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

net/publication/325655703

Coker Gas Plant Design

Technical Report · May 2017

DOI: 10.13140/RG.2.2.18685.87526

CITATIONS READS
0 3,231

5 authors, including:

Erik Magnuson Bruce Edward Kirkpatrick

University of Colorado Boulder University of Colorado
1 PUBLICATION   0 CITATIONS    30 PUBLICATIONS   333 CITATIONS   

SEE PROFILE SEE PROFILE

Galen Roda Connor Miksch

University of Colorado University of Colorado Boulder
8 PUBLICATIONS   13 CITATIONS    5 PUBLICATIONS   41 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Undergraduate Projects View project

Master's Thesis View project

All content following this page was uploaded by Bruce Edward Kirkpatrick on 09 June 2018.

The user has requested enhancement of the downloaded file.

 University of Colorado Boulder Department of Chemical Engineering, CHEN 4530, Spring 2017
Senior Design Project, ExxonMobil (PS7), Team 8

Coker Gas Plant Design

Erik Magnuson1 *, Andrew Burns2 , Bruce Kirkpatrick2 , Galen Roda2 , Connor Miksch2
ExxonMobil Project Coordinator: Jay Westcott

Executive Summary
Coking facilities are refineries that upgrade residual oils from crude distillation to more valuable light hydrocarbons.
These processes usually involve fractionation columns, which separate the coking products into fuel gas, liquefied
petroleum gas, naphtha, and heavier ends. Using Aspen Plus R
and Aspen HYSYS R
, we designed a coking gas
plant to separate light gases, C3 s, C4 s, and heavy mixing components in a sour feed stream originating from the
top of a coking process fractionator.
Our finalized design recovers nearly 100% of the fed refinery fuel gas, 92% propylene, 96% propane, 92%
butylene, 74% butane, and 99% gasoline mixing component at full capacity. All product purities are above 90%
with the exception of refinery fuel gas and butylene which have purities of 65% and 73% respectively. Product
purity meets all prescribed specifications, including successful extraction of H2 S and CO2 contaminants.
The sale of these products corresponds to an annual market value of $292.8M. With annual fixed costs of $49.3M
and variable costs of $70.3M, our plant yields an investor rate of return of 23.6%, a 0.5% improvement on the
alternative design. This is based on an estimated total permanent investment The payback period is estimated at
4 years with an annual return on investment of 27.0% The primary design also has the added benefit of improved
product qualities and simplification of the sour gas sweetening system. Investment viability is most sensitive
to product sales price and operating costs. The sensitivity of sales is of particular concern given the current
economic landscape surrounding petroleum futures.
Sour gas and the amine treatment required for H2 S removal represent the primary safety concerns in the process,
scoring 80/125 in a system FMEA analysis and achieving the maximum weighted environmental index factor
using Biwer’s method when compared to other process components. These concerning elements were mitigated
through process control implementation or on-site handling outside of the scope of this project.
Keywords
Coker gas — Separation process design — Chemical engineering
1 Department
of Chemical Engineering, University of Colorado, Boulder, CO
2 Department
of Chemical and Biological Engineering, University of Colorado, Boulder, CO
*Corresponding author: erik.magnuson@colorado.edu
 Coker Gas Plant Design — 2/28

Contents Acknowledgments 16
References 16
1 Introduction 3
Appendix A: Process Specifications 18
1.1 Project Goal and Scope . . . . . . . . . . . . . . . . . .3
Inlet Conditions . . . . . . . . . . . . . . . . . . . . . . . 18
1.2 Specifications . . . . . . . . . . . . . . . . . . . . . . . . .3 Outlet Requirements . . . . . . . . . . . . . . . . . . . 18
1.3 Assumptions . . . . . . . . . . . . . . . . . . . . . . . . . .3 Utility Parameters . . . . . . . . . . . . . . . . . . . . . . 18
1.4 Thermodynamics . . . . . . . . . . . . . . . . . . . . . . .3 R
Appendix B: Aspen Plus
Initial Model 19
R

R


2 Aspen Plus and HYSYS Modeling 4 Appendix C: Process Unit Sizing 20
2.1 Pressurization Scheme . . . . . . . . . . . . . . . . . . 4 FUGK Method . . . . . . . . . . . . . . . . . . . . . . . . 20
Design • Process Benefit Decanter Sizing . . . . . . . . . . . . . . . . . . . . . . . 20
2.2 Absorber and Stripping Column . . . . . . . . . . . . 5
Appendix D: Process Unit Costing 21
Design • Process Benefit
Columns . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.3 Debutanizer . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Decanters . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Design • Process Benefit
Compressors . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.4 C3 -C4 Separators . . . . . . . . . . . . . . . . . . . . . . . 6
Pumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Design • Process Benefit
Heat Exchangers . . . . . . . . . . . . . . . . . . . . . . 21
2.5 Amine Sweetening System . . . . . . . . . . . . . . . . 6
Design • Process Benefit
Itemized Total Cost . . . . . . . . . . . . . . . . . . . . 22
Equipment Materials . . . . . . . . . . . . . . . . . . . 22
2.6 Enriching Column . . . . . . . . . . . . . . . . . . . . . . 7
Design • Process Benefit Appendix E: Material Balances 23
2.7 Miscellaneous Units . . . . . . . . . . . . . . . . . . . . . 7 Pressurization Scheme . . . . . . . . . . . . . . . . . 23
Design • Process Benefit Absorber-Stripper-Enriching Column System . . 23
2.8 Other Analyses . . . . . . . . . . . . . . . . . . . . . . . . 7 Debutanizer . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Economic Analysis • Environmental Survey • Safety Study C3 -C4 Separator Scheme . . . . . . . . . . . . . . . . 23
3 Alternative Model 8 Individual Absorber Material Balance . . . . . . . 23
3.1 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Individual Stripping Column Material Balance . 23
Pressurization Scheme • Absorber and Stripping Column • Individual Enriching Column Material Balance . 23
Debutanizer • C3 -C4 Separators • Amine Sweetening Systems
Appendix F: Energy Balances 25
3.2 Product Recoveries . . . . . . . . . . . . . . . . . . . . 10
Appendix G: Failure Mode Effects Analysis 26
3.3 Utilities Summary . . . . . . . . . . . . . . . . . . . . . . 11
Appendix H: Environmental Impact 27
3.4 Economic Analysis . . . . . . . . . . . . . . . . . . . . . 11
Appendix I: Economic Sensitivity Plots 28
4 Finalized Model 11
4.1 Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Pressurization Scheme • Absorber, Stripping and Enriching
Columns • Debutanizer • C3 -C4 Separators • Amine Sweetening
Systems
4.2 Product Recoveries . . . . . . . . . . . . . . . . . . . . 13
4.3 Utilities Summary . . . . . . . . . . . . . . . . . . . . . . 13
Energy Analysis
4.4 Economic Analysis . . . . . . . . . . . . . . . . . . . . . 13
4.5 Environmental and Safety Survey . . . . . . . . . . 14
Modified Failure Mode Effects Analysis • Abridged Biwer’s Method
• MEA Considerations
4.6 Social Impact Recommendation . . . . . . . . . . . 15
5 Conclusions 15
5.1 Cost-Benefit Analysis . . . . . . . . . . . . . . . . . . . 15
Costs • Benefits
5.2 SWOT Analysis . . . . . . . . . . . . . . . . . . . . . . . 15
Strengths • Weaknesses • Opportunities • Threats

