See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/270453894

Dynamic simulation of process controls

Article in Petroleum Technology Quarterly · January 2015

CITATIONS READS

0 687

2 authors:

Sepehr Sadighi Reza Seif Mohaddecy

Research Institute of Petroleum Industry (RIPI) Research Institute of Petroleum Industry (RIPI)
180 PUBLICATIONS 599 CITATIONS 65 PUBLICATIONS 205 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Sepehr Sadighi on 06 January 2015.

The user has requested enhancement of the downloaded file.

 Dynamic simulation of process controls

Simulation of a catalytic naphtha reformer confirmed the accurate performance

of control instrumentation

SEPEHR SADIGHI and S REZA SEIF MOHADDECY

RIPI

C
atalytic reforming is used to effective tool for DCS operator modernised unit or process sooner,
upgrade a significant portion training and design verification. and complete it faster
of the low octane heavy naph- Dynamic simulation is useful • Train and develop operators on
tha that is unsuitable as a high throughout the entire lifetime of a procedures and systems that are
octane gasoline blending compo- plant: from conception to decom- safe, optimised, and environmen-
nent. A control system is essential missioning. Some of the benefits of tally friendly
to guarantee high productivity dynamic simulation for chemical • Provide simulation applications
whilst the safety aspects of the processes are: for improving real-time operation
process should be considered. In • Identify unit or process operating and control.
process control, it is often difficult constraints at the conceptual design Modern dynamic simulators can
to measure some important process phase before capital expenditures be used to provide this type of
variables such as composition of are committed dynamic information. These flow
product or yield of process because • Improve the process sheet simulators are supported by
of the limitations of measurement. • Identify and correct weaknesses in strong databases, complete sets of
This task is usually carried out by the automation system design before modules and flexible simulation
simulating the process in dynamic commissioning and operation tools, so the user is capable of
mode. • Demonstrate control applications developing process-wide simula-
For almost 20 years, dynamic before deployment tors, and studying the performance
simulation has been used as an • Begin start-up of a new or of the whole plant.

Recycle logic
Recycle
R

Reactor Reactor Reactor

No. 1 No. 2 No. 3

Separator
Recycle gas
compressor Splitter

Naphtha

Heater Heater Heater Cooler

No. 1 No. 2 No. 3 Separator Off
Stabiliser gas

Heat exchanger
LPG

Heater

Gasoline

Figure 1 Catalytic reforming unit of the target oil refinery

www.eptq.com PTQ Q1 2015 31

 Specifications of heavy naphtha feed Properties of catalyst for naphtha Operating conditions of the catalytic
reforming process naphtha reforming process
Variable Value
Feed flow rate, m3/h 95.29 Property Value Variable Value
Density, kg/m3 664.4 Surface area, m2/g 200.62 Inlet temperatures, °C 497-515
Pore size, nm 311.37 Hydrogen pressure, bar 24.5-31
PONA Pore volume, cm3/g 0.57 Hydrogen to hydrocarbon molar ratio 5-6
Paraffins, vol% 62.81 Mechanical strength, mPa >2 Liquid hourly space velocity (LHSV) 1.6-2.0
Naphthenes, vol% 25.68 Pt content, wt% 0.18 Yield, vol% 70-80
Aromatics, vol% 11.51

ASTM D 86 Table 2 Table 4

IBP, °C 92
5%, °C 98
10%, °C 101 Catalyst distribution in reforming reactors
30%, °C 107
50%, °C 114
70%, °C 122 First reactor Second reactor Third reactor
90%, °C 136 Catalyst weight, kg 8648 14 223 22 452
FBP, °C 150 Catalyst distribution, wt% 20 30 50

