International Journal of Chemical Engineering and Applications, Vol. 8, No.

1, February 2017

Simulation and Performance Analysis of Propylene-Propane

Splitter in Petroleum Refinery Case Study

Aniediong M. Umo and Etim N. Bassey

 the on-purpose propylene technology; propane

Abstract—The separation of propane and propylene mixture dehydrogenation, olefin metathesis, and methanol to
requires high capital cost and high energy consumption. This propylene [2]-[4]. However, irrespective of the propylene
challenge is obvious since propane and propylene have similar manufacturing route, the product is always a
molecular size and physical properties. Therefore in the last few
years, many processes have been surveyed as alternatives to
propylene/propane mixture which must be separated. The
distillation for this separation. However, researches carried out purity of the propylene will depend on the grade; for the
on these techniques indicated that they have not yet shown polymer grade a minimum of 99.5 % purity is required, for
sufficient advantage over the distillation process. Therefore the chemical grade 90-95 % purity and the refinery grade
objective of this work is to explore the possibility of establishing 50-70 % purity are required [3].
optimum operating parameters for better operation of the Since propane and propylene have similar molecular size
distillation process for realizing higher product quality and
improved energy consumption. In this paper the simulation and and physical properties, their separation is challenging. This
analysis of an existing propylene-propane distillation column separation is mostly carried out in a distillation column with
with 150 trays were carried out. The result of the response about 150-200 trays, with a reflux ratio as high as 10-20 and at
optimization showed a product yield of 95% propylene in the a high pressure of about 16-26 atm. The process requires high
distillate from a minimum of 94% yield, reduced reflux ratio capital cost and high energy consumption. To maximize profit,
from 10.2 to 10.02 and about 0.4 ×106 kcal/h of reboiler energy
as energy prices rise, energy efficiency, together with
input was saved. It was evident from these results that optimum
parameters had been established for better control of the splitter optimizing product quality and maximizing high–value
column case study with the potential of achieving higher product product yields have become more important to producers [5].
quality and improved energy consumption. However, for many years, the standard operating philosophy
for distillation process was to maximize reflux for maximum
Index Terms—Energy improvement, performance analysis, separation. When energy was reconsidered inexpensive, this
petroleum refinery, simulation, splitter. philosophy held sway [6]. Therefore in the last few years,
many processes like absorption, adsorption,
extractive-distillation, and membrane techniques have been
I. INTRODUCTION
surveyed as alternatives to distillation for this separation [3],
Propylene is one of the most important intermediate [7], [8]. However, researches carried out on these techniques
petrochemical products. Over the years, it has maintained a indicated that they have not yet shown sufficient advantage
remarkable growth in the market on account of increasing over the distillation process. For example in the use of
demand for polypropylene which absorbs more than 60% of extractive-distillation, the thermodynamic properties of the
all the propylene produced worldwide [1]. Other important investigated extraction solvents tend to exhibit insufficient
propylene derivatives include acrylic acid, acrylonitrile, selectivity to yield an economically viable process. Similarly,
cumene, phenol, gasoline blend, trimmers and tetramers for the use of adsorption method usually requires several steps to
detergents. Propylene is commonly produced as a by-product reach a desirable product purity in addition to the use of
of naphtha and light gas oil feedstock in the steam cracking complicated regeneration cycles, low olefin loadings and high
(SC) and the fluidized catalytic cracking (FCC) units. capital costs requirements thus making the system less
Currently SC accounts for about 60% while FCC accounts for attractive compared to distillation process [3], [9].
about 30% of propylene production [2]. Until better alternatives are developed or the present
The high cost of crude oil in the past few years made it techniques are improved to a level that shows sufficient
necessary for producers to seek cheaper feedstock for the advantage over distillation, effort must be made to optimize
refineries. The use of natural gas liquids (NGLs) and shale oil the distillation column to achieve the required result.
which have high composition of ethane oil have greatly Therefore the objective of this work is to explore the
reduced the by-product propylene. Thus to meet the possibility of establishing optimum operating parameters for
propylene market demand producers are gradually adopting better operation of the process for realizing higher product
quality and improved energy consumption.
Manuscript received June 20, 2016; revised October 17, 2016.
Aniediong M. Umo was with the Department of Chemical and Petroleum
Engineering, University of Uyo, Nigeria (e-mail: II. METHODOLOGY
[email protected]).
Etim N. Bassey is with the Department of Chemical/Petrochemical A. Problem Definition
Engineering, Akwa Ibom State University, Ikot Akpaden, Nigeria (e-mail:
[email protected]). The aim of this work is to study the interactive effect of