1. Introduction
The coking process is a common refining scheme wherein low
quality, high sulfur, heavy hydrocarbons are converted to more
valuable light products [3, 21, 24]. These processes typically
upgrade residual oils from crude oil distillation by thermal
cracking followed by fractionation to provide a preliminary
product separation [2, 25] (Figure 1A). Usually, the post-
coking fractionation results in several light ends, including
fuel gas, liquified petroleum gas (LPG), and naphtha (Figure
1B), which require further separations to meet specifications
for retail. Sulfur content, composition, and storage conditions
are important specifications for fuel quality (and subsequent
price determination) in the petroleum industry [46].
Refining the fuel gas and LPG exiting a fractionation
column in a coking scheme is often achieved by a series of
strippers, absorbers, and other separating units to produce
isolated light ends above C3 hydrocarbons, butane and buty-
lene, propane and propylene, and mixed heavy ends [28]. As
mentioned, sulfur, generally in terms of weight percent hydro-
gen sulfide, and carbon dioxide content must be limited per
industry standards. If not removed before consumption, these
acidic components would cause problems with freezing and
corrosion and create significant environmental, health, and
safety hazards [42]. To sequester CO2 and H2 S from valuable
fuel streams, the following reactions are employed:
H2 S + Amine ↔ [Amine]H+ + HS−
CO2 + H2 O + Amine ↔ [Amine]COOH+ + OH−
Common alkanolamine components for these reactions are
monoethanol amine (MEA), diethanol amine (DEA), di-isopro-
R
R
panol amine (DIPA), and N-methyl diethanol amine (MDEA).
The proton transfer shown in the above reactions results in the
partitioning of “sour” sulfurous and acidic components into
an aqueous dirty MEA stream while sweetened gas streams
can be stored for retail and industrial applications [7, 14].
Naturally, the economics of a refinery are determined by
the qualities (specific compositions and contaminant levels)
of its products and its utilities usage [34]. Overlarge utility
costing can be mitigated by improved heat exchange networks
and non-consumable utilities recycling schemes [68]. Addi-
tionally, environmental and social impact analyses and design
safety surveys are required for any refining process, as these
designs include hazardous components and operate at poten-
tially dangerous temperatures and pressures, posing a possible
threat to surrounding employees, communities, and ecosys-
tems [16, 17, 19]. Ultimately, effectively optimized and safely
implemented coking refineries can provide investment-quality
internal rates of return and competitive annual revenues in the
oil and gas industry.
1.1 Project Goal and Scope
The goal of this project was to design a processing plant for
separating the fuel gases exiting the fractionator of a coking
process, the composition of which was provided by Exxon-
Mobil. As part of a larger project (installation of a coking
Coker Gas Plant Design — 3/28
unit), the processing plant was presumed to require only onsite
equipment design; no cooling towers, electrical substations,
boilers, or sulfur handling facilities were considered in the
design on this project [67].
Included in this project are a complete design of our coker
gas processing scheme (flow diagrams, material and energy
balances, equipment and utilities summaries), a summarized
alternative design, economic analyses with sensitivity testing,
environmental and safety assessments, societal impact survey,
and strengths-weaknesses-opportunities-threats (SWOT) and
cost-benefit evaluations.
1.2 Specifications
Process specifications were provided by ExxonMobil and are
included in full detail in Appendix A: Inlet, Outlet, and Utility
Specifications. Briefly, approximately 232,000 lb/hr of sour
gas at 20 psig and 150◦ F from a coking process feeds our
process: this flow is made up of 7% water, 17% contaminants
(hydrogen sulfide and carbon dioxide), 35% light ends (hy-
drocarbons above C3 ), 16% C3 s, 14% C4 s, and the remainder
heavy ends (below C4 s). We were also provided with nine
acceptable outlet conditions consisting of a gasoline blend-
ing component, four C4 conditions, three C3 conditions, and
a refinery fuel gas. Each condition was provided with cor-
responding composition requirements, maximum hydrogen
sulfide content, product value, and any other relevant spec-
ification. Available utility data were given with according
supply parameters, return requirements, and costs.
1.3 Assumptions
1. Trust Aspen Plus
and Aspen HYSYS
[15, 31].
2. The propane refrigerant R290 is completely recycled at
an efficiency of 0.6 (based on a comparable cycle [44]).
3. Compressors in the system have an isentropic efficiency
of 0.7 and a mechanical efficiency of 0.9.
4. Amine recycling is outside of the project scope (but
included onsite at negligible cost).
5. No MEA leaches into product flows.
6. No land purchase is required for the plant.
7. Plant construction ends and full capacity is achieved by
the beginning of year four.
8. The cost of royalties is 0.5% gross revenue.
9. Modified Accelerated Cost Recovery System (MACRS)
depreciation scheme is 10 years.
10. Full capacity is defined as 8,000 operating hours/year.
1.4 Thermodynamics
Because our system was primarily composed of hydrocar-
bons, we selected the Soave-Redlich-Kwong equation of state
(SRK), which follows the form:
RT a(T )
P= −
vm + c − b (vm + c)(vm + c + b)
where P is pressure, T is temperature, R is the ideal gas con-
stant, vm is molar volume, and a, b, and c relate to compress-
ibility and acentric factor. SRK was devised to fit hydrocar-
bon vapor pressure data and does well for the components in
 Coker Gas Plant Design — 4/28

Figure 1. Basic process flow diagrams of crude oil distillation with highlighted coking process (A) and coking process with highlighted
light product streams (B). Adapted from [21] and [25].

our feed [41]. Although other thermodynamic models like light ends from the gasoline mixing component without a
Peng-Robinson and UNIFAC are appropriate for coker gas fractionator.
separations, the presence of light gases (CO2 , N2 , and H2 ) in With our improved understanding of the process origin, we
the inlet to our system made SRK preferable [1]. turned our attention to a gas-hydrocarbon effluent separation
Our system was mainly modeled with Aspen Plus R
and scheme [28] involving the absorbers and strippers originally
R
the SRK equation of state. Aspen HYSYS was used to present in the distillation train of the fractionator [33]. which
simulate the MEA scrubbing system with Acid Gas Clean- give rise to our alternative and finalized designs.
ing, as this package allows for rigorous modeling of amine
sweetening and carbon dioxide removal systems [15]. 2.1 Pressurization Scheme
2.1.1 Design
2. Aspen Plus R and HYSYS R Modeling Once the fractionator was replaced by a pressurization scheme
[34], a maximum-capability analysis of the compressors was
An initial attempt to develop our process (which can be viewed
performed by combining the return requirements from Exxon-
in Appendix B: Aspen Plus R
Initial Model) erroneously fo-
Mobil with the heuristic maximum individual compressor
cused on implementing a fractionating column. After con-
process outlet temperature.
structing a model that began with the fractionating unit, it
became apparent that the resulting separations in heavy ends
were unnecessary because of several reasons: First, the accept-
able product compositions provided by ExxonMobil do not
include isolated compounds below C4 (rather, all heavy ends
are grouped into a gasoline mixing component with a speci-
fied Reid Vapor Pressure), so extensive processing of these Figure 2. Example feed pressurization scheme with two cycles,
high-mass compounds is fruitless; second, and more impor- where each cycle follows the format: inlet compression, compressed
stream cooling, and cooled stream decanting.
tantly, is the relative absence of truly “heavy” compounds (in
the context of a coking process) in the feed to our separation Following the design shown in Figure 2, a system of three
scheme [21, 46]. compressor cycles was designed such that the following con-
Since our initial PFD was developed through a literature ditions were met:
review (under the presumption that our feed originated from a 1. Maximum individual compressor process outlet temper-
coke drum), we misguidedly added the fractionator to the front ature does not exceed 300◦ F [67].
of our process expecting to produce a product separation like 2. Cooling occurs to extent possible while achieving return
the one shown in Figure 1B, with naphtha, kerosene, diesel oil, requirement for cooling water (< 130◦ F; Appendix A).
and other heavy outlets [29, 30, 36]. After several meetings 3. Final temperature of bulk material flow meets minimum
with Project Coordinator Jay Westcott and further discussion approach temperature of 10◦ F.
regarding the addition of a steam stripper to the fractionator This design resulted in the maximum compression cycle out-
for improved recovery of light ends, we determined that a puts (with a feed at 20 psig) of 119 psig for cycle one, 489
suitable pressurization scheme could separate our valuable psig for cycle two, and 1,719 psig for cycle three.
 Coker Gas Plant Design — 5/28

and C3 s as the heavy key. By setting a basis for the lean oil
entering the absorber, the known feed conditions and desired
absorber vapor separation (C2 s and lighter) can be used to
calculate the stripper liquid flowrate and composition. The
basic concept behind the recycle loop involving the wet gas
and rich oil from the decanter (Figure 4) is that increases in C3
flux through the loop increase the overall C3 product output,
improving the overall process separation efficiency despite
less-effective separations in the units making up the loop.

Figure 3. Example feed pressurization scheme with two cycles and

labelled inlets and outlets.

2.1.2 Process Benefit

The compressor cycle itself is important for the separation
of condensed water and hydrocarbons from higher pressure
hydrocarbon gases [67]. In these cycles, the compression
stages involve rapidly rotating machinery that pressurizes
gases, while interstage coolers remove residual heat from com-
pression. The interstage knockout drums act as post-cooling
three-phase decanters, providing important preliminary water
and oil separation (Figure 3), which reduces the liquid flow
through downstream columns (subsequently reducing the size Figure 4. Example absorber-stripping column scheme illustrating
and cost of the columns). Moreover, the removal of water the repeated separation of condensed oil and water from an
increasingly pressurized gas stream.
improves the overall purity of outlet products and lessens the
mass burden on the system, the initial pressurization provides
the foundation for the necessary product partitioning in later 2.2.2 Process Benefit
units, and the increased pressure at the front of the process The absorber and stripping column process takes advantage
creates a gradient, eliminating the need for downstream large- of a recycle loop to improve the overall recovery of heavy
scale repressurization. compounds from individually less effective unit operations.
In so doing, the system reduces the overall size of the MEA
2.2 Absorber and Stripping Column treatment process by allowing for subsequent separations in
After pressurization, any streams not exiting the process as the C3 s and heavier without additional sweetening (when im-
free water reconvene in the absorber and stripping column plemented at optimal conditions), which simplifies achieving
system for partitioning into a sour dry gas (C2 s and lighter) higher quality heavy product outlets.
from the top of the absorber and a heavy blending component
(C3 s and heavier) from the bottom of the stripping column 2.3 Debutanizer
[4, 5, 33]. A debutanizer follows the stripping column bottoms to sep-
2.2.1 Design
arate the sub-C4 gasoline mixing components from a C3 -C4
The absorber served as the foundation for the design of this mixture, which feeds the remaining units in the pressurization
process. First, we iteratively sized the number of stages in the scheme bottoms distillation train.
absorber in combination with the lean oil flow rate to achieve
maximum separation between C2 s and C3 s in the feed stream
[23]. A mass balance was constructed following the basic
form:
unit feed = absorber vapor + stripper liquid - lean oil
where the unit feed is the combination of rich oil from the
first compression cycle and the high pressure vapor from the
final compression cycle. This mass balance was combined
with the absorber staging to avoid pulling light ends into the
Figure 5. Example debutanizer separating gasoline mixing
stripper liquid and heavy ends into the absorber vapor, while
component from lighter gas hydrocarbons (C3 s and C4 s).
the stripping column was designed with ethane as the light key