Table 1 Table 3

Most of the studies have focused hydrotreated naphtha, called during the reforming process (the
on steady state modelling and the Platcharge, is introduced to the gas stream) is recycled, and is then
reaction section, that is fixed bed catalytic reforming process. The mixed with the naphtha. Finally,
catalytic reactors in series. In the specification of the naphtha feed is the liquid product leaving the
field of dynamic simulation, Hu et shown in Table 1. This feed is separator is introduced to the stabi-
al developed a dynamic mode for treated in a Unifiner (naphtha liser in which LPG and light gases
the catalytic reforming plant and hydrotreating) plant; therefore, the are separated from the gasoline.
tried to study the effect of the feed sulphur, nitrogen, olefin and metals Hence the vapour pressure of the
flow rate and reactor temperatures contents are negligible. gasoline can be set according to
on the yield of the process. Arani et Figure 1 shows that Platcharge is market requirements.
al developed a dynamic model for first preheated by the first furnace The properties of the catalyst and
the catalytic naphtha reformer its distribution for the naphtha
which focused on the reaction Dynamic simulation reforming process are shown in
section of this unit. The action of Tables 2 and 3, respectively.
the controllers and dynamic is useful throughout Moreover, the normal operating
responses were not studied in this conditions of the unit are shown in
work. the entire lifetime Table 4. The selected operating
Therefore, the purpose of this conditions of the process depend
research is to develop a dynamic of a plant: from strongly on the life (activity) of the
simulator to study the action of catalyst.
the main controllers in the plant. In
conception to
this way, a whole commercial fixed
bed catalytic reforming process was
decommissioning Process simulation
Catalytic reforming is often
selected, and was simulated using modelled and simulated based on
the Aspen-Hysys platform. After (Heater No. 1), and then enters the the number of reactive species,
validating the simulation, the action first reactor (Reactor No. 1) where and the type of kinetic model used.
of the controllers is studied in the naphthenes are dehydrogenated to The presence of many components
dynamic mode using the set point aromatics. Then the product stream as reactants or intermediate prod-
tracking or servo method. from the first reactor passes ucts in the reactive mixture, as a
through the second reactor (Reactor result of the presence of new reac-
Process description No. 2), and the outlet stream enters tions, leads to a sophisticated
A commercial fixed bed catalytic the third reactor (Reactor No. 3). challenge in modelling the process.
naphtha reforming unit, called The overall reforming reactions are To decrease these complications,
Platformer licensed from Chevron endothermic; therefore, a preheater reactants in the mixture are
Research Corporation, was chosen (heaters Nos. 1, 2 and 3) should classified in certain and limited
as a case study. The feed of the essentially be provided before each groups called pseudo-components.
plant prior to entering catalytic reforming reactor. Additionally Arrhenius and
reforming should undergo Next, the product stream from Langmuir–Hinshelwood kinetics
hydrodesulphurisation (HDS) in the last reactor enters the separator, are widely used for kinetic-based
the hydrotreating unit. Then the wherein the hydrogen produced catalyst modelling and simulation

32 PTQ Q1 2015 www.eptq.com

 Comparison of product yields: Characterisation of heavy naphtha feed
simulation results and actual data

