20
10
Frequency
Residual
15
0
10
-10
5
-20
0
-20 -10 0 10 20 1 5 10 15 20 25 30 35 40 45 50
R e s id u a l O b s e r v a tio n O r d e r
Figure 95: Residual plots of general wax deposition model
207
These plots all confirmed the validity of the wax deposition model. And its ability to predict the
experimental data to an acceptable degree.
4.44: Scale-Up of Wax deposition Model
The wax deposition rate from the experimental study needs to be scaled-up to predict the wax
deposition rate in the real pipeline. Several attempts were made to use shear rate, shear stress
and Raynold number as a scaler, but these attempts have been futile (Hsu and Brubaker, 1995).
4.45: Modeling Attempts With Critical Wax Tension
In wax deposition process, the hard wax is first deposited due to initial high heat flux conditions
before the gelled crude oil get embedded into the wax network forming a shoe polish like wax
deposit.
The wax deposition rate on the pipe-wall due to molecular diffusion is given in equation (86).
(Burgler et al.,1981; Bern et al., 1986; Weingarten and Metzner, 1986; Brown et al., 1996):
dW dC
  w ADm
dt dr
 k  dC dT
  w A m 
   dT dr (86)
Where
W = total mass of deposited wax on the pipe-wall
T = time
km = constant
208
ρw = density of deposited wax
μ = fluid viscosity
Dm = Molecular diffusion coefficient of the precipitated wax
A = Pipe wall area available for wax deposition = πDL
L = length of the pipe
dC/dr = radial concentration gradient of the precipitated wax
dC/dT = concentration gradient of the precipitated wax with respect to temperature
dT/dr = radial temperature gradient of the precipitated wax.
The wax deposition rate, expressed as thickness y (milli-meter) of deposited wax in a unit time
(hour), is given as (Hsu and Brubaker, 1995):
dy  1  dW
 
dt   w A  dt
(87)
The temperature gradient (dT/dr) from heat transfer equation is given as:
dT   o C p   dT 
 (VD) 
dr  4k o   dL  (88)
Where,
Cp = Crude oil heat capacity
D = tube internal diameter
209
ko = Crude oil thermal conductivity
L = length of tube
Q = flow rate
V = flow velocity
r = radial direction
ρo = crude oil density
substituting equations (86) and (87) into equation (88):
dy
dt  k   o C p  dC 
   m   
VD  dT    4k o  dT 
dL (89)
 = wax deposition tendency, and it’s a function of oil temperature, oil composition and/or
shear rate
k   o C p  dC 
  F1 T , C ,     m   
   4k o  dT  (90)
Similarly the shear effect is given as:
 k   o C p  dC 
F2 (T , C )   m   
   4k o  dT  (91)
Defining critical wax tension as a force required to break a fixed thickness of wax deposit, in
radial direction its dependent on the radial distance (Hsu and Brubaker, 1995)
210
The shear force per unit length of pipe (called shear force gradient) required to break a fixed
thickness of wax deposit is define as critical wax tension at the temperature for a given crude oil
is given by:
PD1n
Critical wax tension =  K 1n  nV 2n
L (92)
For a given crude oil and temperature, the equation (92) can be re-arranged as:
 PD1n 
 
 L F


V 2n 
F
 PD1n 
 

V 2 n 
M
 L M (93)
The subscript F means full scale, and M means model or laboratory data.
To use the critical wax tension as scaler, then flow velocity in the model tube and full scale
pipeline must be equal, therefore, conducting laboratory wax deposition test at the same flow
velocity and ambient temperature as that of the real field pipeline. The data can be used to scale
up to the pipeline operating conditions to predict the wax deposition rate.
The full wax deposition tendency  is:
  F1 T , C   F2 T , C ,V  (94)
T = temperature
C = composition
V = flow velocity
211
The wax deposition rate in a real pipeline is then given as:
 dy   dT 
    M VD 
 dt  F  dL  F (95)
D = internal diameter of the full scale pipeline.
The wax deposition model can be incorporated into a convectional pipe design program to
calculate wax deposition, effective pipe inside diameter, temperature and pressure profiles in real
oil field pipeline.
4.46: Validation of Developed Model with Field Dataand Comparison with Existing Model
Currently in Used.
The developed wax deposition model is used to study wax deposition profile along an existing
Niger Delta (West Africa) pipeline Hsu et al.,(1985). There is no pigging facility for the field.
The pipeline inlet pressure is going up. The production line is transporting stock tank oil (Oil
BB). The pipeline and operating conditions are listed in the Table 35, below
Table 35: Wax Deposition Operating Conditions for a Niger Delta Pipeline
PARAMETERS VALUES
Oil flow rate, BOPD (ib/hr) 60,000
Pipe inlet Temperature, oF 104
Seabed Temperature, oF 60
Pipeline size, Outside diameter, OD 12.750”
Inside diameter , ID 11.375”
Pipeline length, km 4.53
Pipeline insulation Unburied bare
212
The developed model and Matzainet al. model, 2001implemented in OLGA software currently in
used in the petroleum industry were compared to the result of the real field data from the Niger
Delta (West Africa) subsea pipeline (Hsu et al.,1985). The results are as shown in the Table 36
below.
Table 36: Comparison of Developed Model and Matzain et al. Model with the Real Field Data.
Matzain et Absolute deviation.

Time al. Model Developed Field Absolute deviation (Matzain et al.
(days) (2001) Model data (Developed model) Model, 2001 )
2 65 50 38 0.315789474 0.710526316
4 120 102 95 0.073684211 0.263157895
6 175 153 144 0.0625 0.215277778
8 225 200 186 0.075268817 0.209677419
10 275 247 226 0.092920354 0.216814159
12 317 290 265 0.094339623 0.196226415
14 347 320 300 0.066666667 0.156666667
16 370 340 322 0.055900621 0.149068323
18 385 350 330 0.060606061 0.166666667
20 395 360 340 0.058823529 0.161764706
AAD= 0.095649936 0.244584634
From the Table 36 above, the developed model predicts the field data better than the Matzainet
al., model (2001) judging by the average absolute deviation (AAD) of 9.57% returned by the
model as against the AAD value of 24.46 % returned by Matzain et al. model, 2001 implemented
in OLGA software currently in use in the petroleum industry.
The plots of the field data, the predicted wax volume from the developed model and from
Matzainet al. (2001) model are as shown in Figure 96 below,
213
450
Matzain et al. Model (2001)
Total Volume of Wax Deposit (bbl)
400
Developed Model
350 Field data
300
250
200
150
100
50
0
0 5 10 15 20 25
Time (Days)
Figure 96: Comparison of total wax deposit against time of different models with field data
214
CHAPTER FIVE
5.0: CONCLUSIONS AND RECOMMENDATIONS
5.1: Conclusions
a. Wax inhibitors below the crude oils pour points are generally inactive: as crude oils
below their pour points form emulsions with the inhibitors. This prevent the inhibitors
from dissolving in the oil and, thus, from inhibiting wax deposition.
b. Addition of a butyl acrylate (demulsifier) to the solution of the oil and the inhibitors was
discovered to provide an avenue for enhancement of wax inhibitors at low temperature
range (below the crude oil pour point).
c. Pour pointreduction can be achieved by diluting the crudes with Kerosene and Premium
Motor Spirit (PMS), but large volumes (as much as 60% in total volume) of these
petroleum products are needed to obtain even small reductions in pour point (4oC
reduction), and these are economically unviable.
d. Chemical additives (Acrylate ester co-polymer) can be used to inhibit wax deposition in
waxy crudes, as concentration as low as 5% of these additives in the oil gave a good
performance.
e. The shear dispersion effect on wax disposition constituted just about 10% of total wax
deposit, while the molecular diffusion is responsible for more than 70% of the wax
deposit.
f. A new method of using overall wax inhibition percentage has been developed to evaluate
wax inhibitors performance at field operating conditions.
215
g. The optimization of pour point depressant solvent formulation and wax deposits
inhibition process for some crude oil samples can be successively implemented using
response surface methodology central composite design (RSMCCD) in the Minitab-16
software.
h. The proposed model is valid and accurate in representing the actual experimental values
(with AAD values of less than 6%) and also in predicting the inhibition of wax deposit
within the studied range.
i. The developed model predicts the field data better than the Matzain et al. model, (2001)
judging by the average absolute deviation (AAD) of 9.57% returned by the model as
against the AAD value of 24.46% returned by Matzainet al. model (2001) implemented
in OLGA software.
5.2: Summary of Findings
The summary of the findings in reference to each of the objectives are stated in Table 33.
Table 37: Summary of findings of the study
S/N Objectives Findings

1a To study the effect Experimental studies showed that the wax deposition rate reduces
of temperature and as the temperature difference between the coolant water and bulk oil
flow rate on wax reduces when the coolant temperature is around the crude oils pour
deposition rate. point, and the percentage of wax inhibitors decreases. The wax
deposition rate also decreases as the flow rate increases in the
laminar-turbulent transition flow regime, while the flow rate has no
effect in the laminar flow regime.
1b To evaluate the Pour point reduction can be achieved by diluting the crudes with
influence of Kerosene and Premium Motor Spirit (PMS), but large volumes (as
petroleum fractions much as 60% in total volume) of these petroleum products are
and polymeric needed to reduce
compounds on wax the wax deposit by 40 %, while about 5% solution of acrylate ester
deposition rate. copolymer dissolved in xylene reduced the wax deposit by as much
as 65 %.
216
1c To study the effect Addition of a butyl acrylate (demulsifier) to the solution of the oil
of some and the inhibitors was discovered to provide an avenue for
demulsifiers on enhancement of wax inhibitors judging by the increase in the
wax deposition overall average wax inhibition (WIav) percentage from between 46 –
rate. 65% when AEP and Xylene were used in the formulation compared
to WIav percentage of between 67 – 82% in the formulation
involving the use of AEP, xylene and butyl acrylate
2 To develop A new model relating the wax deposition rate ,(v) to its effluent
empirical models variables has been developed. and in un-coded units is given in
relating the wax equation (77):
deposition rate to
its effluent V  22620.5  29.31Tc  117Tb  153.367I w 15.002Q  25.171Wc
variables.
 0.00854Tc2  0.01536Tb2  0.100I w  36.882Q2  0.0633Wc2  0.0757TcTb
 0.1234Tc I w  0.7344Tc Q  0.457Tb I w  0.1419TbWc  4.0885QWc
3 To determine the The optimum operating parameters returned by variables in the

optimum operating turbulent flow regime (since flow is turbulent during pumping) are
parameters at Tc = 268.72K , Tb = 313.45K , Iw = 1.8743% , Q = 1.9265 lit/min ,
turbulent flow Wc = 26.12%. Depositing wax at the rate of 3.4cm3/hour compared
regime for paraffin to wax deposition rate of 7.6 cm3/hour from Matzainet al. model
wax inhibition (2001).
process in subsea
flow lines
4 Employ the model The developed model predicts the field data better than the Salam et
to field data and al., model (2014) judging by the average absolute deviation (AAD)
compared the of 9.57% returned by the model as against the AAD value of 24.46
model with existing % returned by Matzain et al. model (2001) implemented in OLGA
ones currently in software currently in use in the petroleum industry.
use.
217
5.3: Contribution to Knowledge
1. In applying the response surface methodology central composite design (RSMCCD) in
the Minitab-16 software to minimize wax deposition in crude oil flow in cooled straight
pipeline, an improved wax deposition model has been developed for laminar and
turbulent flow regime in both coded (the level values for the variables) and un-coded
units (the real values of variables).
2. In applying the RSMCCD to wax deposition process, wax deposition rate in flow of
crude oil Sample A in cooled straight pipeline was successively reduced to 3.4cm3/hour
at optimum operating parameters of Tc = 268.72 K, Tb = 313.45 K, Iw = 1.87 %, Q = 1.93
lit/min. and Wc = 26.12 % compared to wax deposition rate of 7.6 cm3/hour from
Matzainet al. model (2001),implemented in OLGA software currently in use in the
petroleum industry.
3. A new method of using average wax inhibition percentage (WIav) to evaluate the wax
inhibitors performance at field operating conditions has been developed.
4. Introduction of butyl acrylate as an additive to acrylate ester copolymer/xylene mixture
was able to enhance the overall average wax inhibition percentage of the mixture from
between 46 – 65% to between 77 – 92% for the seven crude oil samples.
5. The shear stripping effect in wax deposition observed to be significant in the turbulent
flow regime was shown to be insignificant in the laminar flow region as the flow rate
term (Q), its square Q2 and interactions with other variables, TcQ, TbQ and IwQ were of
no significant effect to the response due to their returned P-value above 0.05.
218
5.4: Publications from the Study
1. Adeyanju, O.A. and Oyekunle L.O. (2010). New Multi-Solid Thermodynamic Model for
Improved Cloud Point Prediction of Waxy Crudes, Petroleum and Coal, 52(4), 235-242.
2. Oyekunle, L.O and Adeyanju, O.A. (2011).Thermodynamic Prediction of Paraffin wax

precipitation in Crude oil Pipelines,Petroleum Science and Technology 29, 208-217.
3. Adeyanju, O.A. and Oyekunle, L.O. (2011).Simulation of Paraffin Deposition in

petroleum Reservoirs,Journal of Advances in Petroleum Engineering Sciences (ASPES),
2(3), 79-92.
4. Adeyanju, O.A. and Oyekunle, L.O. (2012). A New Unified Model for Predicting non-
Newtonian Viscosity of Waxy Crudes,Petroleum Science and Technology, 30(9), 904-
914.
5. Adeyanju, O.A. and Oyekunle, L.O. (2012). An Experimental study of Rheological

properties of Nigeria Waxy Crude Oil,Petroleum Science and Technology Journal,
30(11), 1102-1111.
6. Adeyanju, O.A. and Oyekunle, L.O. (2013). Experimental Study of Wax Deposition in a
Single Phase Sub-Cooled Pipelines,Journal of Advances in Petroleum Engineering
Sciences (ASPES), 5(2).
7. Adeyanju, O.A. and Oyekunle L.O.,(2014). Modeling and Simulation of Wax Deposition
in Oil Pipeline System, Journal of Advances in Petroleum Engineering Sciences
(ASPES), 6(1).
8. Adeyanju, O.A. and Oyekunle, L.O. , (2015).Optimization of Wax deposition in a Sub-

Cooled pipeline using Response Surface Methodology,Paper presented at the Society of
Petroleum Engineers Conference and Exhibition, held atEko Hotel and Suites, 1st – 5th of
August, 2015.
219
5.5: Future Studies
The study described in this thesis identified some topics for future consideration. These include:
1. Multiphase Wax DepositionExperiments and Modeling
2. Impact of Surface Roughness on Wax Deposition
3. Single Phase Wax Deposition under Low Thermal Driving Force/Heat Flux Conditions
4. Study on the Restart of Crude-oil Flow in Non-uniformly Gelled pipeline.
The direction the future studies should be directed are introduce in the following sub-section.
5.5.1: Multiphase Wax Deposition Experiments and Modeling
Effort should be directed to study the wax deposition in multiphase flow. Flow in subsea pipeline
typically involves crude oil, water and gas. In these scenarios, single –phase (using dead oil)
modelsare inadequate to model multiphase systems because they do not account for the effects of
the water or gas phases. Recent advances in the field of wax deposition modeling and multi
phase flow have made the study of multiphase wax deposition more attractive.
5.5.2.: Impact of Surface Roughness on Wax Deposition
The delineation point between cohesive and adhesive failure would change if the surface
response is changed (Lee, 2008). The adhesive strength will increase as the roughness increases
because of the greater contact area between wax-oil gel and the wall.
220
5.5.3: Single Phase Wax Deposition under Low Thermal Driving Force/Heat Flux
Conditions
In order to minimize the experimental time, the flow loop experiments were performed with
higher thermal gradient than that that operates in the field. This higher thermal gradient result in
higher thermal driving force and higher degree of super-saturation in the deposit-flow interface
that those that operate on the field. Thus, the precipitation kinetics proven to be successful in the
flow loop system may need to be adjusted to the field pipelines. Therefore, a set of single phase
flow loop experiments with low thermal gradient are proposed. As a result of this low heat flux,
the experiments should be performed for longer period of time (e.g months scale)
5.5.4: Study on the Restart of Crude-oil Flow in Non-uniformly Gelled pipeline.
Wax-oil gel have temperature gradient along the radial direction in the field pipelines. During a
shutdown, the oil in the pipeline cools down from the pipe wall towards the centerline. Since
temperature gradient exists during this cooling process, a concentration gradient is also
established,similar to the scenario during wax deposition under flow conditions. The
concentration gradient drives a diffusive flux of wax molecules towards the wall. As a result, the
solid wax content of the final gel is higher near the pipe wall than near the center. Thus, the yield
stress is not uniform across the cross section. The varying cooling rates across the cross section
also contribute to this non-uniformity in yield stress. Hence efforts should be made to develop a
model to make the adequate scale up of the restart model a reality.
221
REFERENCES
Abney, L., Kalman, M., Hoogerhuis, J. and headworth, C., ((2003). Flow Remediation Solution
for Pipeline,in Offshore Technology Conference, Houston, Texas.
Adewusi, V.A. (1997), Prediction of Wax Deposition Potential of Hydrocarbon Systems from
Viscosity-Pressure Correlations, Energy and Fuel, 76(12), 1079 – 1083.
Adewusi,V.A.: (1998). Waxing Tendencies and Rheological Evaluation of Crude Condensate
Blends for an Offshore Pipeline Transportation,Petroleum Science and Technology, 16(7 & 8),
697 – 717.
Ahmed, T., (1991), A Practical Equation of State” Paper SPE 18532, presented at Eastern
Regional Meeting of Society of Petroleum Engineers Conference and Exhibition, Charleston.
Ahmed, T., (2007), On Equation of State” Paper presented at Latin American and Caribbean
Petroleum Engineering Conference, Buenous Aires, Argentina.
Ahn, S., Wang, K.S., Shuler, P.J., Creek, J.L.and Tang, Y., (2005), Paraffin Crystal and
Deposition Control by Emulsification,SPE International Symposium on Oilfield Chemistry,
Society of Petroleum Engineers, The Woodlands, Texas.
Aiyejina, A.,Chakrabarti, D.P., and Pilgrim, A.,(2011), Wax Formation in Oil Pipelines: A
Critical Review , International Journal of Multi-phase Flow, 37(7), 671-694.
Ajienka, J.A., (1981), The Effect of Temperature on Rheology of Waxy Crude Oils and its
Implication in Production Operation,Ph.D. Dissertation, University of Port Harcourt, Nigeria.
Ajienka, J.A. and Ikoku, C.V., (1997), Waxy Crude Oil Handling in Nigeria: Practices, Problems
and Prospects,Energy Source, 12(4), 463 – 478.
Alana, S.J.D., (2003), Investigation of Heavy Oil Single phase Paraffin Deposition
Characteristics, M.Sc thesis, University of Tulsa.
Alboudwarej, H., Felix, J., Taylor, S., Badry, R., Bremmer, C., Brough, B., Skeates, C. ,Baker,
A., Palmer, D., Pattison, K., Beshry, M., Krawchuk, M., Brown, G., Calvo, R., Triana, J.,
Hathcock, R., Koemer, I., Huges, T., Kundu, D., Cardenas, J. and West, C. (2006), Oilfield
Review 18(2), 34-53.
Al-Yaari, M., (2011), Paraffin Wax Deposition: Mitigation & Removal Techniques,Presented
at the 2011 SPE professionals Technical Symposium held in Dhahran, Saudi-Arabia.
Ashton, J.P., Kirspel, L.J., Nguyen, H.T., and Credeur, D.J.,(1989), In-Situ Heat System
Stimulates Paraffinic-Crude Producer in Gulf ofMexico, SPE Production Engineering, 4(2), 157-
160.
Azevedo, L.F., and Teixeira,A.M., (2003), A Critical Review of the Modeling of Wax
Deposition Mechanism,Petroleum Science and Technology, 21(3 & 4), 393-408.
222
Badwaik, L.S., Prasad, K. and Deka, S.C. (2012), Optimization of Extraction Conditions by
Response Surface Methodology for Preparing Partially Defaulted Peanut, International Food
Research Journal, 19(1), 341-346.
Banki, R., Hoteit, H. and Firoozabadi, A., (2008a), Wax Deposition and Ageing in Flowlines
from Irreversible Thermodynamics,Energy and Fuels, 22, 2693-2706.
Banki, R., Hoteit, H. and Firoozabadi, A., (2008b), Mathematical Formulation and Numerical
Modeling of Wax Deposition in Pipelines from Enthalpy-Porosity Approach and Irreversible
Thermodynamics, International Journal of Heat and Mass Transfer, 51, 3387-3398.
Banki, R. and Firoozabadi, A., (2002), Modeling of Wax Deposition in Pipelines from
Irreversible Thermodynamic, SPE Paper No. 77571, Society of Petroleum Engineers Annual
Conference and Exhibition, San Antonio, Texas.
Barker, K.M., (1989), Formation Damage Related to Hot Oiling, SPE Production Engineering,
4(4), 371-375.
Bello, O., Ademodi, B., Akinyemi, P. (2005), Xylene Based Inhibitors Solves Crude Oil Wax
Problems in Niger Delta Pipeline,Oil and Gas Journal, 14, 56-59.
Bello, O., Fasesan, S., Teoderiu, C., and Reinicke, K., (2006), An Evaluation of the Performance
of Selected Wax Inhibitors on Paraffin Deposition in Nigeria Crude Oils,Petroleum Science and
Technology, 24(2), 195 - 206.
Bern, P.A., Winthers, V.R., and Cairns, J., (1980), Wax Deposition in Pipelines,European
Offshore Petroleum Conference and Exhibition, London, 21-24.
Bhattacharya, A. (1991), Effect of Thermal Environment on Wax Deposition in a Crude Oil
Pipeline: Analysis of some Critical Aspects,Proceedings of the first (1991) International
Offshore and Polar Engineering Conference, Edinburgh, United Kingdom.
Bidmus, H.O. and Mehrotra, A.K., (2009), Solids Deposition during “Cold Flow” of Wax
Solvent Mixtures in a Flow Loop Apparatus with Heat Transfer, Energy and Fuels, 23, 3184-
3194.
Blasius, K.R., Cutts, J.A. and Howard, D., (1982), Topography and Stratigraphy of Martian Polar
Layered Deposit” Icarus, 50, 140-160.
Botne, K.K. (2011),Modeling Wax Deposition with Deposition Release Models, Specialization
Project, Department of Petroleum Engineering and Applied Geophysics, Norwegian University
of Science and Technology, Trondheim, . 38.
Box, G.E. and Draper, N.R., (2007), Response Surface Mixture and Ridge Analyses,Second
Edition, A John-Wiley & Son, Inc., Publication.
Brown, T.S.: Niesen, V.G., Erickson, D.D., (1994),The Effect of Light Ends and High Pressure
on Paraffin Formation, SPE Paper No. 28505, Presented at the 69th SPE Annual Technical
Conference and Exhibition held in New Orlean, LA.
223
Bradley, N. (2007), The Response Surface Methodology, M.Sc. Thesis Submitted to the
Department of Mathematical Sciences, Indiana University South Bend, U.S.A.
Burger, E.D., Perkins, T.K., Striegler, J.H., (1981), Studies of Wax Deposition in the Trans-
Alaska Pipeline,American Chemical Society, 33(6), 1075-1086.
Carbognani, L.: Orea, M.,; Fonseca, M.; (1999), Complex Nature of Separated Solid Phases from
Crude Oils,Energy and Fuels, 13, 351-358.
Carnahan, N.F., (1989), Paraffin Deposition in Petroleum production,Society of Petroleum
Engineer Journal of Petroleum Technology, 1989(10), 1024-1025, 1106.
Chen, S., Oye, G., and Sjoblom,O., (2006), Characterization and Rheological Properties of Waxy
Oil,Ann. Trans. Nordic Rheological Science, 14.
Chow, R.; (2000),Precipitation of Wax From Crude Under the Influence of a Magnetic
Field,Journal of Canadian Petroleum Technology, 39(6).
Creek, J.L., Lund, H.J., Brill, J.P.,and Volk, M., (1999), Wax Deposition in Single Phase
Flow,Fluid Phase Equilibria, 158, 801-811.
Dobbs, J.B., (1999), A Unique Method of Paraffin Control in Production Operation, in SPE
Rocky Mountain Regional Meeting , Society of Petroleum engineer Inc., Gillette, Wyoming,
Duncum, S.N., James, K. and Osborne, C.G., (2000), Wax Deposit Inhibitors , Google Patents.
Dwivedi, P., Sarica, C. and Chang, W., (2012), Experimental Study of Wax deposition
Characteristics of a Waxy Crude oil under Single Phase Turbulent Flow Conditions, Paper no.,
OTC 22953, Presented at the 2011 Offshore Technology Conference, Houston, Texas, USA.
Eaton, F.E. and Weeter, G.Y., ((1976), Paraffin Deposition in Flow Lines Paper 1976,
CSME/CSChE-22 Presented at 16th National Heat Transfer Conference, St Louis.
Edmonds, B., Moorwood, T., Szczepanski, R., and Zhang, X., (2008), Simulating Wax
Deposition in Pipelines for Flow Assurance,Energy and Fuels, 22, 729-741.
Erickson, D.D., Niesen, V.G., and Brown, T.S., (1993), Thermodynamic Measurement and
Prediction of Paraffin Precipitation in Crude Oil,Paper Presented at the 68th Society of
Petroleum Engineers Annual Conference and Exhibition, Houston, Texas, USA.
Esaklul, K.A., Fung, G., Harrison, G., Derego, R.; (2003), Active Heating For Flow Assurance
Control in Deep-Water Flow Lines, in Offshore Technology Conference, Houston, Texas.
Fasano, A., Fusi, L., Correra, S., (2004), Mathematical Models for Waxy Crude Oils, Meccanica,
39, 441-482.
Ferworm, K. A.; Hammami, A. Ellis, H., (1997), Control of Wax Deposition: An Experimental
Investigation of Crystal Morphology and An Evaluation of Various Chemical Solvents,SPE
Paper No. 37240, Presented at the International Symposium on Oil Field Chemistry held in
Houston, Texas, Feb. 18-21.
224
Fredenslund, A., Hansen, J.H., Pedensen, K.S., and Ronningsten, H.P. (1988), A
Thermodynamic Model for Predicting Wax Formation in Crude Oil,AIChE Journal, 34, 1937-
1942.
Fung, G., Backhaus, W.D., McDaniel, S., and Erdogmus, M. (2006): To Pig or not To Pig: The
Marlin Experience with Stuck Pig,Offshore Technology Conference, Houston, Texas.
Gentili, D.O. (2005), Evaluation of Polymeric Phosphoric Ester Based Additives as Wax
Deposition Inhibitors,SPE Latin American and Caribbean Petroleum Engineering Conference,
Society of Petroleum Engineers, Rio de Janeiro, Brazil.
Gjermundsen, I., (2003), A Model for Wax Deposition in Offshore Pipelines,Internal Hydro
report, Porsgrunn.
Gjermundsen, I., (2006), State of the Art: Wax Precipitation, Deposition and Aging in Flowing
Hydrocarbon System, Internal Hydro report, Porsgrunn.
Gjermundsen, I. and Duenas, M. (2006), Wax Deposition: A Comparison between
Measurements and Predictions from a Commercial Model,Proceeding of the 7th International
Conference on Petroleum Phase Behaviour and Fouling, 25-29.
Guo, W., Reigan, P., Siegel D., Zirrolli, J., Gustiafson, D., and Ross, D., (2005), Formation of
17- Allylamina Demethoxygel demycin (17-AAG) hydroquinone by NAD(P)H: Quinone
Oxidoreductase 1: Role of 17-AAG Hydroquinone in Heat Shock Protein 90 Inhibition,Cancer
Research, 62, 10006-10015.
Haq, M.A.: (1981), Deposition of Paraffin Wax from Its Solution with Hydrocarbon, SPE
Paper No. 10541, SPE e-library.
Hammami, A. and Raines, M.A. (1999), Paraffin Deposition from Crude Oils: Comparison of
Laboratory Results with Field Data,: SPE Paper No. 54021, SPE Journal, Vol. 4, 9-18.
Huanquan, P.; Abbas, F.; Per, F. ((1997), Pressure and Composition Effect on Wax
Precipitation: Experimental Data and Model Results,SPE Journal, 12(4), 250-258.
Hayduk, W. and Minhas, B.S., (1980),Correlations for prediction of Molecular Diffusivities in
Liquids,Presented at the 30th Canadian Chemical Engineering Conference, Edmonton,
Alberta.
He, S.S., and Huang, Q.Y., (2014), Research on Wax Deposition in the Pipeline Without
Pigging for a Long Time,Petroleum Science and Technology, 32, 316 – 323.
Hernandez, O.C., Hensly, H. and Serica, C., (2004), Improvements in Single-Phase paraffin
Deposition Modeling,SPE Production and Facilities (11): 237-244.
Hoffmann, R. and Amundsen, L., (2010), Single Phase Wax Deposition Experiments,Energy
and Fuels, 24 (2), 1069-1080.
225
Holder, G.A. and Winkler, J., (1965), Wax Crystallization from Distillate Fuels 1 Cloud and
Pour Phenomena Exhibited by Solution of Binary n-Paraffins Mixture,Journal of Institute of
Petroleum, 51 (228).
Hoveden, L., Ronningsen, H.P., Xu, Z.G., Labes-Carrier, C., and Rydahl, A.,(2004), Pipeline
Wax Deposition Models and Model for removal of Wax by Pigging: Comparison between
Model Prediction and Operational Experience,North American Conference on Multi-phase
Technology, Banff, Canada.
Hsu, J.J. and Brubaker, J.P., (1995),Wax deposition Measurement and Scale-Up Modeling for
Waxy Live Crudes Under Turbulent Flow Conditions,SPE Paper 29976 Presented at the
International Meeting on Petroleum Engineering, Beijing, China.
Hsu, J.J.; Santamaria, M.M.; Brubaker, J.P. (1994), Wax Deposition of Waxy Live Crudes
under Turbulent Flow Conditions, SPE Paper No. 28480, Presented at the SPE 69th Annual
Technical Conference and Exhibition held in New Orleans, LA, USA.
Huang, Z., Lee, H.S., Senra, M. and Fogler, H.S., (2011), Investigation of the Impact of
Composition of Wax Deposit on Pigging Frequency Design,American Institute of Chemical
Engineers, 57(11), 2955-2964.
Hunt, E.B.(1962), Laboratory Study of paraffin Deposition,Journal of Petroleum Technology,
Trans AIME, Vol. 225, 1259.
Jang, Y.H., Blanco, M., Creek, J., Tang, Y., (2007), Wax Inhibition by Comb-like Polymers:
Support of the Incorporation-Perturbation Mechanism from Molecular Dynamics
Simulations,Journal of Physical Chemistry B., 11(46), 13173-13179.
Jessen, F.W. and Howell, J.N.: (1958),Effect of Flow Rate on Paraffin Accumulation in plastic,
Steel and Coated Pipes,Journal of Petroleum Technology, Trans AIME, Vol. 213, 80-83.
Jennings, D.W. and Newberry, M. E. (2008), Paraffin Inhibitor Application in Deep-water
Off-shore Development,International Petroleum Technology Conference held in Kuala-
Lumpur, Malaysia.
Jennings, D.W. and Weispfennig, K. (2005), Effect of Shear and Temperature on Wax
Deposition: Cold-finger investigation with a Gulf of Mexico Crude Oil,Energy and Fuels, 19,
1376-1386.
Jorda, R.M. (1966), Paraffin Deposition and Prevention in Oil Wells,Journal of Petroleum
Technology, Trans AIME, Vol. 237, 1605-1612.
Karan, K.; Ratulowski, J.; German, P.(2000), Measurement of Waxy Crude Properties Using
Novel Laboratory Techniques,SPE Paper No. 62945, Presented at SPE Annual Technical
Conference and Exhibitionheld in dallas, Texas.
Kelechukwu, E.M., Salim, H.S., and Yassin, A.M., (2010), Influencing Factors Governing
Paraffin Wax Deposition during Crude Production,International Journal of Physical Sciences,
5(15), 2351-2362.
226
Kelland, M.A., (2009),Production Chemicals for Oil and Gas Industry,CRC Press, Boca
Raton, Florida.
Khuri, A.I. and Mukhopadhyay, A., (2010), Response Surface Methodology,WIREs