2.3.1 Design
The debutanizer separations design began with the selection
of butane as the light key. Heavy key selection proved to be a
more complex endeavor because, although hexane is an abun-
dant and obvious candidate for heavy key, isopentane is nearer
butane in terms of volatility [38, 29, 32, 35, 34] . Ultimately,
we selected isopentane as our heavy key and proceeded with
the Fenske-Underwood- Gilliland-Kirkbride (FUGK) method
for column sizing (which is described in Appendix C: Pro-
cess Unit Sizing). The preliminary FUGK results were input
to our simulation and iterated through the Aspen software,
which continued until a satisfactory separation was achieved
between butane and isopentane.
2.3.2 Process Benefit
As a ‘middleman’ in the bottoms distillation train, the debu-
tanizer provides the heaviest product outlet of the overall
coking gas design (the gasoline mixing component) while di-
recting a vaporous stream of C3 s and C4 s to the three-column
C3 -C4 separator system (Figure 5). The gasoline mixing com-
ponent serves an additional purpose, as the bottoms stream
from the debutanizer is split and partially recycled to the ab-
sorber as the lean oil feed. In the absorber, the presence of
these components improves the transfer of the bulk heavy
ends to the stripping column, while the remainder of the gaso-
line mixing component exiting the debutanizer is cooled and
stored for transport and retail.
2.4 C3 -C4 Separators
The bottoms distillation train concludes with a sequence of
three columns in the C3 -C4 separation scheme, which contains
the largest fractionating columns of the overall design [43,
45]:
Figure 6. Example C3 -C4 separation scheme with isolated alkane
and alkene outlets.
Coker Gas Plant Design — 6/28
2.4.1 Design
As with the debutanizer, light and heavy keys were selected
for each column. The first column had light and heavy keys
of propane and butylene, the top column has light and heavy
keys of propylene and propane, and the bottom column has
light and heavy keys of butylene and butane, all respectively
(Figure 6). These key components were again employed in
the FUGK method which led to sensitivity testing in Aspen,
wherein variables of interest (number of column stages, reflux
ratio) were manipulated to identify optimal conditions for
separation between the key components.
2.4.2 Process Benefit
While the C3 -C4 separation scheme does not provide an unex-
pected benefit or improve separation elsewhere in the process,
it provides the requisite final separations to prepare our C3
and C4 products for storage, transport, and retail. Without
this system, these valuable compounds would be wasted in a
low-grade unrefined gasoline mixing component. Concisely,
the inclusion of the C3 -C4 separators marks the finale of the
process bottoms distillation scheme that began in the bottoms
stream from the feed compressors and the effective production
of isolated propylene, propane, butylene, and butane.
2.5 Amine Sweetening System
Before discussing the design of the amine sweetening system,
the authors would be remiss as concerned engineers to miss an
opportunity to discuss process safety. The sulfurous and acidic
components entering the amine scrubber (H2 S and CO2 ) both
pose exposure hazards to employees and the MEA solution
itself is highly corrosive and toxic. Extra consideration must
be taken by the site managers at the plant where our process is
to be installed, and special considerations are further discussed
in Appendix G: Failure Mode Effects
Figure 7. Example amine treatment system sweetening the sour
dry gas feed originating from the absorber.
2.5.1 Design
Like the pressurization scheme, the amine sweetening system
was largely designed based on project specifications [67]. The
return requirement from ExxonMobil stated that a maximum
0.4 mol of combined H2 S and CO2 per mol MEA, which was
used to calculate the amount of MEA (and MEA solution)
needed to remove all of the H2 S and CO2 from the system.
The computed mass flow of water and MEA was input to

an Aspen HYSYS
R
model of a sweetening column (Figure
7), which was iteratively sized using sensitivity testing on
the number of stages in the column and the amount of MEA
entering the system [7, 22, 52, 18, 46, 42].
2.5.2 Process Benefit
The MEA process must be included in this system to meet
product specifications in terms of contaminant content. The
removal of the acidic and sulfurous compounds from the sour
dry gas results in a valuable sweetened fuel gas, ready for
use. Otherwise, the light ends would be contaminated well
beyond acceptable levels of H2 S and CO2 for consumption
and rendered unusable.
2.6 Enriching Column
Only the finalized design includes the enriching column, the
reasons for which are discussed at the beginning of 4. Final-
ized Design. Otherwise, the alternative and finalized designs
employ the same set of unit operations to achieve acceptable
product separations.
Figure 8. Example enriching column indicating the removal of
heavy ends from sour dry gas in a recycle stream to the stripping
column and subsequently lighter distillate sour gas exiting the top of
the enricher.
2.6.1 Design
The enriching column works in tandem with the absorber-
stripping column scheme to increase the propane flux through
the recycle loop and enhance the overall product recovery of
the process (Figure 8, Figure 9).
The sour gas from the top of the absorber is fed to the
enricher, the enricher separates and removes the light gases
and contaminants for amine treatment, and heavier sub-C2
components from the enricher are recycled to the top of the
stripping column. This general scheme was constructed by
increasing the stripping reflux while reducing the distillate
rate of the enricher until the model converged with all of the
H2 S and CO2 in the dry gas and all of the C3 s and heavier in
the recycle to the stripper.
2.6.2 Process Benefit
Adding an enriching column to the absorber-stripper scheme
improves the overall recovery of heavier components exit-
ing the bottom of the stripping column and improves upon
the original recycle loop design to further increase mass flux
Coker Gas Plant Design — 7/28
Figure 9. Example absorber-stripping column scheme with
enriching column inlet.
through the loop and enhance product separations. The ef-
fects of excluding the enriching column from the design are
apparent in the analysis of our alternative model.
2.7 Miscellaneous Units
Miscellaneous units include heat exchangers, pumps, and
turbines, as these units were deemed interstitial to the main
processes of the model.
2.7.1 Design
Heat exchangers were sized in Aspen based on the cooling
water return specifications and the assumed propane refrig-
eration efficiency [44]. Centrifugal pumps were sized based
on pump head and unit flowrate, and turbines were sized in
accordance with relevant literature [65].
2.7.2 Process Benefit
These units, although not incorporated in the major units
described in 2.1-2.6, were of great importance in the simu-
lation for meeting product specifications and utility return
requirements. They were also instrumental in achieving the
temperature and pressure conditions driving the separations
in downstream unit operations.
2.8 Other Analyses
In the course of developing this report, several manual anal-
yses were performed to determine the economic viability,
environmental impact, and safety considerations.
2.8.1 Economic Analysis
The development of an economic model began with costing
of individual process units. This process is detailed in Process
Unit Costing. The total cost of engineered equipment (pur-
chased and delivered) is calculated as the sum of individual
equipment costs. This value is multiplied by a number of
percentage factors in order to determine an estimate for total
depreciable capital (CT DC ). CT DC is combined with estimates