Paraffins mol%
Simulation Actual 1 NP6 n-Hexane 4.698
Paraffins, vol% 37.5 34 5 IP6 2-Methyl-Pentane 2.129
Naphthenes, vol% 0.93 0.01 2 NP7 n-Heptane 8.859
Aromatics, vol% 61.22 64.4 6 IP7 3-Methyl -Hexane 7.521
3 NP8 n-Octane 8.367
7 IP8 3-Ethyl-Hexane 10.055
Table 6 4 NP9 n-Nonane 5.395
8 IP9 2-Methyl-Octane 8.972
of catalytic naphtha reforming. Total = 55.996
For this study, the kinetic Naphthenes mol%
17 N6C6 Cyclo hexane 2.074
constants presented in previous 13 N6C5 Methyl Cyclo Pentane 2.091
work were used. The feed of the 18 N7C6 Methyl Cyclo Hexane 5.406
catalytic reformer (see Table 1) was 14 N7C5 Ethyl Cyclo Pentane 4.176
lumped according to the structural 19 N8C6 Ethyl cyclo Hexane 2.328
15 N8C5 Methyl Ethyl Cyclo Pentane 3.105
components using a discrete 20 N9C6 Iso Propyl Cyclo Hexane 4.618
lumping procedure with 26 16 N9C5 n-Butyl Cyclo Pentane 0.786
pseudo-components (see Table 5). Total = 24.584
Aromatics mol%
9 A6 Benzene 0.894
Results and discussions 10 A7 Toluene 4.724
Steady state simulation results 11 A8 Ethyl Benzene 6.641
From start of run to end of run, 12 A9 Iso Propyl Benzene 2.906
actual test results were gathered Total = 15.165
Lights hydrocarbon mol%
from the target reforming unit. 22 C1 0
Because for a specified catalyst life 23 C2 0
the simulator was used for predict- 24 C3 0
ing the steady state condition of the 25 C4 0
26 C5 0.171
Platformer, the data used for simu-
lation should be selected from the
normal condition when no abnor- Table 5
malities, such as tower flooding,
emergency depressurisation, or Constants of the controllers selected for the catalytic reformer
pump or compressor shutdown,
occurred. Before using these data to
Constant LIC-251 PIC-255 PIC-256 LIC-253 PIC-255 TIC-252
estimate the tuning parameters, it Kc, 0C/% 5 1.5 1.5 5 1.5 2
was necessary to validate them. If a τI, h ∞ ∞ ∞ ∞ ∞ ∞
reasonable overall mass balance τD, h 0 0.1 0.1 0 0.1 0.1
(±5%) cannot be calculated, the
validity of the test run is Table 7
compromised.
After validating the data, the through faults such as signal trans- addition, the controllers of the cata-
average absolute deviation (AAD%) mission, calibration and power lytic reforming plant were the
of the simulation for the product fluctuation of instruments which proportional-integral-differential
flow rate, outlet temperatures of could not be excluded from the (PID) type. The corresponding
the first, second and third reactors, actual data. control constants are shown in
and the volume yield were 2.4%, Table 7. Figure 2 shows the simula-
0.98%, 0.1%, 0.04% and 2.39%, Dynamic simulation results tion after these values are applied.
respectively. A comparison Before starting the dynamic simula- In the catalytic reforming plant,
between the simulated paraffinic, tion, it was necessary to define the the main phenomenon creating
naphthenic and aromatic contents size of equipment, according to its dynamic disturbance is deactiva-
of the product against actual data is design values, in the Aspen-Hysys tion of the catalyst. Therefore, the
shown in Table 6. simulator. Moreover, the degree of controllers of the plant should be
From the simulation results, it freedom (DOF) of the dynamic able to lead the plant to steady
can be concluded that the simula- simulation should be set to zero. conditions and so maintain produc-
tion program was reliable enough This task was performed by setting tivity and robustness. For the unit
for predicting the behaviour of the the flow rate of the naphtha feed under study, the reliability of the
catalytic reforming unit being stud- and recycle streams (F streams), controllers was examined using the
ied. It was assumed that the main and also the pressures of all prod- servo or set point tracking method.
source of deviations was the possi- uct streams, leaving the tower (P) The main instruments studied
bility of errors in gathering data streams as fixed variables. In were:

www.eptq.com PTQ Q1 2015 33

 Recycle R
PIC-256
TIC-258 Splitter Condenser
Mixer 2V-251 2V-252 @COL 1
TIC-253 TIC-263 Off gas
Platcharge Purge out
TIC-101
2V-253 Vapour LIC-255

LPG out
PIC-255 TIC-252
2H-251 2H-252 2H-252
LIC-251

2E-252 2E-253
2V-255 Reformate
TIC-100 Out R3 Reboiler
R @COL 1

LIC-253

Figure 2 Dynamic simulator of catalytic naphtha reformer

ing of these controllers up to steady

70
state conditions is shown. The
65
changes on the set point of control-
2V-255 liquid level, %

60 lers were implemented according to

55 the design criteria of operating
50 conditions in the target plant.
45
Figure 3 shows the results of set
point tracking of LIC-251 (level of
40
the flash tank). At first, this was set
35 at the 50% level, then manipulated
30 to 30%, 60% and, again, to 50%.
25 The curve illustrated confirms that
20 the simulator could successfully
0 5 10 15 20 25 control step variation in the level of
Time, minutes the flash tank. Moreover, the
controlling parameters – gain,
Figure 3 Set point tracking for the flash tank level integral and differential constants –
were appropriately tuned to keep
up with the variations in the set
2592
point of LIC-251.
2V-255 vessel pressure, kPa