Computational Statistics, (2), 128-149.
Lashkarbolook, M.A., Seyfaee, F., Esmaeilzadeh, D., and Mowla, M, (2010), Experimental
Investigation of Wax Deposition in Kermanshah Crude Oil through a Monitored Flow Loop
Apparatus,Energy and Fuels, 24(2), 1234-1241.
Lee, H.S., (2008), Computational and Rheological Study of Wax Deposition and Gelation in
Subsea Pipeline, A PhD Dissertation submitted to the University of Michigan.
Lee, J., Ye, L., Landen, W.O. and Eitenmillcr, R.R., (2000), Optimization of an Extraction
Procedure for the Quantification of Vitamin E, in Tomato and Broccoli using Response
Surface Methodology, Journal of Food Compos. Anal. 13, 45-57.
Leelavanichkul, P., Deo, M.D., Hanson, F.V., (2004), Crude Oil Characterization and Regular
Solution Approach to Thermodynamic Modeling of Solid precipitation at Low
Pressure,Petroleum Science and Technology, 22, (7/8), 973-990.
Leiroz, A.T., and Azevedo, L.F., (2005), Studies on the Mechanisms of Wax Deposition in
Pipelines, in Offshore Technology Conference, Houston, Texas.
Leiroz, A.T., and Azevedo, L.F., (2007), Paraffin Deposition in a Stagnant Fluid Layer Inside a
Cavity Subject to a Temperature Gradient,Heat Transfer Engineering, 9, 567-575.
Leontaritis, K. J. and Gerouulis, E., (2011), Wax Deposition Correlation-Application in
Multiphase Wax Deposition Models,Paper no. OTC 21623, Presented at the 2011 Offshore
Technology Conference, Houston, Texas, USA.
Lindeloff , N and Krejbjerg, K. (2002), A Compositional Model Simulating Wax Deposition in

Pipeline System,Energy & Fuel, 16, 892 – 903.
Lindeman, O.E. and Allenson, S.J., (2005), Theorectical Modeling of Tertiary Structure of
Paraffin Inhibitors,In SPE 93090 Presented at SPE International Symposium on Oil Field
Chemistry, The Woodlands, Texas, 2-4.
Luo, K.; Li, S.; Zheng, X., Liu, H., Zhong, T., and Zhu, Y. (2001), Phase Behaviour of Highly
Waxy Gas Condensate Systems,SPE Paper No. 68228, Presented at SPE Middle East Oil
Showheld in Bahrain.
Machado, A. L., Lucas, E.F., and Gonzalez, G., (2001), Poly(ethylene-co-vinyl acetate)
(EVA) as Wax Inhibitor of Brazilian Crude Oil: Oil Viscosity, Pour Point and Phase Behaviour
of Organic Solution, Journal of Petroleum Science and Engineering, 32(2-4), 159-165.
Majeed, A.B., Bringedal, B. and Overa, S. (1990), Model Calculates Wax Deposition for
North Sea Oils,Oil & Gas, 88, 63.
227
Manka, J.S. and Ziegler, K.L. (2001), Factors Affecting the Performance of Crude-Oil Wax
Control Additives,, SPE Production and Operations Symposium, Society of Petroleum
Engineers, Oklahoma City, Oklahoma, USA.
Martin, R.L., Becker, H.L. and Galvan, D. (2010), Method of Reducing Paraffin Deposition
with Imidazolines, Texas.
Matzain, A., Apte, M.S., Zhang, H., Volk, M., Redus, C.L., Brill, J.P. and Creek, J.L., (2001),
Multiphase Flow Wax Deposition Modeling, Proceeding ETCE, Houston, USA.
Mehrotra, A.K. and Bhat, N.V., (2007), Modeling the Effect of Shear Stress on Deposition
from Waxy Mixtures Under Laminar Flow with Heat Transfer,Energy & Fuels, 21(3), 1277-
1286.
Misra, S.; Baruah, S.; and Singh, K., (1995), Paraffin Problems in Crude Oil Production and
Transportation: A Review, Society of Petroleum EngineersProduction and Operations,
1995(1), 50-54.
Mokhatab, S., (2009), Wax Prevention and Remedial in Subsea pipelines and flow-lines,World
Oil, No. 11, 230.
Moritis, G., (2002), Flow Assurance Challenges Production from Deeper Water,Oil and Gas
Journal, 99, 66.
Motz, K.L., Latham, R.A. and Statz, R.J., (1990), Low Pour Crude Oil Compositions, Google
Patents.
Myers, R.H. and Montrogomery, D.C. (2002), Response Surface Methodology Process and
Product Optimization Using Design Experiments,2nd Edition,Canada: John Wiley and Sons.
Nguyen, A.D., Fogler, H.S. and Sumaeth, C. (2001), Fused Chemical Reaction 2
Encapsulation: Application to Remediation of Paraffin Plugged Pipelines,Industrial
Engineering Chemistry Research,40, 50-58.
Norland, A.K., (2012), Organic Flow Assurance: Pour Point Depressant Development through
Experimental Design,M.Sc Thesis, submitted to the University of Stavanger, Norway.
Noville, I. and Noveira, L., (2012), Comparison between Real Field data and the Results of
Wax Deposition Simulation SPE 152575, Presented at the 2012 Society of Petroleum
Engineers Latin American and Caribbean Petroleum Engineering Conference held in Mexico
City, Mexico.
Osokogwu, U. and Otung, D.A., (2012), A CFD Approach for Predicting Paraffin Deposition
in Oilfield Installations,International Journal of Engineering Sciences and Research
Technology, 3, 73-95.
Oyekunle, L.O and Adeyanju, O.A. (2011),Thermodynamic Prediction of Paraffin wax
precipitation in Crude oil Pipelines,Petroleum Science and Technology, 29, 208-217.
228
Paso, K and Fogler, H.S., (2004), Bulk Stabilization in Wax Deposition Systems, Energy and
Fuels, 18, 1005-1013.
Patton, C.C. and Casad, B.M..(1970),Paraffin Deposition from Refined Wax-Solvent System,
Society of Petroleum Engineers Journal, 17.
Pedensen, K.S. and Ronningsen, H.P., (2003), Influence of Wax Inhibitors on Wax
Appearance Temperature, Pour Point, and Viscosity of Waxy Crude Oil,Energy and Fuel,
17(2), 321-328.
Pedersen, S.K., Skovborg, P., and Hans, P.R., (1991), Wax Precipitation from North-Sea Crude
Oils (4) Thermodynamic Modeling,Energy and Fuels, 5, 924.
Peng, D.Y. and Robinson, D.B. (1976), A New Two-constant Equation of State, Ind. and Eng.
Chem., 15, 1, 59-64.
Ramirez-Jaramillo, E., Lira-Galean, C. and Manero, O., (2004), Modeling Wax Deposition in
Pipelines,Petroleum Science and Technology, 22, 821 – 861.
Rana, D.P., Bateja, S., Biswas, S.K., Kumar, A., and Lai, B.,(2010), Novel Microbial Process
for Mitigating Wax Deposition in Down-Hole Tubular and Surface Flow Lines, in Society of
Petroleum Engineers, Oil and Gas India Conference and Exhibition, Society of Petroleum
Engineers, Mumbai, India.
Rosvold, K. (2008), Wax Deposition Model,M.Sc. Thesis Submitted to the Department of
Petroleum Engineering and Applied Geophysics, Norwegian University of Science and
Technology, Trondheim, 104.
Rygg, O.B., Rydahl, A.K. and Ronningsen, H.P., (1998), “Wax Deposition in Off-shore
Pipeline System” Proceeding 1st North American Conference on Multiphase Technology,
Banff. Canada.
Sabet, J.K., Ghotbi, C., and Dorkoosh, F., (2012), Application of Response Surface
methodology for Optimization of Paracetamol Particles Formation by RESS Method,Journal of
Nano-materials, Article ID 340379, 1-15.
Sadeghazad, A and Christiansen, R.L., (1998), The Effect of Cloud Point Temperature on Wax
Deposition,SPE Paper No. 49467, presented at the 8th Abu Dhabi International Petroleum
Exhibitio and Conference held in Abu Dhabi, , UAE.
Sadeghazad, A. and Ghaemi, N., (2003), Microbial Prevention of Wax Precipitation in Crude
Oil by Biodegradation Mechanism,, in SPE Asia Pacific Oil and Gas Conference and
Exhibition, Society of Petroleum Engineers, Jakarta, Indonesia.
Salam, K.K., Arinkoola, A.O., Oke, E.O., and Adeleye, J.O., (2014), Optimization of
Operating parameters using Response Surface Methodology for Paraffin-Wax Deposition in
Pipeline, Petroleum and Coal, 56 (1), 19-28.
Semenov, A., (2012), Wax Deposition Forecast,SPE 149793, Presented at the 2012 SPE North
Africa Annual Technical Conference and Exhibition held in Cairo, Egypt.
229
Singh, P., Venkatesan, R., Fogler, H.S., and Nagarajan, N.R., (2000), Formation and Aging of
Incipient Thin Film Wax-Oil Gels,AIChE Journal, 46(5).
Singh, P., Venkatesan, R., Fogler, H.S., and Nagarajan, N.R., (2001),Morphological Evaluation
of Thick Wax Deposit during Aging,AIChE Journal, 47(1).
Soave, G. (1972), Equilibrium Constant from a Modified Redlich-Kwong Equation of
State,Chemical Engineering Science, 27, 1197-1203.
Solaimany-Nazar, A., Dabir, B and Islam, M.R., (2001), Measurement and Modeling of Wax
Deposition in Crude Pipelines, SPE Paper No. 69425.
Soni, H.P., Kiranbala, S. and Bharambe, D.P., (2008), Performance-Based Designing of Wax
Crystal Growth Inhibitors,Energy and Fuel, 22(6), 3930-3938.
Straub, T.I., Autry, S.W. and King, G.E., (1989), An investigation Into Practical Removal of
Down-hole Paraffin by Thermal Method and Chemical Solvents,in SPE Production Operation
Symposium , Society of Petroleum Engineers Inc., Oklahoma City, Oklahoma.
Svendsen, J.A. (1993), Mathematical Modeling of Wax Deposition in Oil Pipeline Systems
AIChE Journal, 39(8), 1377-1388.
Thant, M.M., Taufik, M., Mohd, S., Geoffrey, H., hale, C., and Quarini, J. (2011), Mitigating
Flow Assurance Challenges in Deep-Water Fields Using Active Heating Methods,in SPE
Middle East Oil and Gas Show and Conference, Society of Petroleum Engineer,Manama,
Bahrain.
Thierheimer, C.L., (1990),Solvent for Paraffin Removal from Oil Field Equipment,Google
Patent.
Thomason, W.H., (2000), Start-up and Shut in Issues for Subsea Production of High Paraffinic
Crudes,Offshore Technology Conference, Houston, Texas.
Tinsley, J.F. (2007), Novel Laboratory Cell for Fundamental Studies of the Effect of polymer
Additives on Wax Deposition from Model Crude Oil,Energy and Fuels, 21(3), 1301-1308.
Todi, S. and Deo, M., (2006), Experimental and modeling Studies of Wax Deposition in Crude
Oil Carrying Pipelines, Paper OTC-18368, presented at Offshore Technology Conference
(OTC), in Houston, Texas, USA.
Tronov, J.P.:(1969),Effect of Flow Rate and Other factors on the Formation of Paraffin
Deposits’ Tr. Tatar Neft. Nauch-Issled, Inst. 13, 207.
Tung, N.P., Vuong, N.V., Long, B.Q., Vinh, N.Q., Hung, P.V., Hue, V.T., and Hoe, L.D.,
(2001), Studying the Mechanism of Magnetic Field Influence on Paraffin Crude Oil Viscosity
and Wax Deposition Reduction,in SPE Asia Pacific oil and Gas Conference and Exhibition,
Society of Petroleum Engineers, Jakarta, Indonesia.
230
Van Bergen, P.F., Van Dijk, M.A., and Zeeman, A.J., (2005), Method for Improving Flow
Ability of a Mixture that Contains Wax and other Hydrocarbon,Google Patents.
Venkatesan, R., (2004), The Deposition and Rheology of Organic Gels,PhD thesis, University
of Michigan.
Venkatesan, R. and Creek, J.L., (2007), Wax Production during Production operation: SOTA
Paper OTC 18798 Presented at 2007 offshore Technology Conference, Houston Texas,
USA.Society of Petroleum Engineers Inc., New Orleans, Louisiana.
Wang, W.D., Huang, Q.Y., Huang, J., Peng, Q., Fu, J., and Wang, F., (2014), Study of paraffin
Wax Deposition in Seasonally Pigged Pipeline,Chemistry and Technology of Fuel and Oils, 50,
39-50.
Wang, W.D., Huang, Q.Y., Wang, C.H., Li, S., Qu, W.X., Zhao, J.D., and He, M.Q., (2015),
Effect of Operating Conditions on Wax Deposition in the Laboratory Flow Loop Characterized
with DSC Techniques,Journal of Thermal Analysis and Calorimetry, 119, 478-485.
Wang, Q., Sarica, C. and Chen, T.X., (2001), An Experimental Study on Mechanism of Wax
Removal in Pipeline,in SPE Technical Conference and Exhibition.Society of Petroleum
Engineer Incorporated, New Orleans, Louisiana.
Wilke, C.R., and Chang, P., (1955), Correlation of Diffusion Coefficient in Dilute
Solutions,AIChE Journal, June 1955.
Won, K.W., (1986), Thermodynamic for Solid-Liquid-Vapour Equilibria: Wax Phase
Formation from Heavy Hydrocarbon Mixtures,Fluid Phase Equilibria, 30, 265-279.
Woo, G.T.: Garbis, S.J.: and Gray, T.C., (1984), Long Term Control of Paraffin Deposition”
in SPE Annual Technical Conference and Exhibition, Society of Petroleum Engineers of AIME,
Houston, Texas.
Zhang, Y., Gong, J., Wu, H., (2011), An Experimental Study on Wax Deposition of Water in
Waxy Crude Oil Emulsions,Petroleum Science and Technology, 28(16), 1653-1664.
Zhu, T., Walker, J.A., and Liang, J. (2008), Evaluation of Wax Deposition and Its Control
During Production of Alaskan North Slope Oils,Final Report Prepared for the United States
Department of Energy (National Energy Technology Laboratory) Office of Fossil Energy.
231
APPENDIX 1: CONTOUR PLOTS OF OVERALL WAX INHIBITION, W.Iav (%) FOR
CRUDE OILS B – G (AEP/XYLENE FORMULATION)
Contour Plots of AEP Vs Xylene for Oil B (2 Comp.) Contour Plots of AEP Vs Xylene for Oil E (2 Comp.)
70 70
OWI (2 OWI (2
Com, Oil Com, Oil
B) E)
< -20 < -20
65 -20 – 0 65 -20 – 0
0 – 20 0 – 20
20 – 40 20 – 40
40 – 60 > 40
Xylene (cl)
Xylene (cl)
> 60
60 60
55 55
50 50
15.0 17.5 20.0 22.5 25.0 15.0 17.5 20.0 22.5 25.0
AEP (cl) AEP (cl)
Contour Plots of AEP Vs Xylene for Oil C (2 Comp.) Contour Plots of AEP Vs Xylene for Oil F (2 Comp.)
70 70
OWI (2 OWI (2
Com, Oil Com, Oil
F)
C)
< 0
< 0
65 0 – 10
65 0 – 20
10 – 20
20 – 40
20 – 30
40 – 60
30 – 40
Xylene (cl)
> 60
Xylene (cl)
> 40
60
60
55
55
50
50 15.0 17.5 20.0 22.5 25.0
15.0 17.5 20.0 22.5 25.0 AEP (cl)
AEP (cl)
Contour Plots of AEP Vs Xylene for Oil G (2 Comp.)