for cost of land, startup cost (Cstart ), and cost of royalties
(Croyal ) to estimate total permanent investment (CT PI ).
CT PI = CT DC +Croyal +Cstart
The cost of startup was selected as 2 % of CT DC based on an
assumption of skilled operators. The cost of royalties was
selected at 0.5% of the cost of sales to account for any patent
leasing or royalties to land owners. Note the cost of land was
negated due to this plant’s classification as a addition to an
existing project.
Working capital (CWC ) is also estimated as a percentage of
CT DC (2.5%). This value includes spare parts and maintenance
as well as costs of refrigerant and utility stores. Fortunately,
the feed is sourced from an earlier section of the same plant
and does not need to be included in the working capital.
With CT PI and CWC characterized, the next step is an esti-
mation of operating costs (COM). In our analysis, this value
was split into fixed and variable costs. Fixed costs can be bro-
ken down into five categories: operations (O), maintenance
(M), operating overhead, depreciation (D), and general ex-
penses (GE). [65] presents an excellent overview of these
categories and methods for their calculation. Depreciation
was calculated based on a MACRS-10 year schedule. The
plant is classified under IRS class 13.3 for petroleum refiner-
ies. Fixed costs are distinct from variable costs in that they do
not scale with production capacity. Since there is no cost of
raw materials for this process, variable costs only include the
price of utilities.
With estimates for investment costs and operating costs in
hand, the analysis shifts to macro scale economic concepts.
Potential investors are often interested in metrics such as
cash flows, net present value (NPV ), investor rate of return
(IRR), and payback period (PBP). Cash flows (and discounted
cash flows) serve as a basis for many of these calculations.
However, in order to determine cash flows, after tax earnings
(AT E) must first be calculated:
AT E = (1 − t) × (S + D −COM)
where t is the tax rate. Cash flow for the year in question can
then be found according to:
Cash Flow = CT DC +CWC +CStart +Croyal + D + AT E
Before moving into other profitability metrics, it is im-
portant to select a minimum attractive rate of return (MARR).
The present analysis relies on a MARR of 10%. The MARR
is used to determine discounted cashflows for each year to
represent the present value of that cashflow.
After cash flows have been discounted, they can be summed
for each year in order to compute the net present value (NPV )
of a project. When the sum of the discounted cash flows, both
negative (costs) and positive (revenues), is equivalent to the
net present value of a project, the internal rate of return is
computed. The internal rate of return (IRR) is then calculated
as the rate of return for the discounted cash flows. If the IRR,
otherwise known as the discounted cash flows rate of return
(DCFRR) is greater than the MARR, the project is profitable.
Coker Gas Plant Design — 8/28
Payback period is calculated as the ratio of capital invest-
ment to yearly after tax earnings assuming a median deprecia-
tion rate:
CT DC
PBP =
AT E
Return on investment is calculated as net earnings over total
capital investment divided by total capital investment:
AT E
ROI =
CT DC +CWC
2.8.2 Environmental Survey
Biwer’s method was implemented for our environmental sur-
vey [17]. Briefly, component property data is collected from
material safety data sheets and other resources. Impact cat-
egories are assigned to the components based on their avail-
ability, complexity of synthesis, critical material, chronic and
acute toxicity, endocrine disruption potential, thermal risks,
and global warming potential. For our analysis, a ranking
system of 1-3 was chosen, where 3 corresponds to the most
dangerous and harmful components and 1 corresponds to rel-
atively harmless components. The rankings in the impact
categories are multiplied on a per-component basis to produce
a component impact product. These products are summed for
the entire process and component impact fractions are calcu-
lated by dividing the impact product by the process impact for
each component. This value is multiplied by the mass fraction
of the species in each stream, which calculates a weighted
environmental factor. These weighted factors are summed and
factor fractions are calculated per species. The environmen-
tal factor fractions are compared, and species with fractions
greater than 0.25 were deemed potential hazards.
Upon identifying potential hazards, MSDS data was re-
consulted and the species of interest were eliminated or con-
firmed as hazards based on the hazard data in the MSDS.
2.8.3 Safety Study
Process safety was assessed by failure mode effects analysis
(FMEA) [19]. Failure modes were analyzed based on the most
probable points of failure in our system; modes were assessed
by both mechanism of failure and effect and impact on process
and surroundings. The three primary concerns of FMEA are
(1) probability, (2) seriousness, and (3) chance of detection
by operator. These categories were assigned rankings 1-5, 5
being most severe, and the product of these categories per
failure was taken as the FME ranking on a 1-125 scale (again,
125 being most severe). Analyzed failure modes were then
appended with process control systems specifically designed
to mitigate the weaknesses inherent to each failure mode.
3. Alternative Model
3.1 Design
The alternative model was a preliminary final draft (originally
created for the purposes of a review presentation) that incor-
porated every unit in 2. Aspen Design except for the enriching
 Coker Gas Plant Design — 9/28

Refinery
Fuel Gas
Clean MEA
Solution
MEA-1
A-1 Dirty MEA
Feed Solution

Quench Water
Clean MEA Refinery Grade
Solution Propylene

HX-1 C-1 HX-2

D-1 C-2 D-2 MEA-2
Dirty MEA
Free Water Solution
P-1

F-3
C-3 HX-5

S-1
Commercial Grade
Propane

HX-7 F-2
HX-4
P-2
F-1
HX-3
T-1

Refinery Grade
Butylene

F-4
P-3 HX-6
T-2

Gasoline Mixing Commercial Grade

Component Butane

Figure 10. PFD of alternative model incorporating multistage compression, an absorber-stripping column scheme, a C3 -C4 separator train,
and amine sweetening systems.

column. In brief, the feed stream is fed through two pressur-
ization cycles (C-1, C-2) to free water from the system and
separate heavy oils and lighter vapors to feed to the top of the
absorber (A-1) and stripper (S-1) and bottom of A-1, respec-
tively. A-1 recycles some rich oil to the decanter feeding the
bottom of the absorber with wet gas and the top of the stripper
with rich oil (D-2) to improve the basic light-heavy separation.
The product originating from the top of the absorber continues
to MEA treatment (MEA-1) and exits the process as refinery
fuel gas. In S-1, the light ends are recycled to D-2 so they
can re-enter the recycle loop created by the (A-1, S-1, D-2)
scheme and improve the overall recovery of the products. The
bottoms stream from S-1 is fed to a debutanizer (F-1) to re-
move the gasoline mixing component in the heavy ends (some
of which is recycled to the top of A-1), while F-1’s light ends
continue to a three-column C3 -C4 separation system (F-2, F-3,
F-4) that results in four product streams of (from lightest to
heaviest) refinery grade propylene, commercial grade propane,
refinery grade butylene, and commercial grade butane. The re-
finery grade propylene requires additional amine sweetening
(MEA-2), as the initial absorber-separating column scheme
does not completely partition H2 S and CO2 into the light ends
exiting the absorber (Figure 10).
3.1.1 Pressurization Scheme
The feed undergoes isobaric cooling from 150◦ F to 108.5◦ F
(HX-1) before isentropic compression from 20 psig to 118.7
psig, resulting in a temperature increase to nearly 300◦ F (C-1).
Before entering the first decanter, this stream undergoes a
second round of isobaric cooling from 300◦ F to 194◦ F (HX-
2), which results in a mixed-phase stream with a mass vapor
fraction of 0.98. The decanter (D-1) removes almost 14,000
lb/hr of water, 9,790 lb/hr of condensed oil, and more than
208,260 lb/hr of vapor; all of these streams exit the decanter
at 112◦ F and 118 psig. The vapor product from D-1 enters
another compressor (C-2), increasing its pressure to 480 psig
(recall that our capability testing gave a maximum two-cycle
compression of 489 psig) and its temperature to 312◦ F. The
outlet from C-2 mixes with the distillate from the stripping

Table 1. Product recoveries, purities, and limited component data for alternative model.
Flow rate in feed (lb/hr) Flow rate in product (lb/hr)
Refinery fuel gas 79,516
Refinery grade propylene 9,538
Commercial grade propane 27,094
Refinery grade butylene 14,270
Commercial grade butane 18,652
Gasoline mixing component 25,058
column (S-1) and the bottoms from the absorber (A-1) before
entering the final decanter (D-2), which cools its three outlets
to 95◦ F. The condensed water and oil products exit at 50,802
lb/hr and 253,214 lb/hr, respectively, and the light vapors
from D-2 flow to the bottom of A-1 at 179,197 lb/hr. The
condensed oil product from D-1 travels through a pump (P-1),
which increases the conditions of the stream to 129◦ F and 480
psig. Finally, this pressurized liquid stream is mixed with the
oil from D-2 before entering the top of S-1.
3.1.2 Absorber and Stripping Column
The top of A-1 is fed by a recycled gasoline mixing component
from the bottom of the debutanizer (F-1), which is pumped
to 465 psig and 85◦ F; the bottom of A-1 is fed by wet gas
from D-2. The absorber itself is made up of 25 stages and
operates at the pressure of its top feed with a ∆P across the
column of 15 psi. Vapor leaves the top of the absorber at just
over 465 psig and 139◦ F at a rate of 135,096 lb/hr to an amine
sweetening system (MEA-1). Liquid from the bottom of the
absorber at 170,528 lb/hr, 115◦ F, and 480 psig is recycled into
the mixer feeding D-2, the condensed oil of which exits to
S-1. S-1 is operated at 480 psig and contains 40 stages, with
each stage experiencing a ∆P of 1 psi. The bottoms-to-feed
ratio in this column is 0.8 (mass basis). Two streams exit S-1:
The vapor distillate exits at 131◦ F, 480 psig, and 54,664 lb/hr
to the mixer feeding D-2; the liquid bottoms enter a turbine
(T-1) at 317◦ F, 519 psig, and 218,656 lb/hr before continuing
on to F-1 for debutanization.
3.1.3 Debutanizer
T-1 and a cooler (HX-3) combine to feed the S-1 bottoms to
F-1 at 100◦ F and 180 psig. F-1 contains 25 stages and runs at
a distillate rate of 70,548 lb/hr. The feed is located in stage
7, the condenser operates at 175 psig, and the stage ∆P is 1
psi. The liquid distillate from F-1 exits at 87◦ F and 175 psig
to a pump and cooler (P-2, HX-4) before entering the C3 -C4
separator (F-2); the bottoms from F-1 are refluxed at a rate
of 2.5. Product bottoms leaving the system at 148,108 lb/hr,
395◦ F, and 199 psig are partially recycled to the top of A-1
at a rate of 126,404 lb/hr. The recycled stream is pressurized
and cooled to 480 psig and 85◦ F while the remaining gasoline
mixing component leaves the process as a refined product.
3.1.4 C3 -C4 Separators
P-2 and HX-4 result in the pressurization and heating of the
F-2 inlet to 300 psig and 100◦ F. F-2 is a 25-stage column with
Coker Gas Plant Design — 10/28
Recovery Purity (Limited component)
73,607 0.93 0.57
4,121 0.43 0.70
16,953 0.63 0.94 (0.06)
14,072 0.99 0.70
12,802 0.69 0.97 (0.00)
21,459 0.86 0.99
a partial vapor condenser and a distillate to feed ratio of 0.53.
The stage ∆P is 1 psi, the column operating pressure is 300 psi,
and the feed enters at stage 15. Vapor C3 s are distilled from
F-2 at 37,390 lb/hr, 102◦ F, and 285 psig while the C4 bottoms
are refluxed at a ratio of 2 (the exiting C4 s leave F-2 at 235◦ F,
309 psig, and 33,157 lb/hr. These streams enter the propane
splitter (F-3) and the butane splitter (F-4), respectively.
Before entering F-3, the inlet stream goes through a com-
pressor and cooler (C-3, HX-5) to achieve a discharge tem-
perature and pressure of 62◦ F and 333 psig. F-3 consists of
125 stages (fed at stage 100), a partial vapor condenser, an
operating pressure of 231 psig, and a column ∆P of 22 psi.
The distillate-to-feed ratio is 0.57 and the reflux ratio is 6.
The distillate and bottoms exit at 216 psig, 55◦ F, 18,653 lb/hr
and 238 psig ,122◦ F, 18,737 lb/hr, respectively. The distillate
continues to an amine sweetening system (MEA-2) while the
bottoms exit as commercial grade propane. The feed to F-4 is
throttled in T-2 to 200 psig and cooled in HX-6 to 100◦ F. F-4
is made up of 150 stages (fed at stage 35), a total condenser,
an operating pressure of 90 psig, a column ∆P of 100 psig, a
distillate rate of 20,000 lb/hr, and a reflux ratio of 15 (which
results in a bottoms flow of 13,157 lb/hr). The light butylene
product exits at 90 psig and 129◦ F while the heavier butane
leaves the system at 205◦ F and 190 psig.
3.1.5 Amine Sweetening Systems
MEA-1 is fed by the A-1 distillate while MEA-2 is fed by the
F-3 distillate. MEA-1 operates at 0 psig, includes 4 stages,
uses a column ∆P of 5 psi, and runs MEA solution at a flowrate
of 1,214 gal/min. In contrast, MEA-2 operates at Ptop = 200
psig, Pbottom = 203 psig, includes 4 stages, and runs MEA
solution at a flowrate of 522 gal/min. MEA solution is fed at
300 psig and 100◦ F to both systems, which exits as dirty MEA
streams. Sweetened refinery fuel gas and propylene leave the
system at 127,537 lb/hr and 5,213 lb/hr, respectively.
3.2 Product Recoveries
Recovery data is compiled in Table 1: Alternative Model Prod-
uct Recovery and Purity Data. In summary, product purity
specifications were met such that we produced refinery grade
butylene and propylene and commercial grade butane and
propane within the H2 S and limited component parameters
(e.g. the butylene content in the product butane stream must
be less than 0.1). Our largest product stream is the refinery
fuel gas, with nearly 75,000 lb/hr of flow; the gasoline mixing
component leaves the process at approximately 20,000 lb/hr,