In Figure 4, the pressure of the

2542
flash tank was first set at 2433 kPa.
2492
Then, the set point was changed to
2500 kPa, 2420 kPa, 2432 kPa and
2442 finally to 2506 kPa. These values
were chosen according to the actual
2392 pressures observed from start of
run (SOR) to end of run (EOR). It
2342 was seen that the controller was
capable of tracing the pressure
2292 during the life of the catalyst.
0 5 10 15 20 25 The recorded results for set point
Time, minutes tracking of the condenser pressure
of the stabiliser are shown in Figure
Figure 4 Set point tracking for the flash tank pressure 5. To perform set point tracking for
this unit, the pressure of the vessel
1. The level controller of the flash stabiliser tower (PIC-256) was stabilised at 1860 kPa. Then it
tank installed after the third reactor 4. The level controller of the was manipulated to 1800 kPa, 1900
(LIC-251) reboiler provided for the stabiliser kPa and again to 1860 kPa, respec-
2. The pressure controller of the tower (LIC-253) tively. The results show that the
flash tank installed after the third 5. The temperature controller of the instrument under study (PIC-256)
reactor (PIC-255) reboiler provided for the stabiliser was properly designed to control
3. The pressure controller of the tower (TIC-252). the pressure of the condenser
condenser provided for the In Figures 3 to 7, set point track- within the operating conditions.

32 PTQ Q1 2015 www.eptq.com

 Figures 6 and 7 demonstrate the
servo study of the temperature 1950
controller of the reboiler (TIC-252),

Condenser vessel pressure,

and the level controller of the 1900
reboiler (LIC-253). For this study,
the temperature of TIC-252 was set
1850
to 230°C, 225°C, 240°C and 220°C,

kPa
respectively. For LIC-253, setting
levels were 40%, 60% and again 1800
50%. These values were chosen
according to the observed data 1750
during the life of the catalyst.
Figures 6 and 7 clearly justify the
1700
accurate function of the under 0 10 20 30 40 50 60
study controllers without any over- Time, minutes
shooting in the process variable.

Conclusions Figure 5 Set point tracking for the tower condenser pressure
The dynamic simulation of a
commercial catalytic naphtha
reforming process using the Aspen- 249
Reboiler vessel temperature,

Hysys simulator is presented in this 244

article. The kinetic parameters of
239
the model were based on the steady
state condition obtained from the 234
previous works.
ºC

229
At first, the simulator was used
to predict the process variables at 224
the steady state condition. 219
Comparing the results with the real
214
outputs showed that the AAD% of
simulation for the product flow 209
rate, outlet temperatures of the 5 10 15 20 25 30 35 40 45 50 55 60 65
first, second and third reactors and Time, minutes
volume yield were 2.4%, 0.98%,
0.1%, 0.04% and 2.39%, respec- Figure 6 Set point tracking for the reboiler temperature
tively. Moreover, comparison of the
simulated paraffinic, naphthenic
and aromatic contents of the prod- 71.5
Reboiler vessel liquid level, %

uct against actual data was 66.5

acceptable at steady conditions.
61.5
The main instruments of the cata-
56.5
lytic reformer were also studied in
the dynamic model. The results, 51.5
according to the set point tracking 46.5
or servo method, confirmed the 41.5
accurate function of the controllers 36.5
without any overshooting or offset
31.5
in the process variable. This
26.5
approach could be applied for
developing an operator training 21.5
120 130 140 150 160 170 180 190 200 210 220 230
simulator for the catalytic reform-
Time, minutes
ing plant.

Figure 7 Set point tracking for the reboiler level

Reza Seif Mohaddecy is a Project Manager in engineering from Sharif University of Catalytic Reaction Engineering Department,
the Catalysis and Nanotechnology Division, Technology. Research Institute of Petroleum Industry (RIPI).
Catalytic Reaction Engineering Department, Email: Seifsr @ripi.ir He holds a PhD in chemical engineering from
Research Institute of Petroleum Industry Sepehr Sadighi is a Project Manager in the Universiti Teknologi Malaysia.
(RIPI), Tehran, Iran. He holds a MS in chemical Catalysis and Nanotechnology Division, Email: Sadighis @ripi.ir

www.eptq.com PTQ Q1 2015 33

View publication stats