Contour Plots of AEP Vs Xylene for Oil D (2 Comp.) 70
OWI (2
70 Com, Oil
OWI (2
Com, Oil G)
D) < -20
< 10 65 -20 – 0
0 – 20
65 10 – 20
20 – 30 20 – 40
> 40
Xylene (cl)
30 – 40
40 – 50
Xylene (cl)
> 50 60
60
55
55
50
50 15.0 17.5 20.0 22.5 25.0
15.0 17.5 20.0 22.5 25.0 AEP (cl)
AEP (cl)
232
APPENDIX 2: CONTOUR PLOTS OF OVERALL WAX INHIBITION, W.Iav (%) FOR
CRUDE OILS B – G (AEP/XYLENE/BUTYL-ACRYLATE FORMULATION)
Contour plots for Oil B (AEP Vs Xylene) Contour plots for Oil E (AEP Vs Xylene)
80 80
OWI (Oil B) OWI (Oil
< 40 E)
40 – 60 < 20
75 75 20 – 40
60 – 80
80 – 100 40 – 60
100 – 120 60 – 80
70 > 120 70 > 80
Hold Values
Hold Values
Xylene
Xylene
Butyl acrylate 3
Butyl acrylate 3 65
65
60
60
55
55
50
50 15.0 17.5 20.0 22.5 25.0
15.0 17.5 20.0 22.5 25.0 AEP
AEP
Contour plots for Oil F (AEP Vs Xylene)

Contour plots for Oil C (AEP Vs Xylene)
80
OWI (OIl F)
80
OWI (OIl < 40
C) 40 – 60
< 40
75 60 – 80
75 40 – 60 80 – 100
60 – 80 100 – 120
80 – 100 70 > 120
70 100 – 120
Hold Values
Xylene
> 120
Butyl acrylate 3
Xylene
Hold Values
65
65 Butyl acrylate 3
60
60
55
55
50
50 15.0 17.5 20.0 22.5 25.0
15.0 17.5 20.0 22.5 25.0 AEP
AEP
Contour plots for Oil G (AEP Vs Xylene)

Contour plots for Oil D (AEP Vs Xylene) 80
OWI (OIL
80 G)
OWI (Oil D)
< 20 < 40
75 40 – 60
20 – 40
75 40 – 60 60 – 80
60 – 80 80 – 100
80 – 100 70 100 – 120
70 100 – 120 > 120
Xylene
> 120
Hold Values
65
Xylene
Hold Values Butyl acrylate 3

65 Butyl acrylate 3
60
60
55
55
50
50 15.0 17.5 20.0 22.5 25.0
15.0 17.5 20.0 22.5 25.0 AEP
AEP
233
Contour plot for Oil B (AEP Vs Butyl acrylate) Contour plot for Oil E (AEP Vs Butyl acrylate)
5 5
OWI (Oil OWI (Oil
B) E)
< 40 < 20
40 – 60 20 – 30
4 60 – 80 30 – 40
80 – 100
4
40 – 50
Butyl acrylate (cl)
Butyl acrylate (cl)

> 100 50 – 60
60 – 70
Hold Values
> 70
Xylene 60
3 3 Hold Values
Xylene 60
2 2
1
15.0 17.5 20.0 22.5 25.0 1
15.0 17.5 20.0 22.5 25.0
AEP (cl)
AEP (cl)
Contour plot for Oil C (AEP Vs Butyl acrylate) Contour plot for Oil F (AEP Vs Butyl acrylate)
5 5
OWI (OIl
OWI (OIl
C)
F)
< 40 < 40
40 – 60 40 – 60
4 60 – 80 60 – 80
80 – 100
4
80 – 100
Butyl acrylate (cl)
Butyl acrylate (cl)
> 100 > 100
Hold Values Hold Values

Xylene 60 Xylene 60
3 3
2 2
1 1
15.0 17.5 20.0 22.5 25.0
15.0 17.5 20.0 22.5 25.0
AEP (cl)
AEP (cl)
Contour plot for Oil G (AEP Vs Butyl acrylate)

Contour plot for Oil D (AEP Vs Butyl acrylate)
5
5 OWI (OIL
OWI (Oil
G)
D)
< 30
< 20 30 – 45
20 – 40 45 – 60
40 – 60
4
4 60 – 75
Butyl acrylate (cl)
60 – 80 75 – 90
Butyl acrylate (cl)
80 – 100 90 – 105
> 100 > 105
Hold Values 3 Hold Values
3 Xylene 60 Xylene 60
2
2
1
1 15.0 17.5 20.0 22.5 25.0
15.0 17.5 20.0 22.5 25.0 AEP (cl)
AEP (cl)
234
Contour plot for Oil B(Xylene Vs Butyl acrylate) Contour plot for Oil E (Xylene Vs Butyl acrylate)
5 5
OWI (Oil OWI (Oil
B) E)
< 40 < 20
40 – 60 20 – 40
4 60 – 80 4 40 – 60
80 – 100 60 – 80
Butyl acrylate (cl)
Butyl acrylate (cl)

> 100 80 – 100
> 100
Hold Values
AEP 21 Hold Values
3 3 AEP 21
2 2
1 1
50 55 60 65 70 75 80 50 55 60 65 70 75 80
Xylene (cl) Xylene (cl)
Contour plot for Oil C (Xylene Vs Butyl acrylate) Contour plot for Oil F (Xylene Vs Butyl acrylate)
5 5
OWI (OIl OWI (OIl F)
C) < 40
< 40 40 – 60
40 – 60 60 – 80
4 60 – 80 80 – 100
4
80 – 100 100 – 120
Butyl acrylate (cl)
Butyl acrylate (cl)
100 – 120 > 120

> 120
Hold Values
Hold Values AEP 21
3 AEP 21 3
2 2
1 1
50 55 60 65 70 75 80 50 55 60 65 70 75 80
Xylene (cl) Xylene (cl)
Contour plot for Oil D (Xylene Vs Butyl acrylate) Contour plot for Oil G (Xylene Vs Butyl acrylate)
5 5
OWI (Oil OWI (OIL
D) G)
< 20 < 40
20 – 40 40 – 60
4 40 – 60 60 – 80
60 – 80 4
80 – 100
Butyl acrylate (cl)
Butyl acrylate (cl)
80 – 100 100 – 120

> 100 > 120
Hold Values
Hold Values
3 AEP 21 3 AEP 21
2 2
1
50 55 60 65 70 75 80 1
50 55 60 65 70 75 80
Xylene (cl)
Xylene (cl)
235
APPENDIX 3: RESIDUAL PLOTS OF OVERALL WAX INHIBITION, W.Iav (%) FOR
CRUDE-OIL B – G (AEP/XYLENE FORMULATION)
R e s id u a l P lo ts o f W Ia v ( % ) fo r O il B A E P / X y le n e F o r m u la tio n
99
2
90
1
Residual
Percent
50 0
10 -1
1 -2
-2 -1 0 1 2 0 15 30 45 60
R e s id u a l F it te d V a lu e
H is t o g r a m V e rs u s O rd e r
3
2
Frequency
Residual
2
0
1
-1
0 -2
-2 -1 0 1 2 1 2 3 4 5 6 7 8 9 10 11 12 13 14
R e s i d u a l P l o t s o f W I a v ( % ) f o r O i l C A E P / X y l e n e F o r m u l a t i o n
9 9
2
9 0 1
Residual
Percent
5 0 0
- 1
1 0
- 2
1
- 2 - 1 0 1 2 0 1 5 3 0 4 5 6 0
R e s id u a l F it t e d V a lu e
H is t o g r a m V e r s u s O r d e r
4 2
3 1
Frequency
Residual
2 0
1 - 1
- 2
0
- 2 - 1 0 1 2 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4
R e s id u a l O b s e r v a t io n O r d e r
R e s id u a l P lo ts o f W Ia v ( % ) fo r O il D A E P / X y le n e F o r m u la tio n
99
2
90
1
Residual
Percent
50
0
10
-1
1
-2 -1 0 1 2 10 20 30 40 50
4 2
3 1
Frequency
Residual
2
0
1
-1
0
- 1 .5 - 1 .0 - 0 .5 0 .0 0 .5 1 .0 1 .5 2 .0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
236
R e s id u a l P lo ts o f W Ia v ( % ) fo r O il E A E P / X y le n e F o r m u la tio n
99 2
90
1
Residual
Percent
50
0
10
-1
1
-2 -1 0 1 2 0 10 20 30 40
4 2
3 1
Frequency
Residual
2 0
1
-1
0
- 1 .5 - 1 .0 - 0 .5 0 .0 0 .5 1 .0 1 .5 2 .0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
R e s id u a l P lo ts o f W Ia v ( % ) fo r O il F A E P / X y le n e F o r m u la tio n
99 2
90
1
Residual
Percent
50
0
10
-1
1
-2 -1 0 1 2 0 10 20 30 40
8 2
6 1
Frequency
Residual
4
0
2
-1
0
- 1 .0 - 0 .5 0.0 0 .5 1 .0 1 .5 2 .0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
R e s id u a l P lo ts o f W Ia v ( % ) fo r O il G A E P / X y le n e F o r m u la tio n
99 2
90
1
Residual
Percent
50
0
10
-1
1
-2 -1 0 1 2 10 20 30 40 50
4 2
3 1
Frequency
Residual
2
0
1
-1
0
- 1 .5 - 1 .0 - 0 .5 0 .0 0 .5 1 .0 1 .5 1 2 3 4 5 6 7 8 9 10 11 12 13 14
237
APPENDIX 4: RESIDUAL PLOTS OF OVERALL WAX INHIBITION W.Iav FOR CRUDE-
OIL B – G (AEP/XYLENE/BUTYL-ACRYLATE FORMULATION)
R e s id u a l P lo ts fo r O W I ( % ) O il B
N o r m a l P r o b a b ilit y P lo t V e rs u s F it s
9 9 1 0
9 0
Residual
5
Percent
5 0
0
1 0
1 - 5
- 10 -5 0 5 10 20 30 40 50 6 0
1 2 1 0
9
Frequency
Residual
6
0
3
0 - 5
- 2 .5 0 .0 2 .5 5 .0 7 .5 1 0 .0 2 4 6 8 1 0 1 2 1 4 1 6 1 8 20
R e s id u a l P lo ts fo r O W I ( % ) , O il C
99
10
90
Residual
5
Percent
50
0
10
1 -5
-10 -5 0 5 10 20 30 40 50 60
10
1 0 .0
Frequency
Residual
7 .5 5
5 .0
0
2 .5
0 .0 -5
- 2 .5 0 .0 2 .5 5 .0 7 .5 1 0 .0 2 4 6 8 10 12 14 16 18 20
R e s id u a l P lo ts fo r O W I ( % ) , O il D
99
8
90
Residual
Percent
4
50
0
10
1
-5 0 5 10 10 20 30 40 50
1 0 .0 8
7 .5
Frequency
Residual
4
5 .0
0
2 .5
0 .0
-4 -2 0 2 4 6 8 2 4 6 8 10 12 14 16 18 20
238
R e s id u a l P lo ts fo r O W I ( % ) , O il E
99 10
90
5
Residual
Percent
50
0
10
1 -5
-10 -5 0 5 10 10 20 30 40 50
12 10
9
Frequency
Residual
6
0
3
0 -5
- 2 .5 0 .0 2 .5 5.0 7.5 1 0 .0 2 4 6 8 10 12 14 16 18 20
R e s id u a l P lo ts fo r O W I ( % ) , O il F
N o r m a l P r o b a b il it y P lo t V e r s u s F it s
99 10
90
5
Residual
Percent
50
0
10
-5
1
-10 -5 0 5 10 20 30 40 50 60
10
1 0 .0
Frequency
5
Residual
7 .5
5 .0
0
2 .5
-5
0 .0
- 5 .0 - 2 .5 0 .0 2.5 5 .0 7.5 1 0 .0 2 4 6 8 10 12 14 16 18 20
R e s id u a l P lo ts fo r O W I ( % ) , O il G
99
5.0
90
Residual
Percent
2.5
50
0.0
10 - 2.5
1 - 5.0
- 5 .0 - 2.5 0.0 2.5 5 .0 30 40 50 60 70
R e s id u a l F itte d V a lu e
8 5.0
Frequency
Residual
6 2.5
4 0.0
2 - 2.5
0 - 5.0
-4 -2 0 2 4 6 2 4 6 8 10 12 14 16 18 20
R e s id u a l O b se r v a tio n O r d e r
239
APPENDIX 5: CONTOUR PLOTS OF WAX DEPOSITION FOR CRUDE OIL B –G (4-
VARIABLES)
Contour Plot of Wax Deposit for Oil B(Coolant Temp. Vs Wax Inhib. Percent) Contour Plot of Wax Deposit for Oil E(Coolant Temp. Vs Wax Inhib. Percent)
5 5
W.D. (Oil
W.D (Oil B)
E)
< 100
< 60
100 – 150
60 – 90
150 – 200

4 90 – 120
4 200 – 250 120 – 150

250 – 300 150 – 180
> 300 180 – 210
> 210
Hold Values
Oil Temperature 325 3 Hold Values
3 Flow rate 0.5 Oil Temperature 325
Flow rate 0.5
2
2
1
280 284 288 292 296
1 Coolant Temperature (K)
280 284 288 292 296
Coolant Temperature (K)
Contour Plot of Wax Deposit for Oil F(Coolant Temp. Vs Wax Inhib. Percent)
Contour Plot of Wax Deposit for Oil C(Coolant Temp. Vs Wax Inhib. Percent) 5
W.D. (Oil
5 F)
W.D. (Oil < 100
C) 100 – 125
< 100 125 – 150

4
100 – 150 150 – 175
4 150 – 200 175 – 200

200 – 250 200 – 225
250 – 300 225 – 250
> 300 > 250
3
Hold Values Hold Values
3 Oil Temperature 325 Oil Temperature 325
Flow rate 0.5 Flow rate 0.5
2
2
1
280 284 288 292 296
1 Coolant Temperature (K)
280 284 288 292 296
Contour Plot of Wax Deposit for Oil G(Coolant Temp. Vs Wax Inhib. Percent)
Contour Plot of Wax Deposit for Oil D(Coolant Temp. Vs Wax Inhib. Percent) 5
W.D. (Oil
5 G)
W.D. (Oil
< 100
D)
100 – 150
< 100
100 – 150 4 150 – 200

200 – 250
4 150 – 200
200 – 250 > 250
250 – 300
Hold Values
> 300
Oil Temperature 325
3 Flow rate 0.5
Hold Values
3 Oil Temperature 325
Flow rate 0.5
2
2
1
1 280 284 288 292 296
280 284 288 292 296 Coolant Temperature (K)
240
Contour Plot of Wax Deposit for Oil B(Coolant Temp. Vs FlowRate) Contour Plot of Wax Deposit for Oil E(Coolant Temp. Vs FlowRate)
0.7 0.7
W.D (Oil B) W.D. (Oil
< 100 E)
100 – 120 < 80
120 – 140 80 – 100
0.6 140 – 160 0.6 100 – 120

160 – 180 120 – 140
180 – 200 > 140
200 – 220
Hold Values
> 220
Oil Temperature 325
0.5 Hold Values 0.5 Wax Inhibitor Perce. 3
Oil Temperature 325
Wax Inhibitor Perce. 3
0.4 0.4
0.3 0.3
280 284 288 292 296 280 284 288 292 296
Coolant Temperature (K) Coolant Temperature (K)
Contour Plot of Wax Deposit for Oil C(Coolant Temp. Vs FlowRate) Contour Plot of Wax Deposit for Oil F(Coolant Temp. Vs FlowRate)
0.7 0.7
W.D. (Oil W.D. (Oil
C) F)
< 100 < 100
100 – 120 100 – 120
0.6 120 – 140 0.6 120 – 140
140 – 160 140 – 160

160 – 180 160 – 180

180 – 200 > 180
200 – 220
> 220 Hold Values
0.5 0.5 Oil Temperature 325
Hold Values Wax Inhibitor Perce. 3
Oil Temperature 325
0.4 0.4
0.3 0.3
280 284 288 292 296 280 284 288 292 296
Coolant Temperature (K) Coolant Temperature (K)
Contour Plot of Wax Deposit for Oil D(Coolant Temp. Vs FlowRate) Contour Plot of WaxDeposit for Oil G(Coolant Temp. Vs FlowRate)
0.7 0.7
W.D. (Oil W.D. (Oil
D) G)
< 100 < 100
100 – 120 100 – 120
0.6 120 – 140 120 – 140
0.6
140 – 160
140 – 160
160 – 180 160 – 180
180 – 200
180 – 200
200 – 220
> 200
> 220
0.5 0.5 Hold Values
Hold Values
Oil Temperature 325
Oil Temperature 325
0.4
0.4
0.3
280 284 288 292 296 0.3
Coolant Temperature (K) 280 284 288 292 296
241
Contour Plot of Wax Deposit for Oil B(Oil Temp. Vs Wax Inhib. Perc.) Contour Plot of Wax Deposit for Oil E(Oil Temp. Vs Wax Inhib. Perc.)
5 5
< 100 E)
100 – 150 < 90
150 – 200 90 – 120

4 200 – 250 120 – 150
250 – 300
4
150 – 180
> 300 180 – 210
> 210
Hold Values
Coolant Temperature 287 Hold Values
3 Flow rate 0.5 3 Coolant Temperature 287
Flow rate 0.5
2 2
1 1
321 322 323 324 325 326 327 328 329 321 322 323 324 325 326 327 328 329
Oil Temperature (K) Oil Temperature (K)
Contour Plot of Wax Deposit for Oil C(Oil Temp. Vs Wax Inhib. Perc.)
5 Contour Plot of Wax Deposit for Oil F(Oil Temp. Vs Wax Inhib. Perc.)
W.D. (Oil
C)
5
< 150 W.D. (Oil
150 – 200 F)
4 200 – 250 < 100

250 – 300 100 – 125
> 300 125 – 150

4
150 – 175
Hold Values 175 – 200
Coolant Temperature 287 200 – 225
3 Flow rate 0.5 225 – 250
> 250
3
Hold Values
Coolant Temperature 287
2 Flow rate 0.5
1
321 322 323 324 325 326 327 328 329
Oil Temperature (K) 1
321 322 323 324 325 326 327 328 329
Oil Temperature (K)
Contour Plot of Wax Deposit for Oil D(Oil Temp. Vs Wax Inhib. Perc.) Contour Plot of WaxDeposit for Oil G(Oil Temp. Vs WaxInhib. Perc.)
5 5
W.D. (Oil W.D. (Oil
D) G)
< 100 < 100
100 – 150 100 – 150
150 – 200
4 4 150 – 200
200 – 250
200 – 250
250 – 300
> 250
> 300
Hold Values
Hold Values
3 Coolant Temperature 287
3 Flow rate 0.5
Flow rate 0.5
2
2
1
321 322 323 324 325 326 327 328 329 1
Oil Temperature (K) 321 322 323 324 325 326 327 328 329
Oil Temperature (K)
242
Contour Plot of Wax Deposit for Oil B(Oil Temp. Vs Flow rate) Contour Plot of Wax Deposit for Oil E(Oil Temp. Vs Flow rate)
0.7 0.7
< 100 E)
100 – 120 < 80
120 – 140 80 – 100
0.6 140 – 160 0.6 100 – 120

160 – 180 120 – 140
180 – 200 140 – 160
> 200 160 – 180
> 180
Hold Values
0.5 Coolant Temperature 287 0.5 Hold Values
Wax Inhibitor Perce. 3 Coolant Temperature 287
0.4 0.4
0.3 0.3
321 322 323 324 325 326 327 328 329 321 322 323 324 325 326 327 328 329
Contour Plot of Wax Deposit for Oil C(Oil Temp. Vs Flow rate) Contour Plot of Wax Deposit for Oil F(Oil Temp. Vs Flow rate)
0.7 0.7
W.D. (Oil W.D. (Oil
C) F)
< 120 < 100
120 – 140 100 – 120
140 – 160 0.6 120 – 140
0.6 140 – 160
Flow rate (Lit./min)
160 – 180
180 – 200 160 – 180
200 – 220 180 – 200
> 220 > 200
0.5 0.5 Hold Values

Hold Values
Coolant Temperature 287 Coolant Temperature 287
Wax Inhibitor Perce. 3 Wax Inhibitor Perce. 3
0.4 0.4
0.3
0.3
321 322 323 324 325 326 327 328 329
321 322 323 324 325 326 327 328 329
Oil Temperature (K)
Oil Temperature (K)
Contour Plot of WaxDeposit for Oil D(Oil Temp. Vs Flowrate) Contour Plot of WaxDeposit for Oil G(Oil Temp. Vs Flowrate)
0.7 0.7
W.D. (Oil W.D. (Oil
D) G)
< 90 < 100
90 – 120 100 – 120
120 – 150 0.6 120 – 140
0.6 140 – 160
150 – 180
180 – 210 160 – 180

> 210 180 – 200
> 200
Hold Values
0.5 Coolant Temperature 287
0.5 Hold Values
0.4 0.4
0.3 0.3
321 322 323 324 325 326 327 328 329 321 322 323 324 325 326 327 328 329
243
Contour Plot of Wax Deposit for Oil B(Oil Temp. Vs Flow rate) Contour Plot of Wax Deposit for Oil E(Oil Temp. Vs Flowrate)
0.7 0.7
< 100 E)
100 – 120 < 80
120 – 140 80 – 100
0.6 140 – 160 100 – 120
0.6
160 – 180

120 – 140
180 – 200 140 – 160

> 200 160 – 180
> 180
Hold Values
0.5 Coolant Temperature 287 0.5 Hold Values
0.4 0.4
0.3 0.3
321 322 323 324 325 326 327 328 329 321 322 323 324 325 326 327 328 329
Contour Plot of WaxDeposit for Oil C(Oil Temp. Vs Flowrate) Contour Plot of WaxDeposit for Oil F(Oil Temp. VsFlowrate)
0.7 0.7
W.D. (Oil W.D. (Oil
C) F)
< 120 < 100
120 – 140 100 – 120
0.6 140 – 160 120 – 140
0.6
160 – 180
140 – 160
180 – 200 160 – 180
200 – 220 180 – 200
> 220 > 200
0.5 Hold Values 0.5 Hold Values
Coolant Temperature 287 Coolant Temperature 287
0.4 0.4
0.3 0.3
321 322 323 324 325 326 327 328 329 321 322 323 324 325 326 327 328 329
Contour Plot of Wax Deposit for Oil G(Oil Temp. Vs Flowrate)

Contour Plot of Wax Deposit for Oil D(Oil Temp. Vs Flowrate) 0.7
W.D. (Oil
0.7 G)
W.D. (Oil
< 100
D)
100 – 120
< 90
90 – 120 0.6 120 – 140
120 – 150
140 – 160
0.6 160 – 180
150 – 180
180 – 210 180 – 200
> 210 > 200
Hold Values 0.5 Hold Values

0.5 Coolant Temperature 287 Coolant Temperature 287
0.4
0.4
0.3 0.3
321 322 323 324 325 326 327 328 329 321 322 323 324 325 326 327 328 329
244
Contour Plot of Wax Deposit for Oil B(Wax Inhib. Perc. Vs Flow rate) Contour Plot of Wax Deposit for Oil E(Wax Inhib. Perc. Vs Flow rate)
0.7 0.7
< 100 E)
100 – 125 < 80
125 – 150 80 – 100
150 – 175 0.6 100 – 120
0.6 120 – 140

175 – 200
200 – 225 140 – 160
> 225 160 – 180
> 180
Hold Values
0.5 0.5 Hold Values
Oil Temperature 325 Coolant Temperature 287
Oil Temperature 325
0.4 0.4
0.3
0.3 1 2 3 4 5
1 2 3 4 5
Contour Plot of Wax Deposit for Oil C(Wax Inhib. Perc. Vs Flowrate)
0.7
Contour Plot of Wax Deposit for Oil F(Wax Inhib. Perc. Vs Flowrate)
W.D. (Oil
0.7
C) W.D. (Oil
< 100 F)
100 – 120 < 100
0.6 120 – 140 100 – 120
140 – 160 120 – 140

0.6
160 – 180 140 – 160

180 – 200 160 – 180
200 – 220 > 180
> 220
0.5 Hold Values
Hold Values
0.5 Coolant Temperature 287
Coolant Temperature 287 Oil Temperature 325
Oil Temperature 325
0.4
0.4
0.3 0.3
1 2 3 4 5 1 2 3 4 5
Wax Inhibitor Perce.(%) Wax Inhibitor Perce. (%)
Contour Plot of Wax Deposit for Oil D(Wax Inhib. Perc. Vs Flowrate) Contour Plot of Wax Deposit for Oil G(Wax Inhib. Perc. Vs Flowrate)
0.7 0.7
W.D. (Oil W.D. (Oil
D) G)
< 80 < 100
80 – 120 100 – 125
0.6 120 – 160 0.6 125 – 150
160 – 200 150 – 175