doi: 10.18178/ijcea.2017.8.1.621 1
 International Journal of Chemical Engineering and Applications, Vol. 8, No. 1, February 2017

those process variables involved in the operation of the The response surface methodology was used to obtain the
propane-propylene splitter in this case study. Sensitivity surface and contour plots from the simulated data to describe
analysis will provide a tool that is useful in the optimization of the interactive effect of the process variables on percentage
the operating condition of the splitter. propylene yield in the distillate. The column was optimized to
The propylene – propane splitter unit in this case study obtain a yield of 95% propylene and above in the distillate
which is conceptually illustrated in Fig. 1, is part of a which exceeds the 94% minimum specified for the
polypropylene plant which comprises propylene purification, propane-propylene splitter in the case study plant.
polymerization, additive and extrusion units. The plant was
A. Effect of Reflux Ratio and Column Pressure
designed to produce 35,000 metric tonnes of polypropylene
resin per year [10].The purification area upgrades 73mole The effect of reflux ratio and column pressure on the
percent propylene from a fluid catalytic cracking (FCC) unit percentage composition of propylene in the distillate is
to a minimum of 94moles percent in the propylene - propane presented in Fig. 2. It was observed that at a column pressure
splitter. of 22 atm and reflux ratio of 10, the percent of propylene
The propylene-propane splitter is made up of a column of obtained in the distillate was in the range of 94.0% to 94.4%.
3m diameter containing a total of 150 trays, a reboiler and an At a lower pressure of 18atm using the same reflux ratio of 10,
overhead condenser. The purified propylene recovered at the the percentage composition of propylene was observed to
top is condensed, part of which is returned as reflux and the increase to a higher range of 94.8% and 95.2%. Thus lowering
other part fed to the propylene drying unit, while the propane the column pressure of the splitter unit would favor the purity
is drawn off at the reboiler. of propylene in the distillate but would lead to an increased
reboiler duty. Thus it is advisable that, selecting the column
B. Process Simulation operating pressure must ensure that the dew point of the
The simulation of the propane-propylene splitting unit was distillate is above that which can easily be obtained with
carried out using Hysys Simulation tool and the result cooling liquid. In general the contour plots in Fig. 2, Fig. 3
obtained from the simulation was analyzed using Minitab 14 and Fig. 4 indicate that increase in the reflux ratio increases
statistical software. In this work, the number of theoretical the percentage propylene in the distillate while a decrease in
stages required for the separation, feed stages location, reflux pressure will favour high yield of propylene. Thus from the
ratio, column pressure and reboiler duty were evaluated. In plots it was found that for a yield of propylene above 95.2 % a
the performance analysis it was possible to visualize the effect column pressure of less than 20 atm and a reflux ratio of 10.2
of two variables on a specific design condition using response were required.
surface plots. The efficient operating criteria considered was
mole fraction of propylene in the distillate, which should be a TABLE I: INITIAL PLANT OPERATING DATA
minimum of 0.94. Item Feed Product Bottom
Material stream
For the simulation, the initial operating data of the column
Vapour Fraction 1.00 0.00 0.00
consisted of 55.32 kg mole/hr propylene, reflux ratio of 10.2, Temperature [C] 53.00 41.75 54.24
distillate flow rate of 246.98 kg mole/hr, column pressure of Pressure [bar] 18.23 17.22 19.25
18.23atm and feed temperature of 53 °C. Molar Flow 290.70 246.96 43.74
[kgmole/h]
Mass Flow 12344.37 10419.80 1924.57
[kg/h]
Liquid Volume 23.82 20.03 3.79
Flow [m3/h]
Heat Flow -227856.96 -43.83 -1156869.77
[kcal/h]
Component Mole Fraction
Propylene 0.81 0.946 0.043
Propane 0.19 0.054 0.957

B. Effect of Feed Stage Location

The interactive effects of feed with pressure and reflux
ratio in Fig. 3 and 4 are very similar. It was observed that as
the feed stage goes from stage 20 to stage 60 the purity of
propylene in the distillate increases. However, at a feed stage
of 70 a gradual decline in composition of propylene is
observed. The analysis further shows that, for optimum yield
of propylene, the feed stage should be located between stages
56 and 75. However, the operator may choose not to obtain
Fig. 1. Schematic of propane-propylene splitter. the optimum propylene yield of 95.5% and above since the
pressure and reflux ratio may lead to high reboiler duty.