and the remaining components (with the exception of refin-
ery grade propylene, which exits at just over 4,000 lb/hr) all
depart from the system between 12,000-17,000 lb/hr.
3.3 Utilities Summary
Utilities required by the alternative process are notably more
expensive than those in the finalized model because of the
inclusion of a secondary MEA scrubber in the alternative
model, which raises the sweetening expense to nearly half the
total utilities cost (Table 2.)
Table 2. Utility requirements and costs for alternative model.
Utility Hourly requirement Annual expense
Cooling Water 1,282 klb/hr $2,051,200
High Pressure Steam - -
Low Pressure Steam 250 klb/hr $7,989,488
Electricity 49 MWH/hr $19,712,553
MEA Solution 3,600 gal/hr $28,800,000
Total $58,553,241
3.4 Economic Analysis
The total depreciable capital (TDC) was calculated as $318M
while the working capital (WC) was calculated as $7,943,508.
The cost of startup was calculated as $6,354,806. The cost of
royalties was calculated as $1,448,477. The variable operating
costs were calculated as $70,554,165. The fixed operating
costs were calculated as $49,951,619. The total permanent
investment (TPI) was calculated as $33,998,212. The tax rate
used was 39% while the inflation rate used was 2%. The
return on investment (ROI), which is based on the tax rate, de-
preciation rate at year 5 of the 10 years MACRS depreciation
scheme, gross revenues, costs, total depreciable capital and
working capital, was calculated as 26.20%. This is relatively
large ROI, meaning the investment gains compare favorably
to the investment costs. The payback period (PBP), which
is based on the tax rate, depreciation rate at year 5 of the 10
years MACRS depreciation scheme, gross revenues, costs,
and total depreciable capital, was calculated as 4.00 years.
A maximum PBP of 4 years is typical for chemical plants;
thus the calculated PBP is acceptable. The internal rate of
return (IRR), otherwise known as the discounted cash flow
rate of return (DCFRR), was calculated as 23.1%. This is
assuming a minimum attractive rate of return (MARR) of
10% for the refineries investors. Since our IRR is greater than
the MARR, the refinery is considered profitable. In addition,
the net present value (NPV) of the plant was calculated as
$511,210,570. This large net present value scales with the
large costs associated with building a coker gas refinery.
4. Finalized Model
4.1 Design
The finalized model developed out of attempts to improve the
alternative design and limit the overall use of amine sweeten-
ing. Similarly to the alternative model, the feed is pressurized
Coker Gas Plant Design — 11/28
in a two-cycle system and taken to the absorber and strip-
ping column. In the finalized design, however, this absorber-
stripper scheme includes the enriching column as an addi-
tional unit in the recycle loop, which dramatically improves
the overall separation of sub-C2 compounds from the light
fuel gas and contaminants (which, as in the alternative design,
continue on to amine treatment before exiting the process
as dirty MEA solution and refinery fuel gas). The bottoms
distillation train is identical to the system through the C3 -C4
separator in the alternative design, but the distillate stream
from the C3 column no longer requires additional MEA treat-
ment as the enriching column improves the H2 S and CO2
partitioning to process specifications in the bottoms streams.
This allows for a focused effort in the single MEA scrubber
to remove essentially all of the system contaminants at a far
smaller utility cost, which resulted in this design’s selection
over the alternative model. Furthermore, the enhancements to
the recycle loop made by the addition of the enriching column
result in increased recovery of propane and improved quality
of the propylene product (upgrading it from refinery-grade to
chemical-grade, which increases propylene product revenues
by $0.15/lb).
4.1.1 Pressurization Scheme
The feed undergoes isobaric cooling from 150◦ F to 111◦ F
(HX-1) before isentropic compression from 20 psig to 118.7
psig, resulting in a temperature increase to nearly 300◦ F (C-1).
Before entering the first decanter, this stream undergoes a
second round of isobaric cooling from 300◦ F to 95◦ F (HX-
2), which results in a mixed-phase stream with a mass vapor
fraction of 0.9. The decanter (D-1) removes almost 16,000
lb/hr of water, 36,600 lb/hr of condensed oil, and almost
180,000 lb/hr of vapor; all of these streams exit the decanter
at 50◦ F and 118 psig. The vapor product from D-1 enters
another compressor (C-2), increasing its pressure to 480 psig
(recall that our capability testing gave a maximum two-cycle
compression of 489 psig) and its temperature to 242◦ F. The
outlet from C-2 mixes with the distillate from the stripping
column (S-1) and the bottoms from the absorber (A-1) before
entering the final decanter (D-2), which cools its three outlets
to 95◦ F. The condensed water and oil products exit and the
light vapors from D-2 flow to the bottom of A-1 at 304,714
lb/hr. The condensed oil product from D-1 travels through a
pump (P-1), which increases the pressure of the stream to 480
psig. Finally, this pressurized liquid stream is mixed with the
oil from D-2 before entering the top of S-1.
4.1.2 Absorber, Stripping and Enriching Columns
The top of A-1 is fed by a recycled gasoline mixing component
from the bottom of the debutanizer (F-1), which is pumped
to 465 psig; the bottom of A-1 is fed by wet gas from D-2.
The absorber itself is made up of 25 stages and operates at
the pressure of its top feed with a ∆P across the column of
10 psi. Vapor leaves the top of the absorber at 235,333 lb/hr
to the enriching column (E-1). E-1 separates sour refinery
fuel gas from heavier ends, which are recycled to the top of
 Coker Gas Plant Design — 12/28

Refinery
Fuel Gas
Clean MEA
Solution
MEA-1
A-1 Dirty MEA
Feed Solution

E-1

HX-1 C-1 HX-2

D-1 C-2 HX-3 D-2

Free Water
P-1

P-2
Chemical Grade
Propylene
F-3
C-3 HX-7

S-1 Commercial Grade

Propane
HX-9

HX-5 F-2
HX-6
P-4
F-1
HX-4
T-1

Refinery Grade
F-4 Butylene
HX-8
P-3 T-2

HX-11
Gasoline Mixing Commercial Grade
Component Butane
HX-10

Figure 11. PFD of finalized model incorporating multistage compression, an absorber-stripper-enriching column scheme, a C3 -C4 separator
train, and an amine sweetening system.