200 – 240 175 – 200
> 240 200 – 225
> 225
Hold Values
0.5 0.5 Hold Values
Oil Temperature 325 Coolant Temperature 287
Oil Temperature 325
0.4 0.4
0.3
0.3 1 2 3 4 5
1 2 3 4 5
WaxInhibitor Perce. (%)
245
APPENDIX 6 : RESIDUAL PLOTS OF WAX DEPOSITION FOR CRUDE OIL B – G (4
VARIABLES, LAMINAR FLOW)
R e s id u a l P lo ts fo r W .D . ( c m 3 ), O IL B
99
4
90
Residual
Percent
2
50
0
10 -2
1 -4
- 5 .0 - 2 .5 0 .0 2 .5 5 .0 50 100 150 200 250
6 .0
4
4 .5
Frequency
Residual
2
3 .0
0
1 .5 -2
0 .0 -4
-4 -2 0 2 4 2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0 2 2 2 4 2 6 2 8 3 0
R e s id u a l P lo ts fo r W .D . ( c m 3 ), O IL C
99
5 .0
90
2 .5
Residual
Percent
50 0 .0
- 2 .5
10
- 5 .0
1
-8 -4 0 4 8 100 150 200 250
4
5 .0
3
Frequency
2 .5
Residual
2 0 .0
1 - 2 .5
- 5 .0
0
-4 -2 0 2 4 6 2 4 6 8 10 12 1 4 1 6 18 20 2 2 2 4 26 28 3 0
R e s id u a l P lo ts fo r W .D . ( c m 3 ) , O IL D
99 10
90 5
Residual
Percent
50 0
10 -5
1 - 10
- 10 -5 0 5 10 50 100 150 200 250
8 10
6 5
Frequency
Residual
4 0
2 -5
0 - 10
-8 -4 0 4 8 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
246
R e s id u a l P lo ts fo r W .D . ( c m 3 ), O IL E
N o r m a l P r o b a b ili t y P lo t V e r s u s F it s
99 4
90 2
Residual
Percent
0
50
-2
10
-4
1
- 5 .0 - 2 .5 0 .0 2 .5 5.0 60 90 120 150 180
4
4 .8
2
Frequency
3 .6
Residual
0
2 .4
-2
1 .2
-4
0 .0
-4 -2 0 2 4 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
R e s id u a l P lo ts fo r W .D . ( c m 3 ), O IL F
99
5 .0
90
2 .5
Residual
Percent
50 0 .0
- 2 .5
10
- 5 .0
1
-8 -4 0 4 8 100 125 150 175 200
1 0 .0
5 .0
7 .5
Frequency
2 .5
Residual
5 .0 0 .0
- 2 .5
2 .5
- 5 .0
0 .0
-6 -4 -2 0 2 4 6 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
R e s id u a l P lo ts fo r W .D . ( c m 3 ) , O IL G .
99
5
90
Residual
Percent
0
50
-5
10
1 - 10
-10 -5 0 5 10 50 100 150 200 250
8
5
6
Frequency
Residual
0
4
-5
2
0 - 10
-8 -4 0 4 8 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
247
APPENDIX 7: RESIDUAL PLOTS OF WAX DEPOSITION FOR CRUDE OIL B – G (4
VARIABLES, TURBULENT FLOW)
R e s id u a l P lo ts fo r O il B ( 4 C o m p o n e n ts , T u r b .)
99
4
90
Residual
Percent
2
50
0
10 -2
1 -4
- 5 .0 - 2 .5 0 .0 2 .5 5 .0 50 100 150 200 250
6 .0
4
4 .5
Frequency
Residual
2
3 .0
0
1 .5 -2
0 .0 -4
-4 -2 0 2 4 2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0 2 2 2 4 2 6 2 8 3 0
R e s id u a l P lo ts fo r O il C ( 4 C o m p o n e n ts , T u r b .)
9 9
5 .0
9 0
2 .5
Residual
Percent
5 0 0 .0
- 2 .5
1 0
- 5 .0
1
- 8 - 4 0 4 8 1 0 0 1 5 0 2 0 0 2 5 0
4
5 .0
3
Frequency
2 .5
Residual
2 0 .0
1 - 2 .5
- 5 .0
0
- 4 - 2 0 2 4 6 2 4 6 8 1 0 1 2 1 4 1 6 1 8 2 0 2 2 2 4 2 6 2 8 3 0
R e s id u a l P lo ts fo r O il D ( 4 C o m p o n e n ts , T u r b .)
99 10
90 5
Residual
Percent
50 0
10 -5
1 -10
-10 -5 0 5 10 50 100 150 200 250
8 10
6 5
Frequency
Residual
4 0
2 -5
0 -10
-8 -4 0 4 8 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
248
R e s id u a l P lo ts fo r O il E ( 4 C o m p o n e n ts , T u r b .)
99 4
90 2
Residual
Percent
0
50
-2
10
-4
1
- 5 .0 - 2 .5 0 .0 2 .5 5.0 60 90 120 150 180
4
4 .8
2
Frequency
3 .6
Residual
0
2 .4
-2
1 .2
-4
0 .0
-4 -2 0 2 4 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
R e s id u a l P lo ts fo r O il F ( 4 C o m p o n e n ts , T u r b .)
99
5 .0
90
2 .5
Residual
Percent
50 0 .0
- 2 .5
10
- 5 .0
1
-8 -4 0 4 8 100 125 150 175 200
1 0 .0
5 .0
7 .5
Frequency
2 .5
Residual
5 .0 0 .0
- 2 .5
2 .5
- 5 .0
0 .0
-6 -4 -2 0 2 4 6 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
R e s id u a l P lo ts fo r O il G , ( 4 C o m p o n e n ts , T u r b .)
N o r m a l P r o b a b ili t y P lo t V e r s u s F it s
99
5
90
Residual
Percent
0
50
-5
10
1 -10
-10 -5 0 5 10 50 100 150 200 250
8
5
6
Frequency
Residual
0
4
-5
2
0 -10
-8 -4 0 4 8 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30
249
APPENDIX 8: ANALYSIS OF VARIANCE (ANOVA) RESULTS OF OVERALL WAX
INHIBITION FOR CRUDE OIL A-G AEP/XYLENE FORMULATION
————— 5/2/2014 10:34:27 PM ————————————————————
Response Surface Regression: OWI (%), Oil, OWI (%), Oil, ... versus Block, AEP
Response Surface Regression: OWI (%), Oil A versus Block, AEP (cl), Xylene (cl)
The analysis was done using coded units.
Estimated Regression Coefficients for OWI (%), Oil A
Term Coef SE Coef T P

Constant 33.9000 0.5290 64.081 0.000
Block -0.3000 0.3797 -0.790 0.455
AEP (cl) 6.5000 0.7705 8.436 0.000
Xylene (cl) 6.8333 0.7705 8.869 0.000
AEP (cl)*AEP (cl) 9.4000 1.1298 8.320 0.000
Xylene (cl)*Xylene (cl) -17.6000 1.1298 -15.579 0.000
AEP (cl)*Xylene (cl) -31.0000 2.6690 -11.615 0.000
S = 1.33452 PRESS = 98.8862

R-Sq = 99.03% R-Sq(pred) = 92.31% R-Sq(adj) = 98.20%
Analysis of Variance for OWI (%), Oil A
Source DF Seq SS Adj SS Adj MS F P

Blocks 1 1.14 1.11 1.112 0.62 0.455
Regression 5 1272.10 1272.10 254.421 142.86 0.000
Linear 2 266.83 266.83 133.417 74.91 0.000
AEP (cl) 1 126.75 126.75 126.750 71.17 0.000
Xylene (cl) 1 140.08 140.08 140.083 78.66 0.000
Square 2 765.02 765.02 382.511 214.78 0.000
AEP (cl)*AEP (cl) 1 332.80 123.29 123.293 69.23 0.000
Xylene (cl)*Xylene (cl) 1 432.22 432.22 432.223 242.69 0.000
Interaction 1 240.25 240.25 240.250 134.90 0.000
AEP (cl)*Xylene (cl) 1 240.25 240.25 240.250 134.90 0.000
Residual Error 7 12.47 12.47 1.781
Lack-of-Fit 3 7.13 7.13 2.378 1.78 0.289
Pure Error 4 5.33 5.33 1.333
Total 13 1285.71
Estimated Regression Coefficients for OWI (%), Oil A using data in uncoded
units
Term Coef
Constant -1199.30
Block -0.300000
AEP (cl) 21.1167
Xylene (cl) 32.6533
AEP (cl)*AEP (cl) 0.261111
Xylene (cl)*Xylene (cl) -0.176000
AEP (cl)*Xylene (cl) -0.516667
250
Response Surface Regression: OWI (%), Oil B versus Block, AEP (cl), Xylene (cl)
Estimated Regression Coefficients for OWI (%), Oil B

Constant 34.4571 0.5544 62.149 0.000
Block 0.7048 0.3979 1.771 0.120
AEP (cl) 9.0000 0.8075 11.146 0.000
Xylene (cl) 3.3333 0.8075 4.128 0.004
AEP (cl)*AEP (cl) 16.6833 1.1840 14.090 0.000
AEP (cl)*Xylene (cl) -38.0000 2.7972 -13.585 0.000
S = 1.39861 PRESS = 51.0827

Analysis of Variance for OWI (%), Oil B

Blocks 1 28.57 6.14 6.14 3.14 0.120
Regression 5 2584.59 2584.59 516.92 264.26 0.000
Linear 2 276.33 276.33 138.17 70.63 0.000
AEP (cl) 1 243.00 243.00 243.00 124.23 0.000
Xylene (cl) 1 33.33 33.33 33.33 17.04 0.004
Square 2 1947.26 1947.26 973.63 497.73 0.000
AEP (cl)*AEP (cl) 1 943.82 388.37 388.37 198.54 0.000
Interaction 1 361.00 361.00 361.00 184.55 0.000
AEP (cl)*Xylene (cl) 1 361.00 361.00 361.00 184.55 0.000
Residual Error 7 13.69 13.69 1.96
Lack-of-Fit 3 2.36 2.36 0.79 0.28 0.840
Pure Error 4 11.33 11.33 2.83
Total 13 2626.86
Estimated Regression Coefficients for OWI (%), Oil B using data in uncoded
units
Term Coef
Constant -1576.07
Block 0.704762
AEP (cl) 20.0361
Xylene (cl) 45.8133
AEP (cl)*AEP (cl) 0.463426
Response Surface Regression: OWI (%), Oil C versus Block, AEP (cl), Xylene (cl)
Estimated Regression Coefficients for OWI (%), Oil C
251
Constant 29.0857 0.6028 48.248 0.000
Block -0.0762 0.4327 -0.176 0.865
AEP (cl) 6.8333 0.8780 7.783 0.000
Xylene (cl) 7.8333 0.8780 8.922 0.000
AEP (cl)*AEP (cl) 21.7167 1.2874 16.869 0.000
AEP (cl)*Xylene (cl) -31.0000 3.0414 -10.193 0.000
S = 1.52072 PRESS = 84.4708

Analysis of Variance for OWI (%), Oil C

Blocks 1 0.64 0.07 0.072 0.03 0.865
Regression 5 2288.67 2288.67 457.734 197.93 0.000
Linear 2 324.17 324.17 162.083 70.09 0.000
AEP (cl) 1 140.08 140.08 140.083 60.57 0.000
Xylene (cl) 1 184.08 184.08 184.083 79.60 0.000
Square 2 1724.25 1724.25 862.126 372.80 0.000
AEP (cl)*AEP (cl) 1 1178.14 658.07 658.066 284.56 0.000
Interaction 1 240.25 240.25 240.250 103.89 0.000
AEP (cl)*Xylene (cl) 1 240.25 240.25 240.250 103.89 0.000
Residual Error 7 16.19 16.19 2.313
Lack-of-Fit 3 5.52 5.52 1.840 0.69 0.604
Pure Error 4 10.67 10.67 2.667
Total 13 2305.50
Estimated Regression Coefficients for OWI (%), Oil C using data in uncoded
units
Term Coef
Constant -1139.00
Block -0.0761905
AEP (cl) 6.80278
Xylene (cl) 35.3733
AEP (cl)*AEP (cl) 0.603241
Response Surface Regression: OWI (%), Oil D versus Block, AEP (cl), Xylene (cl)
Estimated Regression Coefficients for OWI (%), Oil D

Constant 31.6714 0.5209 60.804 0.000
Block -0.6524 0.3739 -1.745 0.124
AEP (cl) 4.1667 0.7586 5.492 0.001
Xylene (cl) 6.5000 0.7586 8.568 0.000
AEP (cl)*AEP (cl) 16.9333 1.1124 15.223 0.000
AEP (cl)*Xylene (cl) -13.0000 2.6280 -4.947 0.002
S = 1.31398 PRESS = 60.7823
252
Analysis of Variance for OWI (%), Oil D

Blocks 1 4.57 5.26 5.257 3.05 0.124
Regression 5 1446.77 1446.77 289.354 167.59 0.000
Linear 2 178.83 178.83 89.417 51.79 0.000
AEP (cl) 1 52.08 52.08 52.083 30.17 0.001
Xylene (cl) 1 126.75 126.75 126.750 73.41 0.000
Square 2 1225.69 1225.69 612.844 354.96 0.000
AEP (cl)*AEP (cl) 1 770.24 400.10 400.099 231.74 0.000
Interaction 1 42.25 42.25 42.250 24.47 0.002
AEP (cl)*Xylene (cl) 1 42.25 42.25 42.250 24.47 0.002
Residual Error 7 12.09 12.09 1.727
Lack-of-Fit 3 5.42 5.42 1.806 1.08 0.451
Pure Error 4 6.67 6.67 1.667
Total 13 1463.43
Estimated Regression Coefficients for OWI (%), Oil D using data in uncoded
units
Term Coef
Constant -737.879
Block -0.652381
AEP (cl) -6.06111
Xylene (cl) 26.8800
AEP (cl)*AEP (cl) 0.470370
Response Surface Regression: OWI (%), Oil E versus Block, AEP (cl), Xylene (cl)
Estimated Regression Coefficients for OWI (%), Oil E

Constant 26.2714 0.4928 53.308 0.000
Block -0.1857 0.3537 -0.525 0.616
AEP (cl) 5.6667 0.7178 7.895 0.000
Xylene (cl) 4.3333 0.7178 6.037 0.001
AEP (cl)*AEP (cl) 9.2000 1.0525 8.741 0.000
AEP (cl)*Xylene (cl) -34.0000 2.4864 -13.674 0.000
S = 1.24321 PRESS = 53.6335

Analysis of Variance for OWI (%), Oil E

Blocks 1 1.79 0.43 0.426 0.28 0.616
Regression 5 1148.61 1148.61 229.722 148.63 0.000
Linear 2 152.67 152.67 76.333 49.39 0.000
AEP (cl) 1 96.33 96.33 96.333 62.33 0.000
253
Xylene (cl) 1 56.33 56.33 56.333 36.45 0.001
Square 2 706.94 706.94 353.471 228.70 0.000
AEP (cl)*AEP (cl) 1 313.12 118.10 118.102 76.41 0.000
Interaction 1 289.00 289.00 289.000 186.99 0.000
AEP (cl)*Xylene (cl) 1 289.00 289.00 289.000 186.99 0.000
Residual Error 7 10.82 10.82 1.546
Lack-of-Fit 3 6.82 6.82 2.273 2.27 0.222
Pure Error 4 4.00 4.00 1.000
Total 13 1161.21
Estimated Regression Coefficients for OWI (%), Oil E using data in uncoded
units
Term Coef
Constant -1225.66
Block -0.185714
AEP (cl) 24.2111
Xylene (cl) 32.4933
AEP (cl)*AEP (cl) 0.255556
Response Surface Regression: OWI (%), Oil F versus Block, AEP (cl), Xylene (cl)
Estimated Regression Coefficients for OWI (%), Oil F

Constant 21.8714 0.3913 55.901 0.000
Block -0.7190 0.2808 -2.561 0.038
AEP (cl) 3.5000 0.5698 6.142 0.000
Xylene (cl) 4.5000 0.5698 7.897 0.000
AEP (cl)*AEP (cl) 13.9667 0.8355 16.716 0.000
AEP (cl)*Xylene (cl) -21.0000 1.9740 -10.638 0.000
S = 0.986990 PRESS = 11.0918

Analysis of Variance for OWI (%), Oil F

Blocks 1 7.14 6.387 6.387 6.56 0.038
Regression 5 991.75 991.752 198.350 203.61 0.000
Linear 2 97.50 97.500 48.750 50.04 0.000
AEP (cl) 1 36.75 36.750 36.750 37.73 0.000
Xylene (cl) 1 60.75 60.750 60.750 62.36 0.000
Square 2 784.00 784.002 392.001 402.40 0.000
AEP (cl)*AEP (cl) 1 509.21 272.188 272.188 279.41 0.000
Interaction 1 110.25 110.250 110.250 113.18 0.000
AEP (cl)*Xylene (cl) 1 110.25 110.250 110.250 113.18 0.000
Residual Error 7 6.82 6.819 0.974
Lack-of-Fit 3 0.15 0.152 0.051 0.03 0.992
Pure Error 4 6.67 6.667 1.667
Total 13 1005.71
254
Estimated Regression Coefficients for OWI (%), Oil F using data in uncoded
units
Term Coef
Constant -792.487
Block -0.719048
AEP (cl) 5.28889
Xylene (cl) 24.6400
AEP (cl)*AEP (cl) 0.387963
Response Surface Regression: OWI (%), Oil G versus Block, AEP (cl), Xylene (cl)
Estimated Regression Coefficients for OWI (%), Oil G

Constant 37.1857 0.4525 82.176 0.000
Block -0.7762 0.3248 -2.390 0.048
AEP (cl) 6.0000 0.6591 9.104 0.000
Xylene (cl) 7.6667 0.6591 11.633 0.000
AEP (cl)*AEP (cl) 7.3167 0.9664 7.571 0.000
AEP (cl)*Xylene (cl) -30.0000 2.2830 -13.140 0.000
S = 1.14152 PRESS = 49.0858

Analysis of Variance for OWI (%), Oil G

Blocks 1 0.64 7.44 7.442 5.71 0.048
Regression 5 1424.59 1424.59 284.919 218.65 0.000
Linear 2 284.33 284.33 142.167 109.10 0.000
AEP (cl) 1 108.00 108.00 108.000 82.88 0.000
Xylene (cl) 1 176.33 176.33 176.333 135.32 0.000
Square 2 915.26 915.26 457.630 351.20 0.000
AEP (cl)*AEP (cl) 1 289.12 74.70 74.698 57.33 0.000
Interaction 1 225.00 225.00 225.000 172.67 0.000
AEP (cl)*Xylene (cl) 1 225.00 225.00 225.000 172.67 0.000
Residual Error 7 9.12 9.12 1.303
Lack-of-Fit 3 4.45 4.45 1.485 1.27 0.397
Pure Error 4 4.67 4.67 1.167
Total 13 1434.36
Estimated Regression Coefficients for OWI (%), Oil G using data in uncoded
units
Term Coef
Constant -1332.79
Block -0.776190
AEP (cl) 22.4639
Xylene (cl) 36.6867
AEP (cl)*AEP (cl) 0.203241
255
APPENDIX 9: ANALYSIS OF VARIANCE (ANOVA) RESULTS OF OVERALL WAX
INHIBITION FOR CRUDE OIL A-G AEP/XYLENE/BUTYL-ACRYLATE FORMULATION
Response Surface Regression: OWI (%), Oil versus Block, AEP (cl), Xylene (cl),
The analysis was done using uncoded units.
Estimated Regression Coefficients for OWI (%), Oil A

Constant 1195.08 197.388 6.054 0.000
Block 1 0.99 1.667 0.592 0.570
Block 2 0.82 1.667 0.492 0.636
AEP (cl) -56.75 8.753 -6.484 0.000
Xylene (cl) -15.44 3.628 -4.256 0.003
Butyl-acrylate (cl) -98.14 25.340 -3.873 0.005
AEP (cl)*AEP (cl) 0.87 0.115 7.546 0.000
Xylene (cl)*Xylene (cl) 0.07 0.019 3.975 0.004
Butyl-acrylate (cl)* 4.70 1.037 4.533 0.002
Butyl-acrylate (cl)
AEP (cl)*Xylene (cl) 0.26 0.118 2.194 0.040
AEP (cl)*Butyl-acrylate (cl) 1.96 0.589 3.327 0.010
Xylene (cl)*Butyl-acrylate (cl) 0.53 0.353 1.487 0.033
S = 4.99463 PRESS = 7554.43

Analysis of Variance for OWI (%), Oil A
Source DF Seq SS Adj SS Adj MS

Blocks 2 149.72 29.27 14.63
Regression 9 3155.26 3155.26 350.58
Linear 3 1079.75 1294.27 431.42
AEP (cl) 1 189.06 1048.72 1048.72
Xylene (cl) 1 775.13 451.77 451.77
Butyl-acrylate (cl) 1 115.56 374.20 374.20
Square 3 1624.14 1624.14 541.38
AEP (cl)*AEP (cl) 1 920.96 1420.46 1420.46
Xylene (cl)*Xylene (cl) 1 190.65 394.20 394.20
Butyl-acrylate (cl)*Butyl-acrylate (cl) 1 512.52 512.52 512.52
Interaction 3 451.38 451.38 150.46
AEP (cl)*Xylene (cl) 1 120.12 120.13 120.13
AEP (cl)*Butyl-acrylate (cl) 1 276.13 276.13 276.13
Xylene (cl)*Butyl-acrylate (cl) 1 55.13 55.13 55.13
Residual Error 8 199.57 199.57 24.95
Lack-of-Fit 5 194.07 194.07 38.81
Pure Error 3 5.50 5.50 1.83
Total 19 3504.55
Source F P
Blocks 0.59 0.578
Regression 14.05 0.001
Linear 17.29 0.001
AEP (cl) 42.04 0.000
Xylene (cl) 18.11 0.003
256
Butyl-acrylate (cl) 15.00 0.005
Square 21.70 0.000
AEP (cl)*AEP (cl) 56.94 0.000
Xylene (cl)*Xylene (cl) 15.80 0.004
Butyl-acrylate (cl)*Butyl-acrylate (cl) 20.55 0.002
Interaction 6.03 0.019
AEP (cl)*Xylene (cl) 4.82 0.030
AEP (cl)*Butyl-acrylate (cl) 11.07 0.010
Xylene (cl)*Butyl-acrylate (cl) 2.21 0.042
Residual Error
Lack-of-Fit 21.17 0.015
Pure Error
Total
Predicted Response for New Design Points Using Model for OWI (%), Oil A
Point Fit SE Fit 95% CI 95% PI

1 57.0601 3.90824 (48.0477, 66.0725) (42.4355, 71.6847)
2 27.6897 3.95883 (18.5606, 36.8188) (12.9929, 42.3865)
3 21.4134 1.93594 (16.9492, 25.8777) ( 9.0609, 33.7660)
4 26.6897 3.95883 (17.5606, 35.8188) (11.9929, 41.3865)
5 31.8101 3.90824 (22.7977, 40.8225) (17.1855, 46.4347)
6 21.4134 1.93594 (16.9492, 25.8777) ( 9.0609, 33.7660)
7 21.4134 1.93594 (16.9492, 25.8777) ( 9.0609, 33.7660)
8 45.8416 4.15957 (36.2496, 55.4335) (30.8528, 60.8303)
9 62.3623 5.10709 (50.5853, 74.1393) (45.8895, 78.8351) X
10 23.4075 3.69902 (14.8776, 31.9375) ( 9.0752, 37.7399)
11 45.5916 4.15957 (35.9996, 55.1835) (30.6028, 60.5803)
12 34.8416 4.15957 (25.2496, 44.4335) (19.8528, 49.8303)
13 21.4134 1.93594 (16.9492, 25.8777) ( 9.0609, 33.7660)
14 59.5916 4.15957 (49.9996, 69.1835) (44.6028, 74.5803)
15 21.4134 1.93594 (16.9492, 25.8777) ( 9.0609, 33.7660)
16 21.4134 1.93594 (16.9492, 25.8777) ( 9.0609, 33.7660)
17 34.6851 3.90824 (25.6727, 43.6975) (20.0605, 49.3097)
18 30.6851 3.90824 (21.6727, 39.6975) (16.0605, 45.3097)
19 39.3147 3.95883 (30.1856, 48.4438) (24.6179, 54.0115)
20 38.5647 3.95883 (29.4356, 47.6938) (23.8679, 53.2615)
X denotes a point that is an outlier in the predictors.
Response Surface Regression: OWI (%) Oil versus Block, AEP (cl), Xylene (cl),
Estimated Regression Coefficients for OWI (%) Oil B