III. RESULT AND DISCUSSION C. Reboiler Duty Consideration

The effect of the pressure and reflux ratio on the reboiler
The initial operating conditions, material and energy
duty (Qreboiler) on the one hand and of the pressure and feed
balances obtained from the simulation are shown in Table I.

2
 International Journal of Chemical Engineering and Applications, Vol. 8, No. 1, February 2017

stage on the reboiler duty on the other hand are shown in the requirement for the column obtained in the simulation was
surface plots in Fig. 5 and 6 respectively. As it is evident from 7.5×106 Kcal/h and this was considered as upper limit for
Fig. 5, high percent yield of propylene requires high reflux reboiler heat duty which was subject to be minimised.
ratio and lower pressure which result in increased reboiler
duty. A similar surface plot in Fig. 6, shows that a high
propylene percent in the distillate is promoted by a high feed
stage and a reduction in column pressure which again increase
the reboiler heat duty. In this respect more utilities will be
required which will negatively affect the operating cost and
profitability of the plant.
The aim of any process plant is to make profit by
minimising the cost of production. The cost of energy is a
major factor to be considered especially in a developing
country like Nigeria where energy is very expensive. The
simulation and the analysis of the unit should offer an avenue
for proper monitoring and optimization of the operation.
Fig. 4. Contour plot of % propylene in distillate vs. feed stage, reflux ratio.

Fig. 5. Surface plot of qreboiler (e+6kcal/h) vs reflux ratio, pressure (atm).

(hold value; feed stage 50)
Fig. 2. Contour plot of % propylene in distillate vs. reflux ratio, pressure
(atm).
Contour Plot of % propylene in distillate vs feed stage, pressure
80
% propy lene
in distillate
< 92.5
70 92.5 - 93.0
93.0 - 93.5
93.5 - 94.0
60 94.0 - 94.5
94.5 - 95.0
feed stage

95.0 - 95.5
> 95.5
50
Hold Values
reflux 10.5
40
Fig. 6. Surface plot of qreboiler (e+6kcal/h) vs feed stage, pressure (atm).
30
(hold value; reflux ratio 10.5)

20
17 18 19 20 21 22 23
pressure
Fig. 3. Contour plot of % propylene in distillate vs. feed stage, pressure
(atm).

D. Response Optimization
Fig. 7 is the response optimization of the
propane-propylene splitter unit carried out with Minitab 14
statistical software to maximize propylene yield in the
distillate and minimize the reboiler duty.
The efficient operating criteria considered for the splitter
unit were a minimum of 0.94 mole fraction and a target of Fig. 7. Response optimization for propylene yield.
0.95 mole fraction of propylene in the distillate which
represented the desirability limits for propylene maximization. The result of the optimization gave >95% propylene in the
The simulated propylene-propane plant in the case study distillate and 7.08×106 Kcal/h heat flow in the reboiler was
produced 9747.8kg/h of polymer grade propylene with a achieved at a column pressure of 20.199 atm, reflux ratio
quality of 94.72% propylene. The reboiler energy 10.023 and at a feed stage of 52 as the optimum condition of