Table 3. Product recoveries, purities, and limited component data for finalized model.
Flow rate in feed (lb/hr) Flow rate in product (lb/hr) Recovery Purity (Limited component)
Refinery fuel gas 79,516 79,517 1.00 0.65
Chemical grade propylene 9,538 8,786 0.92 0.94
Commercial grade propane 27,094 26,104 0.96 0.98 (0.01)
Refinery grade butylene 14,270 13,098 0.92 0.73
Commercial grade butane 18,652 13,759 0.74 0.91 (0.06)
Gasoline mixing component 25,058 24,912 0.99 0.99

S-1 through P-2, which pressurizes the bottoms to 480 psig.
E-1 consists of 20 stages, a partial vapor condenser, a stage
∆P of 1 psi, an operating pressure of 460 psig, and 122,000
lb/hr of distillate flow. Liquid from the bottom of the absorber
at 219,407 lb/hr is recycled into the mixer feeding D-2, the
condensed oil of which exits to S-1. S-1 is operated at 480
psig and contains 60 stages, with each stage experiencing a
∆P of 1 psi. The bottoms rate in this column is 244,291 lb/hr.
Two streams exit S-1: The vapor distillate exits at 280,807
lb/hr to the mixer feeding D-2; the liquid bottoms enter a
turbine (T-1) at 244,291 lb/hr before continuing on to F-1 for
debutanization.
4.1.3 Debutanizer
T-1 and a cooler (HX-4) combine to feed the S-1 bottoms to
F-1 at 100◦ F and 180 psig. F-1 contains 25 stages and runs
at a distillate rate of 69,000 lb/hr and 175 psig. The feed is
located in stage 7, the top of the column includes a complete
condenser, and the stage ∆P is 1 psi. The liquid distillate
from F-1 exits at 87◦ F and 175 psig to a pump and cooler
(P-4, HX-6) before entering the C3 -C4 separator (F-2); the

bottoms from F-1 are refluxed at a rate of 2. Product bottoms
leaving the system at 175,291 lb/hr are partially recycled to
the top of A-1 at a rate of 126,404 lb/hr. The recycled stream
is pressurized and cooled to 480 psig and 100◦ F while the
remaining gasoline mixing component leaves the process as a
refined product.
4.1.4 C3 -C4 Separators
P-4 and HX-6 result in the pressurization and heating of the
F-2 inlet to 300 psig and 100◦ F. F-2 is a 25-stage column with
a partial vapor condenser and a distillate rate of 35,880 lb/hr.
The stage ∆P is 1 psi, the column operating pressure is 300
psi, and the feed enters at stage 15. Vapor C3 s are distilled
from F-2 at 35,880 lb/hr while the C4 bottoms are refluxed at
a ratio of 2.5 (the exiting C4 s leave F-2 at 33,120 lb/hr. These
streams enter the propane splitter (F-3) and the butane splitter
(F-4), respectively.
Before entering F-3, the inlet stream goes through a com-
pressor and cooler (C-3, HX-7) to achieve a discharge tem-
perature and pressure of 62◦ F and 333 psig. F-3 consists of
200 stages (fed at stage 100), a partial vapor condenser, an
operating pressure of 231 psig, and a column ∆P of 22 psi.
The distillate rate is 9,323 lb/hr and the reflux ratio is 25. The
distillate and bottoms exit at 9,329 lb/hr and 26,551 lb/hr,
respectively. The distillate exits as chemical grade propylene
while the bottoms exit as commercial grade propane. The
feed to F-4 is throttled in T-2 to 200 psig and cooled in HX-8
to 100◦ F. F-4 is made up of 150 stages (fed at stage 35), a
total condenser, an operating pressure of 90 psig, a column
∆P of 100 psig, a distillate rate of 18,000 lb/hr refinery grade
butylene, and a reflux ratio of 15 (which results in a bottoms
flow of 15,120 lb/hr commercial grade butane).
4.1.5 Amine Sweetening Systems
MEA-1 is fed by the E-1 distillate. MEA-1 operates at 10
psig, includes 4 stages, uses a column ∆P of 5 psi, and runs
MEA solution at a flowrate of 1,761 gal/min (a specified
lifetime of 25 recycles allows for a feed of only 68 gal/min
MEA solution). MEA solution is fed at 300 psig and 100◦ F
to the system, which exits as a dirty MEA stream. Sweetened
refinery fuel gas leaves the system at roughly 122,000 lb/hr.
4.2 Product Recoveries
Recovery data is compiled in Table 3: Finalized Model Prod-
uct Recovery and Purity Data. In summary, product purity
specifications were met such that we produced refinery grade
butylene, chemical grade propylene, and commercial grade
butane and propane within the H2 S and limited component
parameters (e.g. the butylene content in the product butane
stream must be less than 0.1). Our largest product stream is
still the refinery fuel gas, with nearly 80,000 lb/hr of flow (an
increase of approximately 5,000 lb/hr RFG flow); the gaso-
line mixing component leaves the process at approximately
25,000 lb/hr (also an increase of 5,000 lb/hr), and the re-
maining components show marked improvements in product
flow and recovery. The propylene is enhanced to chemical-
Coker Gas Plant Design — 13/28
grade and exits the process at 8,800 lb/hr, effectively doubling
the alternative model’s propylene outlet. Commercial-grade
propane leaves at over 26,000 lb/hr, a hugely improved re-
covery (a 33% increase compared to the alternative model).
Butane and butylene outlets are essentially unchanged, which
presents an opportunity to further improve on this design in
later studies.
4.3 Utilities Summary
Table 4. Utility requirements and costs for final model.
Utility Hourly requirement Annual expense
Cooling Water 4,177 klb/hr $6,683,298
High Pressure Steam 29 klb/hr $1,411,811
Low Pressure Steam 57 klb/hr $1,834,006
Electricity 70 MWH/hr $27,876,384
MEA Solution 4,064 gal/hr $32,513,477
Total $70,318,975
4.3.1 Energy Analysis
In an effort to save utility costs in the plant, an attempt was
made to perform heat integration via a Heat Exchanger Net-
work (HEN). The concept of a HEN pairs streams that are
in temperature ranges allowing for heat transfer to occur be-
tween the two process streams, thereby circumventing the
use of expensive utilities. This can provide considerable cost
savings in long-term processes. Normally, this methodology
involves the construction of composite curves, along with the
declaration of a minimum approach temperature, the identi-
fication of a pinch point, and a shift in the cold composite
curve so that the pinch point has a gap equal to the minimum
approach temperature. This method allows energy savings to
be calculated from the overlap created in the composite curves.
From this, a HEN can be determined through the use of heat
capacities of the process streams along with various heuristics.
This process can be done by hand but was automated after
considering the extraneous nature of this portion of the project
and time concerns.
4.4 Economic Analysis
The total depreciable capital (TDC) was calculated as $318.5M
while the working capital (WC) was calculated as $7,961,763.
The cost of startup was calculated as $6,369,410. The cost of
royalties was calculated as $1,463,910. The variable operating
costs were calculated as $70,318,975. The fixed operating
costs were calculated as $49,428,718. The total permanent
investment (TPI) was calculated as $34,076,344. The tax rate
used was 39% while the inflation rate used was 2%. The
return on investment (ROI), which is based on the tax rate, de-
preciation rate at year 5 of the 10 years MACRS depreciation
scheme, gross revenues, costs, total depreciable capital and
working capital, was calculated as 26.85%. This is relatively
large ROI, meaning the investment gains compare favorably
to the investment costs. The payback period (PBP), which
is based on the tax rate, depreciation rate at year 5 of the 10