Constant 1360.78 225.018 6.047 0.000
Block 1 1.15 1.900 0.605 0.562
Block 2 1.15 1.900 0.605 0.562
AEP (cl) -60.94 9.978 -6.107 0.000
Xylene (cl) -18.63 4.135 -4.506 0.002
AEP (cl)*AEP (cl) 0.90 0.131 6.827 0.000
Butyl-acrylate (cl)
AEP (cl)*Xylene (cl) 0.28 0.134 2.049 0.045
257
S = 5.69377 PRESS = 9892.37
Analysis of Variance for OWI (%) Oil B

Blocks 2 151.88 47.14 23.57
Regression 9 3336.52 3336.52 370.72
Linear 3 773.64 1471.05 490.35
AEP (cl) 1 264.06 1209.13 1209.13
Xylene (cl) 1 470.51 658.22 658.22
Square 3 1877.51 1877.51 625.84
AEP (cl)*AEP (cl) 1 874.28 1510.85 1510.85
Interaction 3 685.38 685.38 228.46
AEP (cl)*Xylene (cl) 1 136.12 136.13 136.13
Residual Error 8 259.35 259.35 32.42
Lack-of-Fit 5 258.35 258.35 51.67
Pure Error 3 1.00 1.00 0.33
Total 19 3747.75
Source F P
Blocks 0.73 0.513
Linear 15.13 0.001
AEP (cl) 37.30 0.000
Xylene (cl) 20.30 0.002
Square 19.30 0.001
AEP (cl)*AEP (cl) 46.60 0.000
Residual Error
Lack-of-Fit 155.01 0.001
Pure Error
Total
Predicted Response for New Design Points Using Model for OWI (%) Oil B

1 59.0282 4.45530 (48.7542, 69.3021) (42.3564, 75.6999)
2 30.4740 4.51298 (20.0670, 40.8809) (13.7200, 47.2280)
3 23.4264 2.20693 (18.3373, 28.5156) ( 9.3448, 37.5081)
4 30.4740 4.51298 (20.0670, 40.8809) (13.7200, 47.2280)
5 36.2782 4.45530 (26.0042, 46.5521) (19.6064, 52.9499)
6 23.4264 2.20693 (18.3373, 28.5156) ( 9.3448, 37.5081)
7 23.4264 2.20693 (18.3373, 28.5156) ( 9.3448, 37.5081)
8 47.5852 4.74182 (36.6506, 58.5199) (30.4984, 64.6720)
9 65.2087 5.82196 (51.7833, 78.6342) (46.4301, 83.9873) X
10 32.4783 4.21680 (22.7543, 42.2022) (16.1397, 48.8168)
258
11 46.8352 4.74182 (35.9006, 57.7699) (29.7484, 63.9220)
12 40.5852 4.74182 (29.6506, 51.5199) (23.4984, 57.6720)
13 23.4264 2.20693 (18.3373, 28.5156) ( 9.3448, 37.5081)
14 63.8352 4.74182 (52.9006, 74.7699) (46.7484, 80.9220)
15 23.4264 2.20693 (18.3373, 28.5156) ( 9.3448, 37.5081)
16 23.4264 2.20693 (18.3373, 28.5156) ( 9.3448, 37.5081)
17 36.4032 4.45530 (26.1292, 46.6771) (19.7314, 53.0749)
18 26.4032 4.45530 (16.1292, 36.6771) ( 9.7314, 43.0749)
19 46.8490 4.51298 (36.4420, 57.2559) (30.0950, 63.6030)
20 46.5990 4.51298 (36.1920, 57.0059) (29.8450, 63.3530)
Estimated Regression Coefficients for OWI (%), Oil C

Constant 1015.38 231.788 4.381 0.002
Block 1 0.46 1.957 0.237 0.818
Block 2 0.63 1.957 0.322 0.755
AEP (cl) -52.66 10.278 -5.123 0.001
Xylene (cl) -10.72 4.260 -2.515 0.036
AEP (cl)*AEP (cl) 0.91 0.135 6.758 0.000
Butyl-acrylate (cl)
AEP (cl)*Xylene (cl) 0.15 0.138 1.085 0.030
S = 5.86507 PRESS = 10394.1

Analysis of Variance for OWI (%), Oil C

Blocks 2 195.16 10.80 5.40
Regression 9 3616.85 3616.85 401.87
Linear 3 1480.46 1318.54 439.51
AEP (cl) 1 81.00 902.87 902.87
Xylene (cl) 1 1110.46 217.65 217.65
Square 3 1723.39 1723.39 574.46
AEP (cl)*AEP (cl) 1 1137.76 1570.90 1570.90
Interaction 3 413.00 413.00 137.67
AEP (cl)*Xylene (cl) 1 40.50 40.50 40.50
Residual Error 8 275.19 275.19 34.40
Lack-of-Fit 5 272.19 272.19 54.44
Pure Error 3 3.00 3.00 1.00
Total 19 4087.20
259
Source F P
Blocks 0.16 0.857
Linear 12.78 0.002
AEP (cl) 26.25 0.001
Xylene (cl) 6.33 0.036
Square 16.70 0.001
AEP (cl)*AEP (cl) 45.67 0.000
Residual Error
Lack-of-Fit 54.44 0.004
Pure Error
Total
Predicted Response for New Design Points Using Model for OWI (%), Oil C

1 63.2379 4.58934 (52.6548, 73.8209) (46.0645, 80.4112)
2 30.7185 4.64876 (19.9985, 41.4386) (13.4604, 47.9766)
3 26.1505 2.27332 (20.9082, 31.3928) (11.6452, 40.6558)
4 28.7185 4.64876 (17.9985, 39.4386) (11.4604, 45.9766)
5 39.2379 4.58934 (28.6548, 49.8209) (22.0645, 56.4112)
6 26.1505 2.27332 (20.9082, 31.3928) (11.6452, 40.6558)
7 26.1505 2.27332 (20.9082, 31.3928) (11.6452, 40.6558)
8 54.5695 4.88448 (43.3059, 65.8331) (36.9686, 72.1704)
9 67.0803 5.99713 (53.2509, 80.9097) (47.7367, 86.4239) X
10 22.1039 4.34367 (12.0874, 32.1205) ( 5.2738, 38.9341)
11 53.5695 4.88448 (42.3059, 64.8331) (35.9686, 71.1704)
12 36.5695 4.88448 (25.3059, 47.8331) (18.9686, 54.1704)
13 26.1505 2.27332 (20.9082, 31.3928) (11.6452, 40.6558)
14 63.5695 4.88448 (52.3059, 74.8331) (45.9686, 81.1704)
15 26.1505 2.27332 (20.9082, 31.3928) (11.6452, 40.6558)
16 26.1505 2.27332 (20.9082, 31.3928) (11.6452, 40.6558)
17 36.2379 4.58934 (25.6548, 46.8209) (19.0645, 53.4112)
18 42.2379 4.58934 (31.6548, 52.8209) (25.0645, 59.4112)
19 40.7185 4.64876 (29.9985, 51.4386) (23.4604, 57.9766)
20 42.7185 4.64876 (31.9985, 53.4386) (25.4604, 59.9766)
Estimated Regression Coefficients for OWI (%), Oil D

Constant 1256.28 191.065 6.575 0.000
Block 1 1.95 1.613 1.211 0.261
Block 2 -1.21 1.613 -0.752 0.473
AEP (cl) -57.08 8.472 -6.737 0.000
Xylene (cl) -16.33 3.511 -4.651 0.002
AEP (cl)*AEP (cl) 0.82 0.112 7.364 0.000
260
Butyl-acrylate (cl)
AEP (cl)*Xylene (cl) 0.27 0.114 2.340 0.047
S = 4.83464 PRESS = 6760.34

Analysis of Variance for OWI (%), Oil D

Blocks 2 300.08 34.97 17.49
Regression 9 3348.68 3348.68 372.08
Linear 3 955.68 1443.79 481.26
AEP (cl) 1 121.00 1060.97 1060.97
Xylene (cl) 1 545.68 505.69 505.69
Square 3 1754.49 1754.49 584.83
AEP (cl)*AEP (cl) 1 686.26 1267.40 1267.40
Interaction 3 638.50 638.50 212.83
AEP (cl)*Xylene (cl) 1 128.00 128.00 128.00
Residual Error 8 186.99 186.99 23.37
Lack-of-Fit 5 177.99 177.99 35.60
Pure Error 3 9.00 9.00 3.00
Total 19 3835.75
Source F P
Blocks 0.75 0.504
Linear 20.59 0.000
AEP (cl) 45.39 0.000
Xylene (cl) 21.63 0.002
Square 25.02 0.000
AEP (cl)*AEP (cl) 54.22 0.000
Residual Error
Lack-of-Fit 11.87 0.034
Pure Error
Total
Predicted Response for New Design Points Using Model for OWI (%), Oil D

1 54.1385 3.78304 (45.4148, 62.8622) (39.9823, 68.2946)
2 24.5395 3.83202 (15.7028, 33.3761) (10.3134, 38.7655)
3 15.4624 1.87393 (11.1411, 19.7837) ( 3.5055, 27.4193)
4 19.0395 3.83202 (10.2028, 27.8761) ( 4.8134, 33.2655)
261
5 26.6385 3.78304 (17.9148, 35.3622) (12.4823, 40.7946)
6 15.4624 1.87393 (11.1411, 19.7837) ( 3.5055, 27.4193)
7 15.4624 1.87393 (11.1411, 19.7837) ( 3.5055, 27.4193)
8 39.5309 4.02633 (30.2462, 48.8157) (25.0223, 54.0395)
9 54.1372 4.94349 (42.7374, 65.5369) (38.1920, 70.0823) X
10 19.7326 3.58053 (11.4758, 27.9893) ( 5.8593, 33.6058)
11 49.0309 4.02633 (39.7462, 58.3157) (34.5223, 63.5395)
12 32.0309 4.02633 (22.7462, 41.3157) (17.5223, 46.5395)
13 15.4624 1.87393 (11.1411, 19.7837) ( 3.5055, 27.4193)
14 50.5309 4.02633 (41.2462, 59.8157) (36.0223, 65.0395)
15 15.4624 1.87393 (11.1411, 19.7837) ( 3.5055, 27.4193)
16 15.4624 1.87393 (11.1411, 19.7837) ( 3.5055, 27.4193)
17 25.1385 3.78304 (16.4148, 33.8622) (10.9823, 39.2946)
18 25.6385 3.78304 (16.9148, 34.3622) (11.4823, 39.7946)
19 36.5395 3.83202 (27.7028, 45.3761) (22.3134, 50.7655)
20 37.0395 3.83202 (28.2028, 45.8761) (22.8134, 51.2655)
Estimated Regression Coefficients for OWI (%), Oil E

Constant 633.775 230.516 2.749 0.025
Block 1 0.488 1.946 0.251 0.808
Block 2 0.155 1.946 0.080 0.939
AEP (cl) -32.690 10.222 -3.198 0.013
Xylene (cl) -6.260 4.237 -1.478 0.018
AEP (cl)*AEP (cl) 0.596 0.135 4.431 0.002
Butyl-acrylate (cl)
AEP (cl)*Xylene (cl) 0.108 0.137 0.788 0.043
S = 5.83289 PRESS = 10467.6

Analysis of Variance for OWI (%), Oil E

Blocks 2 99.34 4.03 2.014
Regression 9 2191.03 2191.03 243.447
Linear 3 1141.63 636.34 212.114
AEP (cl) 1 33.06 347.99 347.988
Xylene (cl) 1 993.00 74.28 74.280
Square 3 854.03 854.03 284.675
AEP (cl)*AEP (cl) 1 484.46 667.99 667.994
Interaction 3 195.38 195.38 65.125
AEP (cl)*Xylene (cl) 1 21.13 21.13 21.125
262
Residual Error 8 272.18 272.18 34.023
Lack-of-Fit 5 267.68 267.68 53.536
Pure Error 3 4.50 4.50 1.500
Total 19 2562.55
Source F P
Blocks 0.06 0.943
Linear 6.23 0.017
AEP (cl) 10.23 0.013
Xylene (cl) 2.18 0.017
Square 8.37 0.008
AEP (cl)*AEP (cl) 19.63 0.002
Residual Error
Lack-of-Fit 35.69 0.077
Pure Error
Total
Predicted Response for New Design Points Using Model for OWI (%), Oil E

1 45.7772 4.56416 (35.2522, 56.3022) (28.6981, 62.8563)
2 19.4579 4.62325 ( 8.7967, 30.1192) ( 2.2945, 36.6213)
3 18.4251 2.26085 (13.2116, 23.6387) ( 3.9994, 32.8509)
4 22.4579 4.62325 (11.7967, 33.1192) ( 5.2945, 39.6213)
5 25.5272 4.56416 (15.0022, 36.0522) ( 8.4481, 42.6063)
6 18.4251 2.26085 (13.2116, 23.6387) ( 3.9994, 32.8509)
7 18.4251 2.26085 (13.2116, 23.6387) ( 3.9994, 32.8509)
8 37.0165 4.85768 (25.8147, 48.2184) (19.5122, 54.5209)
9 49.6612 5.96422 (35.9077, 63.4147) (30.4237, 68.8987) X
10 9.0678 4.31984 (-0.8938, 19.0293) (-7.6700, 25.8055)
11 39.7665 4.85768 (28.5647, 50.9684) (22.2622, 57.2709)
12 29.0165 4.85768 (17.8147, 40.2184) (11.5122, 46.5209)
13 18.4251 2.26085 (13.2116, 23.6387) ( 3.9994, 32.8509)
14 42.7665 4.85768 (31.5647, 53.9684) (25.2622, 60.2709)
15 18.4251 2.26085 (13.2116, 23.6387) ( 3.9994, 32.8509)
16 18.4251 2.26085 (13.2116, 23.6387) ( 3.9994, 32.8509)
17 28.4022 4.56416 (17.8772, 38.9272) (11.3231, 45.4813)
18 36.4022 4.56416 (25.8772, 46.9272) (19.3231, 53.4813)
19 26.0829 4.62325 (15.4217, 36.7442) ( 8.9195, 43.2463)
20 22.3329 4.62325 (11.6717, 32.9942) ( 5.1695, 39.4963)
Estimated Regression Coefficients for OWI (%), Oil F

Constant 1177.51 229.112 5.139 0.001
263
Block 1 0.22 1.934 0.112 0.914
Block 2 0.72 1.934 0.370 0.721
AEP (cl) -48.79 10.159 -4.802 0.001
Xylene (cl) -17.10 4.211 -4.061 0.004
AEP (cl)*AEP (cl) 0.60 0.134 4.476 0.002
Butyl-acrylate (cl)
AEP (cl)*Xylene (cl) 0.34 0.137 2.500 0.037
S = 5.79736 PRESS = 9739.83

Analysis of Variance for OWI (%), Oil F

Blocks 2 175.96 8.53 4.264
Regression 9 2976.17 2976.17 330.685
Linear 3 1341.04 1112.52 370.841
AEP (cl) 1 315.06 775.09 775.092
Xylene (cl) 1 761.91 554.29 554.289
Square 3 1117.76 1117.76 372.585
AEP (cl)*AEP (cl) 1 304.47 673.21 673.207
Interaction 3 517.38 517.38 172.458
AEP (cl)*Xylene (cl) 1 210.12 210.13 210.125
Residual Error 8 268.87 268.87 33.609
Lack-of-Fit 5 255.87 255.87 51.175
Pure Error 3 13.00 13.00 4.333
Total 19 3421.00
Source F P
Blocks 0.13 0.883
Linear 11.03 0.003
AEP (cl) 23.06 0.001
Xylene (cl) 16.49 0.004
Square 11.09 0.003
AEP (cl)*AEP (cl) 20.03 0.002
Residual Error
Lack-of-Fit 11.81 0.055
Pure Error
Total
Predicted Response for New Design Points Using Model for OWI (%), Oil F
264
1 63.1333 4.53636 (52.6725, 73.5942) (46.1583, 80.1084)
2 30.8583 4.59509 (20.2620, 41.4546) (13.7995, 47.9172)
3 25.2333 2.24708 (20.0516, 30.4151) (10.8955, 39.5712)
4 29.6083 4.59509 (19.0120, 40.2046) (12.5495, 46.6672)
5 27.6333 4.53636 (17.1725, 38.0942) (10.6583, 44.6084)
6 25.2333 2.24708 (20.0516, 30.4151) (10.8955, 39.5712)
7 25.2333 2.24708 (20.0516, 30.4151) (10.8955, 39.5712)
8 37.9083 4.82809 (26.7747, 49.0419) (20.5106, 55.3060)
9 63.8333 5.92789 (50.1636, 77.5031) (44.7131, 82.9536) X
10 26.0333 4.29352 (16.1325, 35.9342) ( 9.3975, 42.6692)
11 55.4083 4.82809 (44.2747, 66.5419) (38.0106, 72.8060)
12 39.1583 4.82809 (28.0247, 50.2919) (21.7606, 56.5560)
13 25.2333 2.24708 (20.0516, 30.4151) (10.8955, 39.5712)
14 55.6583 4.82809 (44.5247, 66.7919) (38.2606, 73.0560)
15 25.2333 2.24708 (20.0516, 30.4151) (10.8955, 39.5712)
16 25.2333 2.24708 (20.0516, 30.4151) (10.8955, 39.5712)
17 37.5083 4.53636 (27.0475, 47.9692) (20.5333, 54.4834)
18 34.7583 4.53636 (24.2975, 45.2192) (17.7833, 51.7334)
19 40.2333 4.59509 (29.6370, 50.8296) (23.1745, 57.2922)
20 38.7333 4.59509 (28.1370, 49.3296) (21.6745, 55.7922)
Estimated Regression Coefficients for OWI (%), Oil G

Constant 1279.82 174.676 7.327 0.000
Block 1 0.85 1.475 0.576 0.580
Block 2 -0.32 1.475 -0.215 0.835
AEP (cl) -60.43 7.745 -7.802 0.000
Xylene (cl) -14.94 3.210 -4.653 0.002
AEP (cl)*AEP (cl) 0.82 0.102 8.000 0.000
Butyl-acrylate (cl)
AEP (cl)*Xylene (cl) 0.33 0.104 3.200 0.013
S = 4.41994 PRESS = 5699.98

Analysis of Variance for OWI (%), Oil G

Blocks 2 176.88 6.62 3.31
Regression 9 3235.38 3235.38 359.49
Linear 3 953.30 1797.92 599.31
AEP (cl) 1 361.00 1189.22 1189.22
Xylene (cl) 1 502.05 422.87 422.87
Square 3 1401.58 1401.58 467.19
265
AEP (cl)*AEP (cl) 1 900.57 1250.42 1250.42
Interaction 3 880.50 880.50 293.50
AEP (cl)*Xylene (cl) 1 200.00 200.00 200.00
Residual Error 8 156.29 156.29 19.54
Lack-of-Fit 5 149.79 149.79 29.96
Pure Error 3 6.50 6.50 2.17
Total 19 3568.55
Source F P
Blocks 0.17 0.847
Linear 30.68 0.000
AEP (cl) 60.87 0.000
Xylene (cl) 21.65 0.002
Square 23.91 0.000
AEP (cl)*AEP (cl) 64.01 0.000
Residual Error
Lack-of-Fit 13.83 0.028
Pure Error
Total
Predicted Response for New Design Points Using Model for OWI (%), Oil G

1 68.1730 3.45854 (60.1976, 76.1484) (55.2311, 81.1149)
2 27.1864 3.50332 (19.1078, 35.2651) (14.1807, 40.1922)
3 27.3417 1.71319 (23.3911, 31.2924) (16.4105, 38.2730)
4 31.9364 3.50332 (23.8578, 40.0151) (18.9307, 44.9422)
5 33.9230 3.45854 (25.9476, 41.8984) (20.9811, 46.8649)
6 27.3417 1.71319 (23.3911, 31.2924) (16.4105, 38.2730)
7 27.3417 1.71319 (23.3911, 31.2924) (16.4105, 38.2730)
8 47.2116 3.68096 (38.7233, 55.6999) (33.9475, 60.4757)
9 57.7634 4.51945 (47.3415, 68.1853) (43.1860, 72.3408) X
10 25.9673 3.27341 (18.4188, 33.5158) (13.2841, 38.6506)
11 49.9616 3.68096 (41.4733, 58.4499) (36.6975, 63.2257)
12 40.4616 3.68096 (31.9733, 48.9499) (27.1975, 53.7257)
13 27.3417 1.71319 (23.3911, 31.2924) (16.4105, 38.2730)
14 66.2116 3.68096 (57.7233, 74.6999) (52.9475, 79.4757)
15 27.3417 1.71319 (23.3911, 31.2924) (16.4105, 38.2730)
16 27.3417 1.71319 (23.3911, 31.2924) (16.4105, 38.2730)
17 37.9230 3.45854 (29.9476, 45.8984) (24.9811, 50.8649)
18 33.1730 3.45854 (25.1976, 41.1484) (20.2311, 46.1149)
19 47.9364 3.50332 (39.8578, 56.0151) (34.9307, 60.9422)
20 42.1864 3.50332 (34.1078, 50.2651) (29.1807, 55.1922)
266
APPENDIX 10: ANALYSIS OF VARIANCE RESULTS (ANOVA) OF WAX DEPOSITION
FOR CRUDE OIL A-G (4 VARIABLES, LAMINAR FLOW REGIME)
————— 4/24/2014 7:23:41 AM ————————————————————
Results for: Minitab.MTW
Response Surface Regression: Crude Oil A, Crude Oil B, ... versus Block, Coolan
Response Surface Regression: Crude Oil A versus Block, Coolant Temp, Bulk Oil T
Estimated Regression Coefficients for Crude Oil A

Constant 113.825 0.4597 247.619 0.000
Block 0.525 0.2154 2.438 0.029
Coolant Temp. (K) 18.583 0.2270 81.858 0.000
Bulk Oil Temp. (K) 21.250 0.2270 93.604 0.000
Wax Inhib. Perc. (%) 24.250 0.2270 106.819 0.000
Flow Rate (Ltr./min.) 9.583 0.2270 42.214 0.235
Coolant Temp. (K)*Coolant Temp. (K) 5.250 0.2124 24.723 0.000
Bulk Oil Temp. (K)* 5.500 0.2124 25.900 0.000
Bulk Oil Temp. (K)
Wax Inhib. Perc. (%)* 6.500 0.2124 30.609 0.000
Flow Rate (Ltr./min)* 2.500 0.2124 11.773 0.051
Flow Rate (Ltr./min)
Coolant Temp. (K)*Bulk Oil Temp. (K) 18.500 0.2780 66.537 0.000
Coolant Temp. (K)* 7.250 0.2780 26.075 0.000
Coolant Temp. (K)* -0.500 0.2780 -1.798 0.094
Flow Rate (Ltr./min.)
Bulk Oil Temp. (K)* 9.250 0.2780 33.269 0.000
Bulk Oil Temp. (K)* 0.000 0.2780 0.000 1.000
Wax Inhib. Perc. (%)* -0.250 0.2780 -0.899 0.384
S = 1.11216 PRESS = 91.8680