3
 International Journal of Chemical Engineering and Applications, Vol. 8, No. 1, February 2017

operation. The result illustrated that despite an increase in the (NYSC) Engineering Community Development Service, Ego, Benin City,
July 2012-October 2012. He is currently occupied in process and equipment
column pressure from 18 atm to 20.2atm, reduction in reflux design in soap manufacturing. His research results are published in four
ratio from 10.2 to 10.02, the reboiler heat duty reduced from conference papers and journals, which include “Optimization of Ethanol
7.5×106 to 7.08×106Kcal/h. Fermentation of Cassava Wastewater Using Response Surface Method,”
Global Journal of Engineering Research, 2013. His research activities are
process design, modeling, simulation and optimization.
Mr Aniediong was awarded NYSC state merit award in 2012. He served
IV. CONCLUSION as a member of the secretariat committee NSChE annual conference 2013 in
Propylene- propane splitter in a petroleum refining case Uyo.
study has been simulated. Analysis of the data and
optimization of the column have shown a reduction in reflux
ratio from 10.2 to 10.02, an increase in the column pressure Etim N. Bassey was born in 1945 in Oron,
Nigeria. He obtained his BSc
from 18.23atm to 20.2atm, increased percent propylene in the (1972) ; MSc (1974) in chemical and
distillate from 94.7% to >95% and a saving of about 0.4x10 6 petrochemical engineering from the now Lvov
kcal/h in reboiler energy input. The results obtained have led Polytechnic National University, Ukraine and
PhD (1986) in chemical engineering from
to a better understanding of the process and have therefore
University of Bradford, England.
provided a guide to its optimum operation and control. He worked in the steel industry in Ajaokuta
and Warri, Nigeria from 1975-2005. In 2006 he
REFERENCES joined the University of Uyo, Nigeria as a senior
lecturer and became the head of Department of Chemical and Petroleum
[1] M. Heiaritz-Adrain and S. Wenzel, “Advanced propane Engineering from 2010-2012. He joined the Akwa Ibom State University,
dehydrogenation,” PTQ, pp. 83-91, 2008. Ikot Akpaden, Nigeria in 2012 where he is a professor in the Department of
[2] A. Aaron and A. Musaed, “Maximizing propylene production via FCC Chemical /Petrochemical Engineering. His publications include “Study of
Technology,” Appl Petrochem Res, vol. 5, pp. 377–392, 2015. the Contribution of Carbon IV and Carbon II Oxides in Methanol Synthesis
[3] F. Marcos, “Progress in the recovery of gaseous olefins.The combined with Modified Kinetic Model,” International Journal of Scientific Research
role of membranes, facilitated transport and ionic liquids,” Ph.D and Education, 2015; “Kinetic Models Evaluation in Methanol Production
Thesis, University of Cantabria, Spain, 2013. for improved Process Design,” International Journal of Innovative
[4] L. Rafael, “Refining and Petrochemicals: Challenges and solutions,” Research and Development, 2015. His research interests are in areas of
Presented at Haldor Topsoe Catalytic Forum, August, 2015. process design and integration, optimization, materials engineering, energy
[5] B. Hagger “Improving refinery distillation operations,” PTQ, Q2, pp. systems and environmental process engineering.
129-132, 2008. Prof. Bassey is a Registered Professional Engineer in Nigeria, a Fellow
[6] P. Martin, R. Latour, and A. Lynn “Closed–loop optimisation of of both the Nigerian Society of Chemical Engineers and the Nigerian Society
distillation energy,” Chemical Engineering Progress, pp. 33-37, 1981. of Engineers, and a Senior member of the American Institute of Chemical
[7] A. Shaheen “Role of catalyst type in the selective separation of olefinic Engineers. He served as a Country Representative in the International Iron
and paraffinic hydrocarbons using xerogel-based adsorbents,” Elsevier and Steel Institute Committee on Technology from 1990-1996. Prof Bassey
CARBON, vol. 46, pp. 1003-1009, 2008. is a recipient of 1986 Carl Hanson Travel Award of the British Institution of
[8] M. Surya, R. Yamuna, T. Sankarshana, F. Ismail, and S. Sridhar Chemical Engineers for his Research in Solvent Extraction; Nigerian Society
“Separation of binary mixtures of propylene and propane by facilitated of Chemical Engineers “Pinnacle Holdings” Award in 1992, for
transport through silver incorporated poly (ether-block-amide) Development of a process for beneficiation of low grade Nigerian iron ore
membranes,” Oil & Gas Science and Technology – Rev. IFP Energies into superconcentrade for Direct Reduction Steel production, National Prize
nouvelles, vol. 70, no. 2, pp. 381-390, 2015. Winner for RMRDC Process Equipment and Plants Design in 2004.
[9] P. Jeroen, “Membranes for ethylene/ethane separation,” Ph.D Thesis,
University of Twente, Enschede, Netherlands, 2013.
[10] B. Odigo, “Evaluation of Propane – propylene splitter in warri refining
and petrochemical company,” Nigerian Society of Engineers Technical
Report, 2003.

Aniediong M. Umo was born in 1985 in Ukpom

Abak, Nigeria. He received his B.Eng. (2011) and
M.Eng. (2016) in chemical engineering from the
University of Uyo, Akwa Ibom state, Nigeria.
He had served as a quality control Analyst
(industrial trainee) Peacock Paint Nigeria Ltd (2009),
Presiding Officer Eghosa Grammar School Polling
Unit, Edo State Gubernatorial Election July 2012,
and was President of National Youth Service Corps