years MACRS depreciation scheme, gross revenues, costs,
and total depreciable capital, was calculated as 4.00 years.
A maximum PBP of 4 years is typical for chemical plants;
thus the calculated PBP is acceptable. The internal rate of
return (IRR), otherwise known as the discounted cash flow
rate of return (DCFRR), was calculated as 23.60%. This is
assuming a minimum attractive rate of return (MARR) of
10% for the refineries investors. Since our IRR is greater than
the MARR, the refinery is considered profitable. In addition,
the net present value (NPV) of the plant was calculated as
$528,333,369. This large net present value scales with the
large costs associated with building a coker gas refinery.
The internal rate of return and net present value are cal-
culated in conjunction. Thus, their sensitivity plots illustrate
responsiveness to similar values. IRR and NPV, as seen from
their corresponding plots, are most sensitive to sales and op-
erating costs. Furthermore, IRR and NPV are relatively in-
sensitive to the total depreciable capital, which accounts for
the cost of equipment. Therefore, the profitability of this
project could be improved by adding additional equipment
to improve product yields and accordingly sales. However,
with the addition of this equipment, the cost of utilities must
not change dramatically, or the cost of utilities in addition to
the cost of the extra equipment will negate the potential profit
from improved purities or yields, and sales. The sensitivity of
the IRR and NPV to sales is of particular concern due to the
volatility of the petroleum market.
The goal of the primary design was to further purify
streams at the cost of additional capital investment. This
was successful in part, but the implementation of more capital
came with an unexpected increase in operating costs. The
combined increase of costs and sales in the primary design
slightly improves key economic metrics and simplifies the
process slightly. IRR increases by roughly 0.5% for the 40
year plant lifetime. In either case, the rate of return is more
than acceptable and either plant is viable.
4.5 Environmental and Safety Survey
With its relatively high H2 S and CO2 content, the coking gas
refinery scheme poses potential hazards to its surrounding
operators, community, and environment. For this reason, a
modified failure modes analysis and an abridged environmen-
tal survey by Biwer’s method were deemed necessary for the
thorough completion of our design scheme.
4.5.1 Modified Failure Mode Effects Analysis
Please note that our full FMEA table and updated PFD are in
Appendix G: FMEA.
FMEA determined that many process units required con-
trol systems to mitigate risk of failure and potentially harmful
process operations [6, 25, ?] . Heat exchangers were analyzed
based on most likely mode of failure: Most heat exchangers
had an FMEA factor ranging from 10 to 20, 30 being an ex-
treme case. To mitigate, HX-1 and HX-2 on the process flow
diagram (Appendix G: FMEA were fitted with temperature
and flow feedback control systems. Similarly, temperature
Coker Gas Plant Design — 14/28
feedback controls were implemented for HX-5, 7, 8, 9, 10, and
11 on the process flow diagram. All heat exchangers handling
large concentrations of hydrogen sulfide are constructed with
higher grade stainless steel (higher weight percent chromium)
to mitigate possibility of sulfidation. Compressors were also
analyzed based on most likely mode of failure: Compressors
had an average FMEA factor of 20 (medium), 30 (medium
high) in more extreme cases. Conveniently, corrective action
for compressors 1 and 2 on the process flow diagram is ac-
counted for by the control systems implemented in HX-1 and
HX-2. Corrective action for compressor 3 was accounted for
by a temperature control system, which was also implemented
for the condenser and reboiler of column F2. Decanters 1 and
2 were primarily at risk of sulfidation as passive equipment,
and higher grade stainless steels were correspondingly used
in the construction of these units to mitigate risk. The largest
mode of failure for each pump was insufficient flow because
of its negative effects on the remainder of the process. Each
pump was fitted with a volumetric feedback control system.
Both the absorber and the stripping column had similar modes
of failure (sulfidation and mechanical failure) and relative
FMEA factors, approximately 60 (high). Both unit processes
were constructed of higher grade stainless steels to mitigate
risk of sulfidation, and each column was equipped with real-
time pressure sensors. Each separation column in the separa-
tion train had similar probable modes of failure, which were
either mechanical failure or poor separation. To minimize
this possibility, these columns were each correspondingly fit-
ted with temperature and pressure feedback controls for their
respective reboilers and condensers.
4.5.2 Abridged Biwer’s Method
Biwer’s Chemical Process Environmental Analysis revealed
only a few environmental “hotspots” (i.e. species of concern
within process streams that needed to be accounted for ac-
cordingly) [36, 17]. For overall process flows, each species’
respective weighted-environmental index factor (EIMF) was
investigated. The primary environmental hotspot in the inlet
stream was hydrogen sulfide, with an EIMF of approximately
0.22. This hotspot was accounted for by constructing ini-
tial heat exchangers and compressors that process the feed
through higher grade stainless steel equipment to mitigate
the risk of sulfidation. Similarly, the refinery fuel gas outlet
stream had a large hotspot with respect to hydrogen sulfide
(EIMF = 0.392). This was mitigated by both incorporating
an amine sweetening unit to reduce hydrogen sulfide content,
as well as constructing the unit processes of a higher grade
stainless steel to mitigate the risk of sulfidation.
The chemical grade propylene outlet stream had an envi-
ronmental hotspot with respect to propylene itself, but it was
determined that propylene was not a primary risk source after
consulting material safety data resources. The commercial
grade propane outlet stream had an environmental hotspot
with respect to propane itself, and it was similarly determined
that propane was not a risk source. The refinery grade buty-
lene outlet stream had an environmental hotspot with respect
 Coker Gas Plant Design — 15/28

to 1-butene which was also deemed non-hazardous. How- 5. Conclusions

ever, isobutane has a relatively significant EIMF factor, and
In this report, we rigorously modeled alternative and finalized
handling and processing of this product stream should be as-
designs of a coking gas separations process. We explored the
sessed carefully. The commercial grade butane outlet stream
plant equipment, economics, utility requirements, recoveries,
had an environmental hotspot with respect to the product
and other relevant aspects of refinery design, and ultimately
itself which was deemed not a source of risk. The gaso-
achieved a model with competitive economic statistics and
line blending component outlet stream had an environmen-
efficient separation of the process feed.
tal hotspot with respect to hexanes which was deemed not
a source of risk. Lastly, there is a significant environmen- 5.1 Cost-Benefit Analysis
tal hotspot with respect to the MEA amine sweetening so-
5.1.1 Costs
lution in its respective inlet and outlet streams. Handling
• Feed pricing
and processing of these inlet and outlet MEA streams is as-
• Operating costs
sumed to be performed by an on-site service included in the
• Total capital investment
scope of the larger project (installation of the coking unit)
• Working capital
[17, 8, 10, 9, 11, 12, 13, 20, 26, 27, 37, 40, 39, 50, 49, 48, 47,
51, 58, 57, 63, 53, 56, 55, 54, 62, 64, 59, 61, 60, 66]. 5.1.2 Benefits
• Competitive NPV
4.5.3 MEA Considerations • Competitive ROI
• Competitive IRR
Beyond the health considerations presented in our Biwer’s
• Competitive annual cash flows
analysis, excessive temperatures in the gas sweetening pro-
cess can potentially cause an acid gas breakout and degrade or
5.2 SWOT Analysis
change the pH of the amine scrubbing solution. This can be
5.2.1 Strengths
avoided by adjusting process unit temperatures and reboiler
• Profitability
temperatures to certain thresholds. High inlet velocities in the
• Component recovery
gas sweetening process can cause erosion and/or corrosion by
means of solid particulates or acid gas flashing. Again, this 5.2.2 Weaknesses
can be reconciled by setting the inlet flow velocity and the pip- • Dangerous
ing flow velocity to a certain threshold. Amine concentration • Negative publicity (environmental concerns)
is directly related to the corrosiveness of the MEA solution. • Amine system hazards
MEA strength is in the range of 18-20 weight percent, which • High utility costs
is of little concern in terms of concentration range for amine
sweetening solutions. Acid gas loading is of additional con- 5.2.3 Opportunities
cern because it can be a primary source of acid gas breakout, • Improving butane separations
which can cause significant plant corrosion and health hazards. • Heat integration (utility savings)
This risk can be mitigated by thresholding the acid gas loading • Community outreach
to a recommended ratio. MEA rich amine loading is typically • Capture additional market share
limited to 0.425-0.450 m/m. Sulfidation is a major contrib-
utor to the corrosion of process equipment constructed with 5.2.4 Threats
standard materials, such as carbon steel. This risk is mitigated • Pricing changes
by using higher grade stainless steels, which have a higher • Catastrophic failures
weight percent of chromium, the most important internal or • Feed sourcing issues
metallurgical factor in controlling sulfidic corrosion. • Industry competition

As such, the coking gas refining scheme presented in

4.6 Social Impact Recommendation this report offers an exciting and potentially market-viable
As a refining venture, this project must ensure compliance processing venture to separate light gases from a coking unit.
with industry standards in terms of ethical business tactics
and socially conscious corporate culture [16]. We recommend
that the project add funds for community outreach (vaccina-
tion programs, scholarship and grant funding opportunities),
as well as offering educational programs. Additionally, the
project’s land holdings must comply with the World Bank and
UN guidelines on land acquisition.