267
Analysis of Variance for Crude Oil A

Blocks 1 7.3 7.4 7.4
Regression 14 45288.1 45288.1 3234.9
Linear 4 35443.3 35443.3 8860.8
Coolant Temp. (K) 1 8288.2 8288.2 8288.2
Bulk Oil Temp. (K) 1 10837.5 10837.5 10837.5
Wax Inhib. Perc. (%) 1 14113.5 14113.5 14113.5
Flow Rate (Ltr./min.) 1 2204.2 2204.2 2204.2
Square 4 2153.8 2153.8 538.5
Coolant Temp. (K)*Coolant Temp. (K) 1 381.4 756.0 756.0
Bulk Oil Temp. (K)*Bulk Oil Temp. (K) 1 544.4 829.7 829.7
Wax Inhib. Perc. (%)*Wax Inhib. Perc. (%) 1 1056.6 1158.9 1158.9
Flow Rate (Ltr./min.)*Flow Rate (Ltr./min.) 1 171.4 171.4 171.4
Interaction 6 7691.0 7691.0 1281.8
Coolant Temp. (K)*Bulk Oil Temp. (K) 1 5476.0 5476.0 5476.0
Coolant Temp. (K)*Wax Inhib. Perc. (%) 1 841.0 841.0 841.0
Coolant Temp. (K)*Flow Rate (Ltr./min.) 1 4.0 4.0 4.0
Bulk Oil Temp. (K)*Wax Inhib. Perc. (%) 1 1369.0 1369.0 1369.0
Bulk Oil Temp. (K)*Flow Rate (Ltr./min.) 1 0.0 0.0 0.0
Wax Inhib. Perc. (%)*Flow Rate (Ltr./min.) 1 1.0 1.0 1.0
Lack-of-Fit 10 14.1 14.1 1.4
Pure Error 4 3.3 3.3 0.8
Total 29 45312.8
Source F P
Blocks 5.94 0.029
Linear 7163.72 0.000
Coolant Temp. (K) 6700.73 0.000
Bulk Oil Temp. (K) 8761.79 0.000
Wax Inhib. Perc. (%) 11410.34 0.000
Flow Rate (Ltr./min.) 1782.00 0.000
Square 435.32 0.000
Coolant Temp. (K)*Coolant Temp. (K) 611.20 0.000
Bulk Oil Temp. (K)*Bulk Oil Temp. (K) 670.80 0.000
Wax Inhib. Perc. (%)*Wax Inhib. Perc. (%) 936.90 0.000
Flow Rate (Ltr./min.)*Flow Rate (Ltr./min.) 138.59 0.000
Coolant Temp. (K)*Bulk Oil Temp. (K) 4427.18 0.000
Coolant Temp. (K)*Wax Inhib. Perc. (%) 679.92 0.000
Coolant Temp. (K)*Flow Rate (Ltr./min.) 3.23 0.094
Bulk Oil Temp. (K)*Wax Inhib. Perc. (%) 1106.79 0.080
Bulk Oil Temp. (K)*Flow Rate (Ltr./min.) 0.00 1.000
Wax Inhib. Perc. (%)*Flow Rate (Ltr./min.) 0.81 0.384
Residual Error
Lack-of-Fit 1.73 0.314
Pure Error
Total
Estimated Regression Coefficients for Crude Oil A using data in uncoded units
Term Coef
Constant 113.825
Block 0.525000
Coolant Temp. (K) 18.5833
Bulk Oil Temp. (K) 21.2500
Wax Inhib. Perc. (%) 24.2500
Flow Rate (Ltr./min.) 9.58333
268
Coolant Temp. (K)*Coolant Temp. (K) 5.25000
Bulk Oil Temp. (K)* 5.50000
Bulk Oil Temp. (K)
Wax Inhib. Perc. (%)* 6.50000
Flow Rate (Ltr./min.)* 2.50000
Coolant Temp. (K)*Bulk Oil Temp. (K) 18.5000
Coolant Temp. (K)* 7.25000
Coolant Temp. (K)* -0.500000
Bulk Oil Temp. (K)* 2.65036E-15
Wax Inhib. Perc. (%)* -0.250000
Response Surface Regression: Crude Oil B versus Block, Coolant Temp, Bulk Oil T
Estimated Regression Coefficients for Crude Oil B

Constant 105.592 0.7108 148.562 0.000
Block 0.225 0.3330 0.676 0.510
Coolant Temp. (K) 22.792 0.3510 64.930 0.000
Bulk Oil Temp. (K) 18.292 0.3510 52.110 0.000
Wax Inhib. Perc. (%) 25.208 0.3510 71.815 0.000
Flow Rate (Ltr./min.) 11.042 0.3510 31.456 0.301
Bulk Oil Temp. (K)* 6.635 0.3283 20.208 0.000
Bulk Oil Temp. (K)
Wax Inhib. Perc. (%)* 12.885 0.3283 39.243 0.000
Flow Rate (Ltr./min.)* 5.635 0.3283 17.163 0.012
Coolant Temp. (K)* 9.938 0.4299 23.115 0.000
Coolant Temp. (K)* 0.187 0.4299 0.436 0.669
Bulk Oil Temp. (K)* 9.063 0.4299 21.080 0.000
Bulk Oil Temp. (K)* 0.063 0.4299 0.145 0.886
Wax Inhib. Perc. (%)* 0.188 0.4299 0.436 0.669
S = 1.71963 PRESS = 142.409

Analysis of Variance for Crude Oil B

Blocks 1 1.4 1.4 1.4
Regression 14 57262.5 57262.5 4090.2
269
Linear 4 38674.2 38674.2 9668.5
Coolant Temp. (K) 1 12467.0 12467.0 12467.0
Bulk Oil Temp. (K) 1 8030.0 8030.0 8030.0
Wax Inhib. Perc. (%) 1 15251.0 15251.0 15251.0
Flow Rate (Ltr./min.) 1 2926.0 2926.0 2926.0
Square 4 6428.9 6428.9 1607.2
Interaction 6 12159.4 12159.4 2026.6
Lack-of-Fit 10 13.4 13.4 1.3
Pure Error 4 28.0 28.0 7.0
Total 29 57305.2
Source F P
Blocks 0.46 0.510
Linear 3269.56 0.000
Coolant Temp. (K) 4215.91 0.000
Bulk Oil Temp. (K) 2715.47 0.000
Wax Inhib. Perc. (%) 5157.36 0.000
Square 543.51 0.000
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Crude Oil B using data in uncoded units
Term Coef
Constant 105.592
Block 0.225000
Bulk Oil Temp. (K)
270
Response Surface Regression: Crude Oil C versus Block, Coolant Temp, Bulk Oil T
Estimated Regression Coefficients for Crude Oil C

Constant 119.592 0.4928 242.690 0.000
Block -0.275 0.2309 -1.191 0.253
Coolant Temp. (K) 22.958 0.2434 94.337 0.000
Bulk Oil Temp. (K) 17.542 0.2434 72.080 0.000
Wax Inhib. Perc. (%) 22.208 0.2434 91.256 0.000
Flow Rate (Ltr./min.) 7.208 0.2434 29.620 0.147
Bulk Oil Temp. (K)* 8.885 0.2276 39.032 0.000
Bulk Oil Temp. (K)
Wax Inhib. Perc. (%)* 8.635 0.2276 37.933 0.000
Flow Rate (Ltr./min.)* 6.760 0.2276 29.697 0.068
Coolant Temp. (K)* 11.312 0.2981 37.954 0.000
Coolant Temp. (K)* 0.188 0.2981 0.629 0.539
Bulk Oil Temp. (K)* 9.312 0.2981 31.244 0.000
Bulk Oil Temp. (K)* -0.562 0.2981 -1.887 0.080
Wax Inhib. Perc. (%)* -1.062 0.2981 -3.565 0.063
S = 1.19224 PRESS = 105.011

Analysis of Variance for Crude Oil C

Blocks 1 2.0 2.0 2.0
Regression 14 51787.6 51787.6 3699.1
Linear 4 33119.2 33119.2 8279.8
Coolant Temp. (K) 1 12650.0 12650.0 12650.0
Bulk Oil Temp. (K) 1 7385.0 7385.0 7385.0
Wax Inhib. Perc. (%) 1 11837.0 11837.0 11837.0
Flow Rate (Ltr./min.) 1 1247.0 1247.0 1247.0
271
Square 4 4549.0 4549.0 1137.3
Interaction 6 14119.4 14119.4 2353.2
Lack-of-Fit 10 14.4 14.4 1.4
Pure Error 4 5.5 5.5 1.4
Total 29 51809.5
Source F P
Blocks 1.42 0.253
Linear 5824.98 0.000
Coolant Temp. (K) 8899.53 0.000
Bulk Oil Temp. (K) 5195.51 0.000
Wax Inhib. Perc. (%) 8327.57 0.000
Square 800.08 0.000
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Crude Oil C using data in uncoded units
Term Coef
Constant 119.592
Block -0.275000
Flow Rate (Ltr./hour) 7.20833
Bulk Oil Temp. (K)
Flow Rate (Ltr./hour)* 6.76042
Flow Rate (Ltr./hour)
272
Bulk Oil Temp. (K)* -0.562500
Wax Inhib. Perc. (%)* -1.06250
Response Surface Regression: Crude Oil D versus Block, Coolant Temp, Bulk Oil T
Estimated Regression Coefficients for Crude Oil D

Constant 96.5167 0.6675 144.589 0.000
Block -0.0500 0.3127 -0.160 0.875
Coolant Temp. (K) 20.3750 0.3297 61.805 0.000
Bulk Oil Temp. (K) 22.6250 0.3297 68.630 0.000
Wax Inhib. Perc. (%) 27.1250 0.3297 82.280 0.000
Flow Rate (Ltr./min.) 11.6250 0.3297 35.263 0.182
Bulk Oil Temp. (K)* 8.5104 0.3084 27.598 0.000
Bulk Oil Temp. (K)
Wax Inhib. Perc. (%)* 9.6354 0.3084 31.246 0.000
Flow Rate (Ltr./min.)* 6.0104 0.3084 19.491 0.071
Coolant Temp. (K)* 8.6875 0.4038 21.517 0.000
Coolant Temp. (K)* 0.0625 0.4038 0.155 0.879
Bulk Oil Temp. (K)* 8.9375 0.4038 22.136 0.000
Bulk Oil Temp. (K)* -0.6875 0.4038 -1.703 0.111
Wax Inhib. Perc. (%)* 0.1875 0.4038 0.464 0.650
S = 1.61503 PRESS = 137.458

Analysis of Variance for Crude Oil D

Blocks 1 0.1 0.1 0.1
Regression 14 59276.9 59276.9 4234.1
Linear 4 43150.5 43150.5 10787.6
Coolant Temp. (K) 1 9963.4 9963.4 9963.4
Bulk Oil Temp. (K) 1 12285.4 12285.4 12285.4
Wax Inhib. Perc. (%) 1 17658.4 17658.4 17658.4
Flow Rate (Ltr./min.) 1 3243.4 3243.4 3243.4
Square 4 5306.0 5306.0 1326.5
273
Interaction 6 10820.4 10820.4 1803.4
Lack-of-Fit 10 15.8 15.8 1.6
Pure Error 4 20.8 20.8 5.2
Total 29 59313.5
Source F P
Blocks 0.03 0.875
Linear 4135.83 0.000
Coolant Temp. (K) 3819.82 0.000
Bulk Oil Temp. (K) 4710.05 0.000
Wax Inhib. Perc. (%) 6769.98 0.000
Square 508.56 0.000
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Crude Oil D using data in uncoded units
Term Coef
Constant 96.5167
Block -0.0500000
Bulk Oil Temp. (K)
274
Response Surface Regression: Crude Oil E versus Block, Coolant Temp, Bulk Oil T
Estimated Regression Coefficients for Crude Oil E

Constant 79.3667 0.6082 130.489 0.000
Block -0.1000 0.2850 -0.351 0.731
Coolant Temp. (K) 12.6667 0.3004 42.169 0.000
Bulk Oil Temp. (K) 13.3333 0.3004 44.388 0.000
Wax Inhib. Perc. (%) 18.0000 0.3004 59.924 0.000
Flow Rate (Ltr./min.) 7.0833 0.3004 23.581 0.216
Bulk Oil Temp. (K)* 9.7292 0.2810 34.626 0.000
Bulk Oil Temp. (K)
Wax Inhib. Perc. (%)* 9.9792 0.2810 35.516 0.000
Flow Rate (Ltr./min.)* 6.4792 0.2810 23.059 0.077
Coolant Temp. (K)* 5.8750 0.3679 15.969 0.000
Coolant Temp. (K)* -0.2500 0.3679 -0.680 0.508
Bulk Oil Temp. (K)* 4.8750 0.3679 13.251 0.000
Bulk Oil Temp. (K)* 0.0000 0.3679 0.000 1.000
Wax Inhib. Perc. (%)* 0.5000 0.3679 1.359 0.196
S = 1.47156 PRESS = 174.548

Analysis of Variance for Crude Oil E

Blocks 1 0.3 0.3 0.27
Regression 14 25136.9 25136.9 1795.49
Linear 4 17097.5 17097.5 4274.38
Coolant Temp. (K) 1 3850.7 3850.7 3850.67
Bulk Oil Temp. (K) 1 4266.7 4266.7 4266.67
Wax Inhib. Perc. (%) 1 7776.0 7776.0 7776.00
Flow Rate (Ltr./min.) 1 1204.2 1204.2 1204.17
Square 4 5121.6 5121.6 1280.41
Interaction 6 2917.8 2917.8 486.29
275
Lack-of-Fit 10 23.3 23.3 2.33
Pure Error 4 7.0 7.0 1.75
Total 29 25167.5
Source F P
Blocks 0.12 0.731
Linear 1973.87 0.000
Coolant Temp. (K) 1778.21 0.000
Bulk Oil Temp. (K) 1970.31 0.000
Wax Inhib. Perc. (%) 3590.90 0.000
Square 591.28 0.000
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Crude Oil E using data in uncoded units
Term Coef
Constant 79.3667
Block -0.100000
Bulk Oil Temp. (K)
276
Response Surface Regression: Crude Oil F versus Block, Coolant Temp, Bulk Oil T
Estimated Regression Coefficients for Crude Oil F

Constant 93.8333 0.7102 132.125 0.000
Block -0.5000 0.3327 -1.503 0.155
Coolant Temp. (K) 13.6667 0.3507 38.966 0.000
Bulk Oil Temp. (K) 15.5833 0.3507 44.430 0.000
Wax Inhib. Perc. (%) 17.7500 0.3507 50.608 0.000
Flow Rate (Ltr./min.) 5.0000 0.3507 14.256 0.315
Bulk Oil Temp. (K)* 8.7708 0.3281 26.734 0.000
Bulk Oil Temp. (K)
Wax Inhib. Perc. (%)* 5.3958 0.3281 16.447 0.000
Flow Rate (Ltr./min.)* 4.2708 0.3281 13.018 0.077
Coolant Temp. (K)* 8.2500 0.4296 19.206 0.000
Coolant Temp. (K)* 0.0000 0.4296 0.000 1.000
Bulk Oil Temp. (K)* 10.2500 0.4296 23.861 0.000
Bulk Oil Temp. (K)* 0.0000 0.4296 0.000 1.000
Wax Inhib. Perc. (%)* -0.1250 0.4296 -0.291 0.775
S = 1.71825 PRESS = 205.900

Analysis of Variance for Crude Oil F

Blocks 1 6.7 6.7 6.67
Regression 14 27733.9 27733.9 1980.99
Linear 4 18472.3 18472.3 4618.08
Coolant Temp. (K) 1 4482.7 4482.7 4482.67
Bulk Oil Temp. (K) 1 5828.2 5828.2 5828.17
Wax Inhib. Perc. (%) 1 7561.5 7561.5 7561.50
Flow Rate (Ltr./min.) 1 600.0 600.0 600.00
Square 4 3839.0 3839.0 959.76
Interaction 6 5422.5 5422.5 903.75
Lack-of-Fit 10 28.3 28.3 2.83
Pure Error 4 13.0 13.0 3.25
277
Total 29 27781.9
Source F P
Blocks 2.26 0.155
Linear 1564.19 0.000
Coolant Temp. (K) 1518.32 0.000
Bulk Oil Temp. (K) 1974.06 0.000
Wax Inhib. Perc. (%) 2561.15 0.000
Square 325.08 0.000
Wax Inhib. Perc. (%)*Flow Rate (Ltr./min) 0.08 0.775
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Crude Oil F using data in uncoded units
Term Coef
Constant 93.8333
Block -0.500000
Bulk Oil Temp. (K)
Coolant Temp. (K)* 1.04886E-15
Wax Inhib. Perc. (%)* -0.125000
Response Surface Regression: Crude Oil G versus Block, Coolant Temp, Bulk Oil T
Estimated Regression Coefficients for Crude Oil G
278
Constant 101.967 0.7875 129.489 0.000
Block -0.400 0.3689 -1.084 0.297
Coolant Temp. (K) 19.333 0.3889 49.713 0.000
Bulk Oil Temp. (K) 18.917 0.3889 48.642 0.000
Wax Inhib. Perc. (%) 25.667 0.3889 65.999 0.000
Flow Rate (Ltr./min.) 11.000 0.3889 28.285 0.291
Bulk Oil Temp. (K)* 6.042 0.3638 16.608 0.000
Bulk Oil Temp. (K)
Wax Inhib. Perc. (%)* 9.042 0.3638 24.855 0.000
Flow Rate (Ltr./min.)* 4.792 0.3638 13.172 0.075
Coolant Temp. (K)* 10.000 0.4763 20.995 0.000
Coolant Temp. (K)* -0.250 0.4763 -0.525 0.608
Bulk Oil Temp. (K)* 7.500 0.4763 15.746 0.000
Bulk Oil Temp. (K)* -0.250 0.4763 -0.525 0.608
Wax Inhib. Perc. (%)* -0.375 0.4763 -0.787 0.444
S = 1.90519 PRESS = 232.557

Analysis of Variance for Crude Oil G

Blocks 1 4.3 4.3 4.3
Regression 14 46578.4 46578.4 3327.0
Linear 4 36273.5 36273.5 9068.4
Coolant Temp. (K) 1 8970.7 8970.7 8970.7
Bulk Oil Temp. (K) 1 8588.2 8588.2 8588.2
Wax Inhib. Perc. (%) 1 15810.7 15810.7 15810.7
Flow Rate (Ltr./min.) 1 2904.0 2904.0 2904.0
Square 4 3510.4 3510.4 877.6
Interaction 6 6794.5 6794.5 1132.4
Lack-of-Fit 10 29.3 29.3 2.9
Pure Error 4 21.5 21.5 5.4
Total 29 46633.5
Source F P
Blocks 1.18 0.297
279
Linear 2498.34 0.000
Coolant Temp. (K) 2471.42 0.000
Bulk Oil Temp. (K) 2366.04 0.000
Wax Inhib. Perc. (%) 4355.84 0.000
Square 241.78 0.000
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Crude Oil G using data in uncoded units
Term Coef
Constant 101.967
Block -0.400000
Bulk Oil Temp. (K)
Wax Inhib. Perc. (%)* -0.375000
280
FOR CRUDE OIL A-G (4 VARIABLES, TURBULENT FLOW REGIME)
————— 5/18/2016 9:21:29 AM ————————————————————
Welcome to Minitab, press F1 for help.

Retrieving project from file: 'C:\USERS\USER\DOCUMENTS\SIJI 30 - COPY.MPJ'
Response Surface Regression: Oil A (4 Com, Oil B (4 Com, ... versus Block, Cool
Response Surface Regression: Oil A (4 Com versus Block, Coolant temp, Oil temp.
Estimated Regression Coefficients for Oil A (4 Components, Turb.)

Constant 114.108 1.3110 87.040 0.000
Block -1.325 0.6142 -2.157 0.049
Coolant temp. 38.500 1.2949 29.732 0.000
Oil temp. 42.500 1.2949 32.821 0.000
W.I. percent 46.833 1.2949 36.167 0.000
Flow rate (Turbulent) 17.500 1.2949 13.514 0.000
Coolant temp.*Coolant temp. 19.833 2.4226 8.187 0.000
Oil temp.*Oil temp. 18.333 2.4226 7.568 0.000
W.I. percent*W.I. percent 25.333 2.4226 10.457 0.000
Flow rate (Turbulent)* 7.833 2.4226 3.234 0.076
Flow rate (Turbulent)
Coolant temp.*Oil temp. 71.500 3.1719 22.542 0.000
Coolant temp.*W.I. percent 30.500 3.1719 9.616 0.000
Coolant temp.*Flow rate (Turbulent) -2.000 3.1719 -0.631 0.538
Oil temp.*W.I. percent 39.000 3.1719 12.296 0.000
Oil temp.*Flow rate (Turbulent) -2.500 3.1719 -0.788 0.444
W.I. percent*Flow rate (Turbulent) -0.500 3.1719 -0.158 0.877
S = 3.17186 PRESS = 931.728

Analysis of Variance for Oil A (4 Components, Turb.)

Blocks 1 46.8 46.8 46.8
Regression 14 44150.2 44150.2 3153.6
Linear 4 34728.7 34728.7 8682.2
Coolant temp. 1 8893.5 8893.5 8893.5
Oil temp. 1 10837.5 10837.5 10837.5
W.I. percent 1 13160.2 13160.2 13160.2
Flow rate (Turbulent) 1 1837.5 1837.5 1837.5
Square 4 1847.5 1847.5 461.9
Coolant temp.*Coolant temp. 1 358.4 674.3 674.3
Oil temp.*Oil temp. 1 357.8 576.2 576.2
W.I. percent*W.I. percent 1 1026.1 1100.2 1100.2
Flow rate (Turbulent)*Flow rate (Turbulent) 1 105.2 105.2 105.2
Interaction 6 7574.0 7574.0 1262.3
Coolant temp.*Oil temp. 1 5112.3 5112.3 5112.3
281
Coolant temp.*W.I. percent 1 930.3 930.3 930.3
Coolant temp.*Flow rate (Turbulent) 1 4.0 4.0 4.0
Oil temp.*W.I. percent 1 1521.0 1521.0 1521.0
Oil temp.*Flow rate (Turbulent) 1 6.3 6.3 6.3
W.I. percent*Flow rate (Turbulent) 1 0.3 0.3 0.3
Residual Error 14 140.9 140.9 10.1
Lack-of-Fit 10 125.6 125.6 12.6
Pure Error 4 15.3 15.3 3.8
Total 29 44337.9
Source F P
Blocks 4.65 0.049
Linear 862.98 0.000
Coolant temp. 883.98 0.000
Oil temp. 1077.21 0.000
W.I. percent 1308.07 0.000
Flow rate (Turbulent) 182.64 0.000
Square 45.91 0.000
Coolant temp.*Coolant temp. 67.03 0.000
Oil temp.*Oil temp. 57.27 0.000
W.I. percent*W.I. percent 109.36 0.000
Flow rate (Turbulent)*Flow rate (Turbulent) 10.46 0.006
Coolant temp.*Oil temp. 508.14 0.000
Coolant temp.*W.I. percent 92.46 0.000
Coolant temp.*Flow rate (Turbulent) 0.40 0.538
Oil temp.*W.I. percent 151.18 0.000
Oil temp.*Flow rate (Turbulent) 0.62 0.444
W.I. percent*Flow rate (Turbulent) 0.02 0.877
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Oil A (4 Components, Turb.) using data in

uncoded units
Term Coef
Constant 305055
Block -1.32500
Coolant temp. -695.006
Oil temp. -1260.24
W.I. percent -2036.01
Flow rate (Turbulent) 1198.04
Coolant temp.*Coolant temp. 0.198333
Oil temp.*Oil temp. 1.14583
W.I. percent*W.I. percent 6.33333
Flow rate (Turbulent)* 195.833
Coolant temp.*Oil temp. 1.78750
Coolant temp.*W.I. percent 1.52500
Coolant temp.*Flow rate (Turbulent) -1.00000
Oil temp.*W.I. percent 4.87500
Oil temp.*Flow rate (Turbulent) -3.12500
W.I. percent*Flow rate (Turbulent) -1.25000
Predicted Response for New Design Points Using Model for Oil A (4 Components,
Turb.)
282
1 124.608 2.43119 (119.394, 129.823) (116.037, 133.180)
2 114.108 1.31099 (111.297, 116.920) (106.747, 121.469)
3 99.775 2.43119 ( 94.561, 104.989) ( 91.204, 108.346)
4 113.275 2.43119 (108.061, 118.489) (104.704, 121.846)
5 99.275 2.43119 ( 94.061, 104.489) ( 90.704, 107.846)
6 93.275 2.43119 ( 88.061, 98.489) ( 84.704, 101.846)
7 114.108 1.31099 (111.297, 116.920) (106.747, 121.469)
8 114.108 1.31099 (111.297, 116.920) (106.747, 121.469)
9 150.108 2.43119 (144.894, 155.323) (141.537, 158.680)
10 223.608 2.43119 (218.394, 228.823) (215.037, 232.180)
11 81.775 2.43119 ( 76.561, 86.989) ( 73.204, 90.346)
12 81.775 2.43119 ( 76.561, 86.989) ( 73.204, 90.346)
13 141.775 2.43119 (136.561, 146.989) (133.204, 150.346)
14 105.608 2.43119 (100.394, 110.823) ( 97.037, 114.180)
15 142.108 2.43119 (136.894, 147.323) (133.537, 150.680)
16 114.108 1.31099 (111.297, 116.920) (106.747, 121.469)
17 125.108 2.43119 (119.894, 130.323) (116.537, 133.680)
18 133.108 2.43119 (127.894, 138.323) (124.537, 141.680)
19 157.275 2.43119 (152.061, 162.489) (148.704, 165.846)
20 238.608 2.43119 (233.394, 243.823) (230.037, 247.180)
21 114.108 1.31099 (111.297, 116.920) (106.747, 121.469)
22 139.442 2.43119 (134.227, 144.656) (130.870, 148.013)
23 89.942 2.43119 ( 84.727, 95.156) ( 81.370, 98.513)
24 104.442 2.43119 ( 99.227, 109.656) ( 95.870, 113.013)
25 92.608 2.43119 ( 87.394, 97.823) ( 84.037, 101.180)
26 95.442 2.43119 ( 90.227, 100.656) ( 86.870, 104.013)
27 114.108 1.31099 (111.297, 116.920) (106.747, 121.469)
28 174.942 2.43119 (169.727, 180.156) (166.370, 183.513)
29 172.442 2.43119 (167.227, 177.656) (163.870, 181.013)
30 186.275 2.43119 (181.061, 191.489) (177.704, 194.846)
Response Surface Regression: Oil B (4 Com versus Block, Coolant temp, Oil temp.
Estimated Regression Coefficients for Oil B (4 Components, Turb.)

Constant 104.183 1.4449 72.103 0.000
Block -0.550 0.6770 -0.812 0.430
Coolant temp. 45.667 1.4272 31.997 0.000
Oil temp. 36.833 1.4272 25.808 0.000
W.I. percent 50.167 1.4272 35.150 0.000
Oil temp.*Oil temp. 29.083 2.6700 10.892 0.000
W.I. percent*Flow rate (Turbulent) 1.500 3.4959 0.429 0.674
S = 3.49592 PRESS = 988.888

283
Analysis of Variance for Oil B (4 Components, Turb.)

Blocks 1 8.1 8.1 8.1
Regression 14 58693.0 58693.0 4192.4
Linear 4 38569.7 38569.7 9642.4
Coolant temp. 1 12512.7 12512.7 12512.7
Oil temp. 1 8140.2 8140.2 8140.2
W.I. percent 1 15100.2 15100.2 15100.2
Square 4 5694.6 5694.6 1423.6
Oil temp.*Oil temp. 1 788.7 1450.0 1450.0
Interaction 6 14428.8 14428.8 2404.8
Residual Error 14 171.1 171.1 12.2
Lack-of-Fit 10 143.1 143.1 14.3
Pure Error 4 28.0 28.0 7.0
Total 29 58872.2
Source F P
Blocks 0.66 0.430
Linear 788.98 0.000
Oil temp. 666.06 0.000
W.I. percent 1235.55 0.000
Square 116.49 0.000
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Oil B (4 Components, Turb.) using data in

uncoded units
Term Coef
Constant 464951
Block -0.550000
Oil temp. -1924.01
284
W.I. percent*Flow rate (Turbulent) 3.75000
Predicted Response for New Design Points Using Model for Oil B (4 Components,
Turb.)

1 128.683 2.67957 (122.936, 134.430) (119.236, 138.131)
2 104.183 1.44493 (101.084, 107.282) ( 96.070, 112.297)
3 101.683 2.67957 ( 95.936, 107.430) ( 92.236, 111.131)
4 130.267 2.67957 (124.520, 136.014) (120.819, 139.714)
5 91.350 2.67957 ( 85.603, 97.097) ( 81.903, 100.797)
6 105.850 2.67957 (100.103, 111.597) ( 96.403, 115.297)
7 104.183 1.44493 (101.084, 107.282) ( 96.070, 112.297)
8 104.183 1.44493 (101.084, 107.282) ( 96.070, 112.297)
9 141.767 2.67957 (136.020, 147.514) (132.319, 151.214)
10 241.267 2.67957 (235.520, 247.014) (231.819, 250.714)
11 78.767 2.67957 ( 73.020, 84.514) ( 69.319, 88.214)
12 72.433 2.67957 ( 66.686, 78.180) ( 62.986, 81.881)
13 148.350 2.67957 (142.603, 154.097) (138.903, 157.797)
14 111.767 2.67957 (106.020, 117.514) (102.319, 121.214)
15 153.100 2.67957 (147.353, 158.847) (143.653, 162.547)
16 104.183 1.44493 (101.084, 107.282) ( 96.070, 112.297)
17 137.683 2.67957 (131.936, 143.430) (128.236, 147.131)
18 121.350 2.67957 (115.603, 127.097) (111.903, 130.797)
19 165.767 2.67957 (160.020, 171.514) (156.319, 175.214)
20 260.183 2.67957 (254.436, 265.930) (250.736, 269.631)
21 104.183 1.44493 (101.084, 107.282) ( 96.070, 112.297)
22 146.933 2.67957 (141.186, 152.680) (137.486, 156.381)
23 96.433 2.67957 ( 90.686, 102.180) ( 86.986, 105.881)
24 103.600 2.67957 ( 97.853, 109.347) ( 94.153, 113.047)
25 97.100 2.67957 ( 91.353, 102.847) ( 87.653, 106.547)
26 96.100 2.67957 ( 90.353, 101.847) ( 86.653, 105.547)
27 104.183 1.44493 (101.084, 107.282) ( 96.070, 112.297)
28 170.100 2.67957 (164.353, 175.847) (160.653, 179.547)
29 187.433 2.67957 (181.686, 193.180) (177.986, 196.881)
30 197.433 2.67957 (191.686, 203.180) (187.986, 206.881)
Response Surface Regression: Oil C (4 Com versus Block, Coolant temp, Oil temp.
Estimated Regression Coefficients for Oil C (4 Components, Turb.)

Constant 119.442 1.8952 63.023 0.000
Block 0.675 0.8879 0.760 0.460
Coolant temp. 45.667 1.8719 24.395 0.000
Oil temp. 36.833 1.8719 19.677 0.000
285
W.I. percent 43.167 1.8719 23.060 0.000
Oil temp.*Oil temp. 33.667 3.5021 9.613 0.000
S = 4.58530 PRESS = 1742.52

Analysis of Variance for Oil C (4 Components, Turb.)

Blocks 1 12.1 12.2 12.2
Regression 14 51629.0 51629.0 3687.8
Linear 4 33243.7 33243.7 8310.9
Coolant temp. 1 12512.7 12512.7 12512.7
Oil temp. 1 8140.2 8140.2 8140.2
W.I. percent 1 11180.2 11180.2 11180.2
Square 4 4332.6 4332.6 1083.1
Oil temp.*Oil temp. 1 1242.6 1943.0 1943.0
Interaction 6 14052.8 14052.8 2342.1
Residual Error 14 294.4 294.4 21.0
Lack-of-Fit 10 233.1 233.1 23.3
Pure Error 4 61.3 61.3 15.3
Total 29 51935.5
Source F P
Blocks 0.58 0.460
Linear 395.29 0.000
Oil temp. 387.17 0.000
W.I. percent 531.76 0.000
Square 51.52 0.000
286
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Oil C (4 Components, Turb.) using data in

uncoded units
Term Coef
Constant 480255
Block 0.675000
Oil temp. -2089.75
Predicted Response for New Design Points Using Model for Oil C (4 Components,
Turb.)

1 142.692 3.51457 (135.154, 150.230) (130.301, 155.083)
2 119.442 1.89520 (115.377, 123.506) (108.800, 130.083)
3 115.358 3.51457 (107.820, 122.896) (102.967, 127.749)
4 144.192 3.51457 (136.654, 151.730) (131.801, 156.583)
5 106.025 3.51457 ( 98.487, 113.563) ( 93.634, 118.416)
6 123.358 3.51457 (115.820, 130.896) (110.967, 135.749)
7 119.442 1.89520 (115.377, 123.506) (108.800, 130.083)
8 119.442 1.89520 (115.377, 123.506) (108.800, 130.083)
9 148.692 3.51457 (141.154, 156.230) (136.301, 161.083)
10 254.525 3.51457 (246.987, 262.063) (242.134, 266.916)
11 99.025 3.51457 ( 91.487, 106.563) ( 86.634, 111.416)
12 90.192 3.51457 ( 82.654, 97.730) ( 77.801, 102.583)
13 165.858 3.51457 (158.320, 173.396) (153.467, 178.249)
14 122.525 3.51457 (114.987, 130.063) (110.134, 134.916)
15 157.525 3.51457 (149.987, 165.063) (145.134, 169.916)
16 119.442 1.89520 (115.377, 123.506) (108.800, 130.083)
17 141.858 3.51457 (134.320, 149.396) (129.467, 154.249)
18 134.358 3.51457 (126.820, 141.896) (121.967, 146.749)
19 177.192 3.51457 (169.654, 184.730) (164.801, 189.583)
20 264.358 3.51457 (256.820, 271.896) (251.967, 276.749)
21 119.442 1.89520 (115.377, 123.506) (108.800, 130.083)
22 166.442 3.51457 (158.904, 173.980) (154.051, 178.833)
23 116.275 3.51457 (108.737, 123.813) (103.884, 128.666)
24 135.775 3.51457 (128.237, 143.313) (123.384, 148.166)
25 102.442 3.51457 ( 94.904, 109.980) ( 90.051, 114.833)
287
26 101.442 3.51457 ( 93.904, 108.980) ( 89.051, 113.833)
27 119.442 1.89520 (115.377, 123.506) (108.800, 130.083)
28 189.942 3.51457 (182.404, 197.480) (177.551, 202.333)
29 192.775 3.51457 (185.237, 200.313) (180.384, 205.166)
30 188.775 3.51457 (181.237, 196.313) (176.384, 201.166)
Response Surface Regression: Oil D (4 Com versus Block, Coolant temp, Oil temp.
Estimated Regression Coefficients for Oil D (4 Components, Turb.)

Constant 97.6000 2.279 42.828 0.000
Block -0.3000 1.068 -0.281 0.783
Coolant temp. 40.5000 2.251 17.993 0.000
Oil temp. 46.8333 2.251 20.806 0.000
W.I. percent 53.8333 2.251 23.916 0.000
Oil temp.*Oil temp. 33.7500 4.211 8.015 0.000
Coolant temp.*Flow rate (Turbulent) 9.0000 5.514 1.632 0.125
S = 5.51362 PRESS = 2073.56

Analysis of Variance for Oil D (4 Components, Turb.)

Blocks 1 2.4 2.4 2.4
Regression 14 60464.8 60464.8 4318.9
Linear 4 44190.0 44190.0 11047.5
Coolant temp. 1 9841.5 9841.5 9841.5
Oil temp. 1 13160.2 13160.2 13160.2
W.I. percent 1 17388.2 17388.2 17388.2
Square 4 5519.1 5519.1 1379.8
Oil temp.*Oil temp. 1 1150.3 1952.7 1952.7
Interaction 6 10755.8 10755.8 1792.6
Residual Error 14 425.6 425.6 30.4
Lack-of-Fit 10 363.1 363.1 36.3
Pure Error 4 62.5 62.5 15.6
288
Total 29 60892.8
Source F P
Blocks 0.08 0.783
Linear 363.40 0.000
Oil temp. 432.90 0.000
W.I. percent 571.98 0.000
Square 45.39 0.000
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Oil D (4 Components, Turb.) using data in

uncoded units
Term Coef
Constant 454042
Block -0.300000
Oil temp. -1999.07
Flow rate (Turbulent) -350.042
Coolant temp.*Flow rate (Turbulent) 4.50000
W.I. percent*Flow rate (Turbulent) -3.99597E-14
Predicted Response for New Design Points Using Model for Oil D (4 Components,
Turb.)

1 117.267 4.22611 (108.203, 126.331) (102.367, 132.166)
2 97.600 2.27889 ( 92.712, 102.488) ( 84.804, 110.396)
3 93.600 4.22611 ( 84.536, 102.664) ( 78.700, 108.500)
4 112.350 4.22611 (103.286, 121.414) ( 97.450, 127.250)
5 94.933 4.22611 ( 85.869, 103.997) ( 80.034, 109.833)
6 89.933 4.22611 ( 80.869, 98.997) ( 75.034, 104.833)
7 97.600 2.27889 ( 92.712, 102.488) ( 84.804, 110.396)
289
8 97.600 2.27889 ( 92.712, 102.488) ( 84.804, 110.396)
9 147.517 4.22611 (138.453, 156.581) (132.617, 162.416)
10 228.350 4.22611 (219.286, 237.414) (213.450, 243.250)
11 62.183 4.22611 ( 53.119, 71.247) ( 47.284, 77.083)
12 76.017 4.22611 ( 66.953, 85.081) ( 61.117, 90.916)
13 135.767 4.22611 (126.703, 144.831) (120.867, 150.666)
14 105.017 4.22611 ( 95.953, 114.081) ( 90.117, 119.916)
15 148.683 4.22611 (139.619, 157.747) (133.784, 163.583)
16 97.600 2.27889 ( 92.712, 102.488) ( 84.804, 110.396)
17 127.433 4.22611 (118.369, 136.497) (112.534, 142.333)
18 128.600 4.22611 (119.536, 137.664) (113.700, 143.500)
19 163.683 4.22611 (154.619, 172.747) (148.784, 178.583)
20 256.267 4.22611 (247.203, 265.331) (241.367, 271.166)
21 97.600 2.27889 ( 92.712, 102.488) ( 84.804, 110.396)
22 147.017 4.22611 (137.953, 156.081) (132.117, 161.916)
23 84.517 4.22611 ( 75.453, 93.581) ( 69.617, 99.416)
24 96.683 4.22611 ( 87.619, 105.747) ( 81.784, 111.583)
25 86.017 4.22611 ( 76.953, 95.081) ( 71.117, 100.916)
26 88.350 4.22611 ( 79.286, 97.414) ( 73.450, 103.250)
27 97.600 2.27889 ( 92.712, 102.488) ( 84.804, 110.396)
28 178.183 4.22611 (169.119, 187.247) (163.284, 193.083)
29 169.350 4.22611 (160.286, 178.414) (154.450, 184.250)
30 193.683 4.22611 (184.619, 202.747) (178.784, 208.583)
Response Surface Regression: Oil E (4 Co versus Block, Coolant temp, Oil temp.
Estimated Regression Coefficients for Oil E (4 Components, Turb.)

Constant 76.7083 1.4998 51.147 0.000
Block 0.3750 0.7027 0.534 0.602
Coolant temp. 25.1667 1.4814 16.989 0.000
Oil temp. 27.0000 1.4814 18.226 0.000
W.I. percent 42.8333 1.4814 28.915 0.000
Oil temp.*Oil temp. 46.8333 2.7714 16.899 0.000
S = 3.62859 PRESS = 1020.24

Analysis of Variance for Oil E (4 Components, Turb.)

Blocks 1 3.7 3.8 3.8
Regression 14 28839.9 28839.9 2060.0
Linear 4 20443.8 20443.8 5111.0
Coolant temp. 1 3800.2 3800.2 3800.2
290
Oil temp. 1 4374.0 4374.0 4374.0
W.I. percent 1 11008.2 11008.2 11008.2
Square 4 5683.6 5683.6 1420.9
Oil temp.*Oil temp. 1 2712.7 3760.0 3760.0
Interaction 6 2712.5 2712.5 452.1
Residual Error 14 184.3 184.3 13.2
Lack-of-Fit 10 151.6 151.6 15.2
Pure Error 4 32.8 32.8 8.2
Total 29 29028.0
Source F P
Blocks 0.28 0.602
Linear 388.17 0.000
Oil temp. 332.20 0.000
W.I. percent 836.06 0.000
Square 107.92 0.000
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Oil E (4 Components, Turb.) using data

in uncoded units
Term Coef
Constant 425763
Block 0.375000
Oil temp. -2229.47
291
Oil temp.*Flow rate (Turbulent) -4.78036E-15
Predicted Response for New Design Points Using Model for Oil E (4 Components,
Turb.)

1 112.167 2.78126 (106.201, 118.132) (102.361, 121.972)
2 76.708 1.49977 ( 73.492, 79.925) ( 68.287, 85.129)
3 78.833 2.78126 ( 72.868, 84.799) ( 69.028, 88.639)
4 86.417 2.78126 ( 80.451, 92.382) ( 76.611, 96.222)
5 85.667 2.78126 ( 79.701, 91.632) ( 75.861, 95.472)
6 71.167 2.78126 ( 65.201, 77.132) ( 61.361, 80.972)
7 76.708 1.49977 ( 73.492, 79.925) ( 68.287, 85.129)
8 76.708 1.49977 ( 73.492, 79.925) ( 68.287, 85.129)
9 120.250 2.78126 (114.285, 126.215) (110.444, 130.056)
10 166.917 2.78126 (160.951, 172.882) (157.111, 176.722)
11 61.083 2.78126 ( 55.118, 67.049) ( 51.278, 70.889)
12 70.417 2.78126 ( 64.451, 76.382) ( 60.611, 80.222)
13 106.333 2.78126 (100.368, 112.299) ( 96.528, 116.139)
14 100.250 2.78126 ( 94.285, 106.215) ( 90.444, 110.056)
15 125.917 2.78126 (119.951, 131.882) (116.111, 135.722)
16 76.708 1.49977 ( 73.492, 79.925) ( 68.287, 85.129)
17 111.500 2.78126 (105.535, 117.465) (101.694, 121.306)
18 109.000 2.78126 (103.035, 114.965) ( 99.194, 118.806)
19 124.083 2.78126 (118.118, 130.049) (114.278, 133.889)
20 180.667 2.78126 (174.701, 186.632) (170.861, 190.472)
21 76.708 1.49977 ( 73.492, 79.925) ( 68.287, 85.129)
22 118.042 2.78126 (112.076, 124.007) (108.236, 127.847)
23 96.542 2.78126 ( 90.576, 102.507) ( 86.736, 106.347)
24 89.042 2.78126 ( 83.076, 95.007) ( 79.236, 98.847)
25 69.208 2.78126 ( 63.243, 75.174) ( 59.403, 79.014)
26 63.375 2.78126 ( 57.410, 69.340) ( 53.569, 73.181)
27 76.708 1.49977 ( 73.492, 79.925) ( 68.287, 85.129)
28 150.542 2.78126 (144.576, 156.507) (140.736, 160.347)
29 113.708 2.78126 (107.743, 119.674) (103.903, 123.514)
30 154.875 2.78126 (148.910, 160.840) (145.069, 164.681)
Response Surface Regression: Oil F (4 Com versus Block, Coolant temp, Oil temp.
Estimated Regression Coefficients for Oil F (4 Components, Turb.)

Constant 91.608 1.8216 50.290 0.000
Block -1.825 0.8535 -2.138 0.051
Coolant temp. 26.083 1.7993 14.497 0.000
Oil temp. 31.750 1.7993 17.646 0.000
W.I. percent 34.083 1.7993 18.943 0.000
Oil temp.*Oil temp. 38.958 3.3661 11.574 0.000
292
Oil temp.*Flow rate (Turbulent) 8.250 4.4072 1.872 0.082
W.I. percent*Flow rate (Turbulent) 2.750 4.4072 0.624 0.543
S = 4.40725 PRESS = 1726.95

Analysis of Variance for Oil F (4 Components, Turb.)

Blocks 1 88.8 88.8 88.82
Regression 14 28313.4 28313.4 2022.39
Linear 4 17793.8 17793.8 4448.46
Coolant temp. 1 4082.0 4082.0 4082.04
Oil temp. 1 6048.4 6048.4 6048.38
W.I. percent 1 6970.0 6970.0 6970.04
Square 4 5393.7 5393.7 1348.43
Oil temp.*Oil temp. 1 1788.6 2601.9 2601.86
Interaction 6 5125.9 5125.9 854.31
Residual Error 14 271.9 271.9 19.42
Lack-of-Fit 10 225.9 225.9 22.59
Pure Error 4 46.0 46.0 11.50
Total 29 28674.2
Source F P
Blocks 4.57 0.051
Linear 229.02 0.000
Oil temp. 311.39 0.000
W.I. percent 358.84 0.000
Square 69.42 0.000
Residual Error
Pure Error
Total
293
Estimated Regression Coefficients for Oil F (4 Components, Turb.) using data in
uncoded units
Term Coef
Constant 404023
Block -1.82500
Oil temp. -1944.40
Oil temp.*Flow rate (Turbulent) 10.3125
W.I. percent*Flow rate (Turbulent) 6.87500
Predicted Response for New Design Points Using Model for Oil F (4 Components,
Turb.)

1 119.525 3.37809 (112.280, 126.770) (107.615, 131.435)
2 91.608 1.82160 ( 87.701, 95.515) ( 81.380, 101.837)
3 97.692 3.37809 ( 90.446, 104.937) ( 85.782, 109.602)
4 109.983 3.37809 (102.738, 117.229) ( 98.073, 121.893)
5 101.858 3.37809 ( 94.613, 109.104) ( 89.948, 113.768)
6 107.358 3.37809 (100.113, 114.604) ( 95.448, 119.268)
7 91.608 1.82160 ( 87.701, 95.515) ( 81.380, 101.837)
8 91.608 1.82160 ( 87.701, 95.515) ( 81.380, 101.837)
9 140.317 3.37809 (133.071, 147.562) (128.407, 152.227)
10 191.150 3.37809 (183.905, 198.395) (179.240, 203.060)
11 89.817 3.37809 ( 82.571, 97.062) ( 77.907, 101.727)
12 90.983 3.37809 ( 83.738, 98.229) ( 79.073, 102.893)
13 122.192 3.37809 (114.946, 129.437) (110.282, 134.102)
14 103.817 3.37809 ( 96.571, 111.062) ( 91.907, 115.727)
15 130.150 3.37809 (122.905, 137.395) (118.240, 142.060)
16 91.608 1.82160 ( 87.701, 95.515) ( 81.380, 101.837)
17 109.192 3.37809 (101.946, 116.437) ( 97.282, 121.102)
18 126.692 3.37809 (119.446, 133.937) (114.782, 138.602)
19 138.317 3.37809 (131.071, 145.562) (126.407, 150.227)
20 210.025 3.37809 (202.780, 217.270) (198.115, 221.935)
21 91.608 1.82160 ( 87.701, 95.515) ( 81.380, 101.837)
22 125.817 3.37809 (118.571, 133.062) (113.907, 137.727)
23 98.817 3.37809 ( 91.571, 106.062) ( 86.907, 110.727)
24 104.317 3.37809 ( 97.071, 111.562) ( 92.407, 116.227)
25 91.983 3.37809 ( 84.738, 99.229) ( 80.073, 103.893)
26 99.483 3.37809 ( 92.238, 106.729) ( 87.573, 111.393)
27 91.608 1.82160 ( 87.701, 95.515) ( 81.380, 101.837)
28 162.317 3.37809 (155.071, 169.562) (150.407, 174.227)
29 151.650 3.37809 (144.405, 158.895) (139.740, 163.560)
30 160.150 3.37809 (152.905, 167.395) (148.240, 172.060)
Response Surface Regression: Oil G, (4 Co versus Block, Coolant temp, Oil temp.
294
Estimated Regression Coefficients for Oil G, (4 Components, Turb.)

Constant 105.775 2.540 41.642 0.000
Block -0.825 1.190 -0.693 0.500
Coolant temp. 36.000 2.509 14.349 0.000
Oil temp. 36.667 2.509 14.614 0.000
W.I. percent 51.000 2.509 20.327 0.000
Oil temp.*Oil temp. 21.250 4.694 4.527 0.000
S = 6.14565 PRESS = 2984.37

Analysis of Variance for Oil G, (4 Components, Turb.)

Blocks 1 18.2 18.1 18.1
Regression 14 44195.9 44195.9 3156.8
Linear 4 34576.8 34576.8 8644.2
Coolant temp. 1 7776.0 7776.0 7776.0
Oil temp. 1 8066.7 8066.7 8066.7
W.I. percent 1 15606.0 15606.0 15606.0
Square 4 2642.0 2642.0 660.5
Oil temp.*Oil temp. 1 400.0 774.1 774.1
Interaction 6 6977.0 6977.0 1162.8
Residual Error 14 528.8 528.8 37.8
Lack-of-Fit 10 478.0 478.0 47.8
Pure Error 4 50.8 50.8 12.7
Total 29 44742.8
Source F P
Blocks 0.48 0.500
Linear 228.87 0.000
Oil temp. 213.58 0.000
295
W.I. percent 413.20 0.000
Square 17.49 0.000
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Oil G, (4 Components, Turb.) using data

in uncoded units
Term Coef
Constant 314206
Block -0.825000
Oil temp. -1350.76
Predicted Response for New Design Points Using Model for Oil G, (4 Components,
Turb.)

1 123.900 4.71055 (113.797, 134.003) (107.292, 140.508)
2 105.775 2.54012 (100.327, 111.223) ( 91.512, 120.038)
3 98.233 4.71055 ( 88.130, 108.336) ( 81.626, 114.841)
4 110.733 4.71055 (100.630, 120.836) ( 94.126, 127.341)
5 96.400 4.71055 ( 86.297, 106.503) ( 79.792, 113.008)
6 88.900 4.71055 ( 78.797, 99.003) ( 72.292, 105.508)
7 105.775 2.54012 (100.327, 111.223) ( 91.512, 120.038)
8 105.775 2.54012 (100.327, 111.223) ( 91.512, 120.038)
9 142.900 4.71055 (132.797, 153.003) (126.292, 159.508)
10 211.567 4.71055 (201.464, 221.670) (194.959, 228.174)
11 68.400 4.71055 ( 58.297, 78.503) ( 51.792, 85.008)
12 78.567 4.71055 ( 68.464, 88.670) ( 61.959, 95.174)
13 126.067 4.71055 (115.964, 136.170) (109.459, 142.674)
14 107.400 4.71055 ( 97.297, 117.503) ( 90.792, 124.008)
15 151.733 4.71055 (141.630, 161.836) (135.126, 168.341)
296
16 105.775 2.54012 (100.327, 111.223) ( 91.512, 120.038)
17 127.233 4.71055 (117.130, 137.336) (110.626, 143.841)
18 127.067 4.71055 (116.964, 137.170) (110.459, 143.674)
19 151.900 4.71055 (141.797, 162.003) (135.292, 168.508)
20 235.400 4.71055 (225.297, 245.503) (218.792, 252.008)
21 105.775 2.54012 (100.327, 111.223) ( 91.512, 120.038)
22 146.358 4.71055 (136.255, 156.461) (129.751, 162.966)
23 90.358 4.71055 ( 80.255, 100.461) ( 73.751, 106.966)
24 100.692 4.71055 ( 90.589, 110.795) ( 84.084, 117.299)
25 87.025 4.71055 ( 76.922, 97.128) ( 70.417, 103.633)
26 87.025 4.71055 ( 76.922, 97.128) ( 70.417, 103.633)
27 105.775 2.54012 (100.327, 111.223) ( 91.512, 120.038)
28 163.692 4.71055 (153.589, 173.795) (147.084, 180.299)
29 159.025 4.71055 (148.922, 169.128) (142.417, 175.633)
30 189.025 4.71055 (178.922, 199.128) (172.417, 205.633)
FOR CRUDE OILS FLOW IN SUBSEA STRAIGHT PIPELINE (FIVE (5) VARIABLES,
LAMINAR FLOW REGIME)
————— 4/11/2015 7:26:52 PM ————————————————————
Response Surface Regression: Wax deposit versus Block, Coolant temp, Bulk oil
Estimated Regression Coefficients for Wax deposit (cm3)

Constant 87.774 13.303 6.598 0.000
Block -20.087 9.791 -2.052 0.048
Coolant temp. (K) -30.500 7.888 -3.867 0.001
Bulk oil temp. (K) -47.412 8.556 -5.542 0.000
Wax inhib. Perc. (%) -22.850 7.888 -2.897 0.007
Flow rate (Lit./min.) 13.800 7.888 1.749 0.090
Wax content (%) 39.534 8.515 4.643 0.000
Coolant temp. (K)*Coolant temp. (K) 25.175 9.072 2.775 0.009
Bulk oil temp. (K)* 47.384 22.309 2.124 0.041
Bulk oil temp. (K)
Wax inhib. Perc. (%)* 34.800 9.072 3.836 0.001
Flow rate (Lit./min.)* 40.925 9.072 4.511 0.000
Wax content (%)*Wax content (%) 35.701 12.319 2.898 0.007
Coolant temp. (K)*Bulk oil temp. (K) 53.313 8.819 6.045 0.000
Coolant temp. (K)* -45.687 8.819 -5.181 0.000
Coolant temp. (K)* -11.312 8.819 -1.283 0.209
Coolant temp. (K)*Wax content (%) -2.813 8.819 -0.319 0.752
Bulk oil temp. (K)* 36.563 8.819 4.146 0.000
Bulk oil temp. (K)* 11.438 8.819 1.297 0.204
Bulk oil temp. (K)*Wax content (%) 29.687 8.819 3.366 0.002
297
Wax inhib. Perc. (%)* -11.438 8.819 -1.297 0.204
Wax inhib. Perc. (%)*Wax content (%) 11.563 8.819 1.311 0.199
Flow rate (Lit./min.)* 23.437 8.819 2.658 0.012
Wax content (%)
S = 49.8881 PRESS = 282664

Analysis of Variance for Wax deposit (cm3)

Blocks 1 11062 10476 10476.4
Regression 20 721550 721550 36077.5
Linear 5 187009 195793 39158.5
Coolant temp. (K) 1 37210 37210 37210.0
Bulk oil temp. (K) 1 76428 76428 76427.8
Wax inhib. Perc. (%) 1 20885 20885 20884.9
Flow rate (Lit./min.) 1 7618 7618 7617.6
Wax content (%) 1 44868 53652 53652.3
Square 5 271238 271238 54247.5
Coolant temp. (K)*Coolant temp. (K) 1 55365 19167 19166.9
Bulk oil temp. (K)*Bulk oil temp. (K) 1 112480 11228 11227.9
Wax inhib. Perc. (%)*Wax inhib. Perc. (%) 1 32013 36624 36624.5
Flow rate (Lit./min.)*Flow rate (Lit./min.) 1 50479 50651 50651.3
Wax content (%)*Wax content (%) 1 20902 20902 20901.7
Interaction 10 263304 263304 26330.4
Coolant temp. (K)*Bulk oil temp. (K) 1 90951 90951 90951.1
Coolant temp. (K)*Wax inhib. Perc. (%) 1 66795 66795 66795.1
Coolant temp. (K)*Flow rate (Lit./min.) 1 4095 4095 4095.1
Coolant temp. (K)*Wax content (%) 1 253 253 253.1
Bulk oil temp. (K)*Wax inhib. Perc. (%) 1 42778 42778 42778.1
Bulk oil temp. (K)*Flow rate (Lit./min.) 1 4186 4186 4186.1
Bulk oil temp. (K)*Wax content (%) 1 28203 28203 28203.1
Wax inhib. Perc. (%)*Flow rate (Lit./min.) 1 4186 4186 4186.1
Wax inhib. Perc. (%)*Wax content (%) 1 4278 4278 4278.1
Flow rate (Lit./min.)*Wax content (%) 1 17578 17578 17578.1
Residual Error 32 79642 79642 2488.8
Lack-of-Fit 22 79600 79600 3618.2
Pure Error 10 43 43 4.3
Total 53 812255
Source F P
Blocks 4.21 0.048
Linear 15.73 0.000
Coolant temp. (K) 14.95 0.001
Bulk oil temp. (K) 30.71 0.000
Wax inhib. Perc. (%) 8.39 0.007
Flow rate (Lit./min.) 3.06 0.090
Wax content (%) 21.56 0.000
Square 21.80 0.000
Coolant temp. (K)*Coolant temp. (K) 7.70 0.009
Bulk oil temp. (K)*Bulk oil temp. (K) 4.51 0.041
Wax inhib. Perc. (%)*Wax inhib. Perc. (%) 14.72 0.001
Flow rate (Lit./min.)*Flow rate (Lit./min.) 20.35 0.000
Wax content (%)*Wax content (%) 8.40 0.007
Coolant temp. (K)*Bulk oil temp. (K) 36.54 0.000
Coolant temp. (K)*Wax inhib. Perc. (%) 26.84 0.000
298
Coolant temp. (K)*Flow rate (Lit./min.) 1.65 0.209
Coolant temp. (K)*Wax content (%) 0.10 0.752
Bulk oil temp. (K)*Wax inhib. Perc. (%) 17.19 0.000
Bulk oil temp. (K)*Flow rate (Lit./min.) 1.68 0.204
Bulk oil temp. (K)*Wax content (%) 11.33 0.002
Wax inhib. Perc. (%)*Flow rate (Lit./min.) 1.68 0.204
Wax inhib. Perc. (%)*Wax content (%) 1.72 0.199
Flow rate (Lit./min.)*Wax content (%) 7.06 0.012
Residual Error
Lack-of-Fit 848.84 1.000
Pure Error
Total
Estimated Regression Coefficients for Wax deposit (cm3) using data in uncoded
units
Term Coef
Constant 485226
Block -20.0875
Coolant temp. (K) -565.484
Bulk oil temp. (K) -2392.28
Wax inhib. Perc. (%) -891.809
Flow rate (Lit./min.) -5987.92
Wax content (%) -475.740
Coolant temp. (K)*Coolant temp. (K) 0.251750
Bulk oil temp. (K)* 2.96151
Bulk oil temp. (K)
Wax inhib. Perc. (%)* 8.70001
Flow rate (Lit./min.)* 1023.13
Wax content (%)*Wax content (%) 0.991708
Coolant temp. (K)*Bulk oil temp. (K) 1.33281
Coolant temp. (K)* -2.28437
Coolant temp. (K)* -5.65625
Coolant temp. (K)*Wax content (%) -0.0468750
Bulk oil temp. (K)*Wax content (%) 1.23698
Wax inhib. Perc. (%)* -28.5938
Wax inhib. Perc. (%)*Wax content (%) 0.963542
Flow rate (Lit./min.)* 19.5312
Wax content (%)
Predicted Response for New Design Points Using Model for Wax deposit (cm3)

1 182.570 25.4130 (130.805, 234.334) ( 68.526, 296.613)
2 151.511 46.5540 ( 56.684, 246.338) ( 12.520, 290.502)
3 87.774 13.3027 ( 60.677, 114.871) (-17.396, 192.943)
4 249.474 38.2587 (171.544, 327.404) (121.413, 377.535)
5 163.010 19.4488 (123.394, 202.626) ( 53.942, 272.078)
6 272.674 38.2587 (194.744, 350.604) (144.613, 400.735)
7 87.746 25.4130 ( 35.982, 139.511) (-26.297, 201.790)
8 87.774 13.3027 ( 60.677, 114.871) (-17.396, 192.943)
299
9 87.774 13.3027 ( 60.677, 114.871) (-17.396, 192.943)
10 279.074 38.2587 (201.144, 357.004) (151.013, 407.135)
11 127.474 38.2587 ( 49.544, 205.404) ( -0.587, 255.535)
12 223.874 38.2587 (145.944, 301.804) ( 95.813, 351.935)
13 181.274 38.2587 (103.344, 259.204) ( 53.213, 309.335)
14 87.774 13.3027 ( 60.677, 114.871) (-17.396, 192.943)
15 116.263 35.3512 ( 44.255, 188.272) ( -8.282, 240.809)
16 87.774 13.3027 ( 60.677, 114.871) (-17.396, 192.943)
17 285.182 35.5444 (212.781, 357.584) (160.409, 409.955)
18 87.774 13.3027 ( 60.677, 114.871) (-17.396, 192.943)
19 267.232 35.5444 (194.831, 339.634) (142.459, 392.005)
20 87.774 13.3027 ( 60.677, 114.871) (-17.396, 192.943)
21 393.082 35.5444 (320.681, 465.484) (268.309, 517.855)
22 386.237 35.3512 (314.229, 458.245) (261.692, 510.782)
23 387.806 35.5444 (315.404, 460.207) (263.032, 512.579)
24 67.837 35.3512 ( -4.171, 139.845) (-56.708, 192.382)
25 324.037 35.3512 (252.029, 396.045) (199.492, 448.582)
26 295.982 35.5444 (223.581, 368.384) (171.209, 420.755)
27 175.663 35.3512 (103.655, 247.672) ( 51.118, 300.209)
28 222.363 35.3512 (150.355, 294.372) ( 97.818, 346.909)
29 365.937 35.3512 (293.929, 437.945) (241.392, 490.482)
30 151.906 35.5444 ( 79.504, 224.307) ( 27.132, 276.679)
31 179.013 35.3512 (107.005, 251.022) ( 54.468, 303.559)
32 382.782 35.5444 (310.381, 455.184) (258.009, 507.555)
33 361.856 35.5444 (289.454, 434.257) (237.082, 486.629)
34 343.837 35.3512 (271.829, 415.845) (219.292, 468.382)
35 87.774 13.3027 ( 60.677, 114.871) (-17.396, 192.943)
36 87.774 13.3027 ( 60.677, 114.871) (-17.396, 192.943)
37 158.163 35.3512 ( 86.155, 230.172) ( 33.618, 282.709)
38 413.937 35.3512 (341.929, 485.945) (289.392, 538.482)
39 131.182 35.5444 ( 58.781, 203.584) ( 6.409, 255.955)
40 87.774 13.3027 ( 60.677, 114.871) (-17.396, 192.943)
41 87.774 13.3027 ( 60.677, 114.871) (-17.396, 192.943)
42 439.156 35.5444 (366.754, 511.557) (314.382, 563.929)
43 369.256 35.5444 (296.854, 441.657) (244.482, 494.029)
44 57.113 35.3512 (-14.895, 129.122) (-67.432, 181.659)
45 274.032 35.5444 (201.631, 346.434) (149.259, 398.805)
46 417.287 35.3512 (345.279, 489.295) (292.742, 541.832)
47 155.487 35.3512 ( 83.479, 227.495) ( 30.942, 280.032)
48 87.774 13.3027 ( 60.677, 114.871) (-17.396, 192.943)
49 319.082 35.5444 (246.681, 391.484) (194.309, 443.855)
50 145.806 35.5444 ( 73.404, 218.207) ( 21.032, 270.579)
51 106.213 35.3512 ( 34.205, 178.222) (-18.332, 230.759)
52 310.006 35.5444 (237.604, 382.407) (185.232, 434.779)
53 466.356 35.5444 (393.954, 538.757) (341.582, 591.129)
54 226.213 35.3512 (154.205, 298.222) (101.668, 350.759)
300
FOR CRUDE OILSFLOW IN SUB SEA STRAIGHT PIPELINE (FIVE (5) VARIABLES,
TURBULENT FLOW REGIME)
————— 5/16/2016 12:11:44 PM ————————————————————
Response Surface Regression: Wax Deposit, versus Block, Coolant Temp, Bulk Oil
Estimated Regression Coefficients for Wax Deposit, (cm3)_5

Constant 144.380 4.537 31.824 0.000
Block 1 2.947 2.973 0.991 0.329
Block 2 -1.453 2.973 -0.489 0.628
Coolant Temp, K 18.665 2.394 7.797 0.000
Bulk Oil Temp., K -3.282 2.394 -1.371 0.015
Flow Rate, Lit./min. -4.994 2.394 -2.086 0.045
Wax Inh. Per., % -66.260 2.394 -27.680 0.000
Wax Content, % 14.846 2.394 6.202 0.000
Coolant Temp, K*Coolant Temp, K 0.850 2.038 0.417 0.009
Bulk Oil Temp., K*Bulk Oil Temp., K 2.458 2.038 1.206 0.007
Flow Rate, Lit./min.* 1.475 2.038 0.724 0.474
Flow Rate, Lit./min.
Wax Inh. Per., %*Wax Inh. Per., % 0.403 2.038 0.198 0.844
Wax Content, %*Wax Content, % 2.279 2.038 1.119 0.272
Coolant Temp, K*Bulk Oil Temp., K 3.031 2.781 1.090 0.284
Coolant Temp, K*Flow Rate, Lit./min. -1.469 2.781 -0.528 0.601
Coolant Temp, K*Wax Inh. Per., % -2.469 2.781 -0.888 0.382
Coolant Temp, K*Wax Content, % 5.156 2.781 1.854 0.073
Bulk Oil Temp., K* 0.594 2.781 0.213 0.832
Bulk Oil Temp., K*Wax Inh. Per., % -3.656 2.781 -1.315 0.198
301
Bulk Oil Temp., K*Wax Content, % -3.406 2.781 -1.225 0.230
Flow Rate, Lit./min.* 1.344 2.781 0.483 0.632
Wax Inh. Per., %
Flow Rate, Lit./min.*Wax Content, % -4.906 2.781 -1.764 0.088
Wax Inh. Per., %*Wax Content, % -2.531 2.781 -0.910 0.370
S = 15.7331 PRESS = 31732.5

Analysis of Variance for Wax Deposit, (cm3)_5

Blocks 2 246 246 123
Regression 20 219764 219764 10988
Linear 5 215758 215758 43152
Coolant Temp, K 1 15049 15049 15049
Bulk Oil Temp., K 1 465 465 465
Flow Rate, Lit./min. 1 1077 1077 1077
Wax Inh. Per., % 1 189647 189647 189647
Wax Content, % 1 9520 9520 9520
Square 5 754 754 151
Coolant Temp, K*Coolant Temp, K 1 19 43 43
Bulk Oil Temp., K*Bulk Oil Temp., K 1 310 360 360
Flow Rate, Lit./min.*Flow Rate, Lit./min. 1 109 130 130
Wax Inh. Per., %*Wax Inh. Per., % 1 5 10 10
Wax Content, %*Wax Content, % 1 310 310 310
Interaction 10 3252 3252 325
Coolant Temp, K*Bulk Oil Temp., K 1 294 294 294
Coolant Temp, K*Flow Rate, Lit./min. 1 69 69 69
Coolant Temp, K*Wax Inh. Per., % 1 195 195 195
Coolant Temp, K*Wax Content, % 1 851 851 851
Bulk Oil Temp., K*Flow Rate, Lit./min. 1 11 11 11
Bulk Oil Temp., K*Wax Inh. Per., % 1 428 428 428
Bulk Oil Temp., K*Wax Content, % 1 371 371 371
Flow Rate, Lit./min.*Wax Inh. Per., % 1 58 58 58
Flow Rate, Lit./min.*Wax Content, % 1 770 770 770
Wax Inh. Per., %*Wax Content, % 1 205 205 205
Residual Error 31 7673 7673 248
Lack-of-Fit 22 6425 6425 292
Pure Error 9 1248 1248 139
Total 53 227683
Source F P
Blocks 0.50 0.614
Linear 174.33 0.000
Coolant Temp, K 60.80 0.000
Bulk Oil Temp., K 1.88 0.180
Flow Rate, Lit./min. 4.35 0.045
Wax Inh. Per., % 766.16 0.000
Wax Content, % 38.46 0.000
Square 0.61 0.694
Coolant Temp, K*Coolant Temp, K 0.17 0.679
Bulk Oil Temp., K*Bulk Oil Temp., K 1.46 0.237
Flow Rate, Lit./min.*Flow Rate, Lit./min. 0.52 0.474
Wax Inh. Per., %*Wax Inh. Per., % 0.04 0.844
Wax Content, %*Wax Content, % 1.25 0.272
Coolant Temp, K*Bulk Oil Temp., K 1.19 0.284
Coolant Temp, K*Flow Rate, Lit./min. 0.28 0.601
302
Coolant Temp, K*Wax Inh. Per., % 0.79 0.382
Coolant Temp, K*Wax Content, % 3.44 0.073
Bulk Oil Temp., K*Flow Rate, Lit./min. 0.05 0.832
Bulk Oil Temp., K*Wax Inh. Per., % 1.73 0.198
Bulk Oil Temp., K*Wax Content, % 1.50 0.230
Flow Rate, Lit./min.*Wax Inh. Per., % 0.23 0.632
Flow Rate, Lit./min.*Wax Content, % 3.11 0.088
Wax Inh. Per., %*Wax Content, % 0.83 0.370
Residual Error
Pure Error
Total
Estimated Regression Coefficients for Wax Deposit, (cm3)_5 using data in

uncoded units
Term Coef
Constant 22620.5
Block 1 2.94689
Block 2 -1.45311
Coolant Temp, K -29.3121
Bulk Oil Temp., K -117.112
Flow Rate, Lit./min. -15.0021
Wax Inh. Per., % 153.367
Wax Content, % 25.1705
Coolant Temp, K*Coolant Temp, K 0.00850068
Bulk Oil Temp., K*Bulk Oil Temp., K 0.153612
Flow Rate, Lit./min.* 36.8824
Wax Inh. Per., %*Wax Inh. Per., % 0.100869
Wax Content, %*Wax Content, % 0.0633100
Coolant Temp, K*Bulk Oil Temp., K 0.0757813
Coolant Temp, K*Flow Rate, Lit./min. -0.734375
Coolant Temp, K*Wax Inh. Per., % -0.123438
Coolant Temp, K*Wax Content, % 0.0859375
Bulk Oil Temp., K* 0.742187
Bulk Oil Temp., K*Wax Inh. Per., % -0.457031
Bulk Oil Temp., K*Wax Content, % -0.141927
Flow Rate, Lit./min.* 3.35938
Wax Inh. Per., %
Flow Rate, Lit./min.*Wax Content, % -4.08854
Wax Inh. Per., %*Wax Content, % -0.210938
Residual Plots for Wax Deposit, (cm3)_5
————— 6/9/2016 2:14:43 PM ————————————————————
Welcome to Minitab, press F1 for help.

Retrieving project from file: 'C:\USERS\USER\DOCUMENTS\SIJI 60.MPJ'
303
304

Improved Prediction

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Improved Prediction

Uploaded by

Copyright:

Available Formats

IMPROVED PREDICTION OF WAX DEPOSITION IN SUBSEA CRUDE-OIL

OLUSIJI AYOADE ADEYANJU,

B.Sc, M.Sc. Petroleum Engineering (University of Ibadan, Ibadan, Nigeria)

A THESIS SUBMITTED TO THE SCHOOL OF POSTGRADUATE STUDIES IN

PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY (Ph.D.), IN CHEMICAL ENGINEERING, AT THE

UNIVERSITY OF LAGOS, LAGOS, NIGERIA.

For the award of the degree of

………………………………….. ……………………… ………………

………………………………….. ……………………… ………………

………………………………….. ……………………… ………………

………………………………….. ……………………… ………………

………………………………….. ……………………… ………………

………………………………….. ……………………… ………………

………………………………….. ……………………… ………………

I dedicate this thesis to the Awesome and Almighty GOD

successfully finish the research.

Department, Igbinedion University, Okada, Edo State.

LIST OF FIGURES xviii

LIST OF TABLES xxvi

GLOSSARY OF TERMS xxix

LIST OF ABBREVIATIONS xxxvii

1.1: Background of the study 2

1.2: Statement of the Problem 3

1.3: Aim and objective of the Study 4

1.4 Scope of the Study 4

1.5: Significance/Justification of the Study 5

2.0: LITERATURE REVIEW 7

2.1: Wax Deposition Mechanisms 7

2.2 Effects of Wax Deposition 8

2.2.1 Technical Effect 8

2.2.2 Economic Effect 9

2.3: Wax Deposition Process 9

2.3.1: Factors Leading to Wax Precipitation and Deposition 11

2.3.1.1: Bulk Oil and Pipe Wall (Coolant) Temperatures 12

2.3.1.2: Flow rate 13

2.3.1.3: Pipe-wall Surface Roughness 14

2.3.1.4: Gas-Oil Ratio 14

2.3.1.5: Crude Oil Composition 14

2.3.2: Wax Crystallization Process 17

2.3.3: Wax Deposition Mechanisms 18

2.3.3.1: Molecular Diffusion Mechanism 19

2.3.3.2: Shear Dispersion Mechanism 19

2.3.4: Production Problems of Wax Deposition 21

2.3.4.1: High Viscosity and Pressure Loss of Crude Oil Flow 22

2.3.4.3: Deposition of Wax Crystals on Surfaces 23

2.3.5: Wax Treatment Techniques 23

2.3.6: Wax Removal Techniques 24

2.3.6.1: Mechanical Removal 24

2.3.6.2: Heat Application 25

2.3.6.3: Application of Wax Removal Chemicals 25

2.3.6.4: Hot Oiling Flow 25

2.3.6.5: Application of Wax Solvent (Inhibitors) 26

2.3.6.6: Application of Thermochemical Packages 27

2.3.7: Wax Preventive Techniques 27

2.3.7.1: Major Methods 27

2.3.7.2: Minor Methods 28

2.3.8: Crude Oil Wax Deposition in Subsea Environment 29

2.3.9: Gulf of Guinea Subsea Environment 31

2.4: Laboratory Studies of Wax Deposition 33

2.4.1: Cold Finger Experiments 34

2.4.2: Flow Loop Experiments 34

2.5: Chemical Wax Prevention 37

2.5.1: Wax Dispersants 38