Acknowledgments
Erik Magnuson and Andrew Burns were the primary design-
ers of our separation scheme, Galen Roda performed the eco-
nomic analysis, Connor Miksch performed FMEA, environ-
mental analysis, and decanter sizing, and Bruce Kirkpatrick
created the final presentation and this report.
The authors would like to thank Dr. Tom Belval for his
instruction as the CHEN 4530 course manager and Jay West-
cott of ExxonMobil for his guidance in the design process, as
well as ExxonMobil and University of Colorado Boulder for
the opportunity to collaborate on this project.
References
[1] An Industrial i-Butane/n-Butane Fractionator. 1–7.
[2] Delayed Coker Unit (DCU). 1–6.
[3]
Oil Refinery Processes. 1–53.
[4]
Design of Absorber. In Absorber Design. Feb. 2017.
[5] Lean Oil Absorption. In Gas Processing (Feb. 2017).
[6] A DAMS , J. Coking 101: An Introduction to Delayed
Coking. Process Engineering Associates, LLC. (Feb.
2014), 1–26.
[7] A DDINGTON , L., AND N ESS , C. An Evaluation of
General “Rules of Thumb” in Amine Sweetening Unit
Design and Operation. Bryan Research and Engineering,
Inc. (Feb. 2010), 1–18.
[8] A IRGAS USA. Hydrogen. Material Safety Data Sheets
(Jan. 2003), 1–10.
[9] A IRGAS USA. Methane. Material Safety Data Sheets
(Jan. 2003), 1–8.
[10]
A IRGAS USA. Nitrogen. Material Safety Data Sheets
(Dec. 2003), 1–10.
[11]
A IRGAS USA. Ethylene. Material Safety Data Sheets
(Nov. 2015), 1–12.
[12]
A IRGAS USA. Isobutane. Material Safety Data Sheets
(Jan. 2017), 1–11.
[13] A LFA A ESAR. 4,4-Dimethyloctane. Material Safety
Data Sheets (Mar. 2017), 1–5.
[14] A LIABAD , Z. Removal of CO2 and H2S using aqueous
alkanolamine solusions. World Academy of Science 37
(2009), 194–203.
[15] A SPEN P LUS. Physical Property Methods and Models.
aspentech (Jan. 1999), 1–458.
[16] BARCLAYS. Environmental and Social Risk Briefing Oil
& Gas. Barclays Bank PLC (Mar. 2015), 1–26.
[17] B IWER , A., AND H EINZLE , E. Environmental assess-
ment in early process development. Journal of Chemical
Technology & Biotechnology 79, 6 (May 2004), 597–609.
Coker Gas Plant Design — 16/28
[18]
B LAUWHOFF , P., K AMPHUIS , B., AND S WAAIJ , W.
Absorber design in sour natural gas treatment plants: Im-
pact of process variables on operation and economics=
Absorberentwurf für Anlagen zur Behandlung von . . . .
Chemical Engineering . . . 19, 1 (1985), 1–25.
[19] C HEMICAL I NDUSTRY D IGEST. FMEA and FMECA
Basic Concepts. Business Insights Global (Mar. 2013).
[20] C HEM S ERVICE. 2,6-Dimethylheptane . Material Safety
Data Sheets (Sept. 2015), 1–9.
[21] C OLWELL , R. Oil Refinery Processes - A Brief Overview.
Process Engineering Associates, LLC. (Jan. 2009), 1–36.
[22] D RNEVICH , R. F., PAPAVASSILIOU , V., S CALISE , J.,
AND WATWE , R. Cost Effective Technology for Organic
Sulfur Removal from Refinery Fuel Gas . NPRA Meeting
San Antonio TX (Feb. 2011), 1–7.
[23] D UFFY, G. The McCabe Thiele Graphical Design. 1–37.
[24] E LLIS , P. J., AND PAUL , C. A. Tutorial: Delayed Coking
Fundamentals. AIChE Spring National Meeting (Feb.
2017), 1–20.
[25] E MERSON P ROCESS M ANAGEMENT. Delayed Coker.
Process Solutions Guide (Jan. 2008), 1–3.
[26] F ERRELLGAS. Propane. Material Safety Data Sheets
(Apr. 2014), 1–6.
[27] F ISCHER S CIENTIFIC. 3-Methylhexane . Material Safety
Data Sheets (Feb. 2015), 1–7.
[28] F ORBES , J. T. Gas Separation From Hydrogen Contain-
ing Hydrocarbon Effluent. US Patent Office.
[29] G ILLIS , D. B. Selective separation of Heavy Coker Gas
Oil. US Patent Office, 2015.
[30] H OEHN , R. K. Method for operating a steam stripped
fractionation column. US Patent Office.
[31] H OLDEN , C. Jump Start: Delayed Coker Model in Aspen
HYSYS. aspentech (Jan. 2112), 1–14.
[32] H USSAIN , D. Y. Equilibrium Separation Column. In
Chemical Process Technology. Oct. 2011, pp. 1–18.
[33] JAEGER P RODUCTS , I NCORPORATED. Steam Stripping.
Jaeger Products, Inc. (Dec. 2000), 1–18.
[34] J ECHURA , J. Refinery Feedstocks & Products – Prop-
erties & Specifications. Colorado School of Mines (Feb.
2017), 1–124.
[35] J ECHURA , J. Supporting Processes. Colorado School of
Mines (Feb. 2017), 1–71.
[36] K UNZ , R. G. Combustion of Refinery Fuel Gas. In En-
vironmental Calculations A Multimedia Approach. Aug.
2010, pp. 1–14.
[37] L INDA G AS P UERTO R ICO. Propylene. Material Safety
Data Sheets (Sept. 2013), 1–9.
[38] L UYBEN , W. L. Distillation column pressure selection.
Separation and Purification Technology 168, C (Aug.
2016), 62–67.
 Coker Gas Plant Design — 17/28

[39] M ATHESON T RI·G AS. 2-Methylpentane. Material [60] S CIENCE L AB . COM. n-Pentane. Material Safety Data
Safety Data Sheets (Mar. 2009), 1–7. Sheets (May 2013), 1–6.
[40] M ATHESON T RI·G AS. Butyl benzene. Material Safety [61] S CIENCE L AB . COM. Octane. Material Safety Data Sheets
Data Sheets (Mar. 2009), 1–6. (May 2013), 1–5.
[41] [62] S CIENCE L AB . COM. p-Xylene. Material Safety Data
M AUHAR , S. M., BARJAKTAROVIC , B. G., AND S O -
VILJ , M. N. Optimization of Propylene-Propane Distilla- Sheets (May 2013), 1–6.
tion Process. 31st International Conference of the Slovak [63] S CIENCE L AB . COM. Toluene. Material Safety Data
Society of Chemical Engineering (Mar. 2005), 1–5. Sheets (May 2013), 1–6.
[42] M ITRA , S. A technical report on gas sweetening system. [64] S CIENCE L AB . COM. Water. Material Safety Data Sheets
Petrofac Engineering I Ltd. (June 2011), 1–58. (May 2013), 1–5.
[43] NAEEM , M., A L - RABIAH , A. A., AND M UGHEES , W. [65] S EIDER , D. W., L EWIN , D., AND W IDAGDO , S. Prod-
Process Simulation of 1-Butene and N- Butane Separa- uct and Process Design Principles: Synthesis, Analysis,
tion By Extractive Distillation. International Journal of and Evaluation (Third ed.). Wiley & Sons, 2009.
Engineering Research Technology (Apr. 2017), 1–4. [66]
[44]
T ESORO. Butane. Material Safety Data Sheets (July
NARAYANAMURTHY, R. Packaged Refrigerant Based 2012), 1–9.
Energy Storage (RBES) Air Conditioning System. Ice [67]
Energy, Inc., 2005. W ESTCOTT, J. PS7 fr Westcott of ExxonMobil 0117.pdf.
[45]
CHEN Design Prompt (Feb. 2017), 1–8.
O GURA , S., AND S OUMAI , M. Process for separating [68]
highly pure butene-1 or butene-1/isobutene mixture from Y I , Q., F ENG , J., AND L I , W. Y. Optimization and
C4 hydrocarbon fraction. US Patent Office, 1985. efficiency analysis of polygeneration system with coke-
[46]
oven gas and coal gasified gas by Aspen Plus. Fuel 96, C
O LSEN , T. An Oil Refinery Walk-Through. Chemical (June 2012), 131–140.
Engineering Progress (May 2014), 1–8.
[47] P RAXAIR. 1-Butene. Material Safety Data Sheets (Dec.
2016), 1–9.
[48] P RAXAIR. Carbon dioxide. Material Safety Data Sheets
(Oct. 2016), 1–9.
[49] P RAXAIR. Ethane. Material Safety Data Sheets (Oct.
2016), 1–9.
[50] P RAXAIR. Hydrogen Sulfide. Material Safety Data
Sheets (Oct. 2016), 1–10.
[51] R EAGENT W ORLD. 3-Methylheptane. Material Safety
Data Sheets (Aug. 2011), 1–2.
[52] R EMESAT, D., AND B ESHARA , M. Debottlenecking a
refinery fuel gas absorber. PTQ Revamps 37, 1 (1996),
69.
[53] S CIENCE L AB . COM. Benzene. Material Safety Data
Sheets (May 2013), 1–6.
[54] S CIENCE L AB . COM. Cumene. Material Safety Data
Sheets (May 2013), 1–5.
[55] S CIENCE L AB . COM. Hexanes. Material Safety Data
Sheets (May 2013), 1–6.
[56] S CIENCE L AB . COM. Isopentane. Material Safety Data
Sheets (May 2013), 1–5.
[57] S CIENCE L AB . COM. n-Decane. Material Safety Data
Sheets (May 2013), 1–6.
[58] S CIENCE L AB . COM. n-Heptane. Material Safety Data
Sheets (May 2013), 1–6.
[59] S CIENCE L AB . COM. n-Nonane. Material Safety Data
Sheets (May 2013), 1–5.
 Coker Gas Plant Design — 18/28

Appendix A
Inlet, Outlet, and Utility Specifications
Feed Conditions
Pressure 20 psig
Temperature 150◦ F
Chemical Formula Common Name Flow Rate (lb/hr) Chemical Formula Common Name Flow Rate (lb/hr)
H2 O Water 16,080 C6 H6 Benzene 394
CO2 Carbon dioxide 2,756 C6 H14 n-Hexane 12,336
N2 Nitrogen 1,568 C6 H14 2-Methylpentane 260
H2 Hydrogen 1,592 C7 H8 Toluene 650
H2 S Hydrogen sulfide 37,454 C7 H16 n-Heptane 1,316
CH4 Methane 42,722 C7 H16 3-Methylhexane 506
C 2 H4 Ethylene 2,670 C8 H10 Paraxylene 1,072
C 2 H6 Ethane 32,532 C8 H10 n-Octane 922
C 3 H6 Propylene 9,538 C8 H18 3-Methylheptane 1,038
C 3 H8 Propane 27,094 C9 H12 Cumene 850
C 4 H8 1-Butene 14,270 C9 H20 n-Nonane 1,684
C4 H10 Isobutane 4,400 C9 H20 2,6-Dimethylheptane 906
C4 H10 Butane 14,252 C10 H14 n-Butylbenzene 134
C5 H12 Isopentane 1,796 C10 H22 n-Decane 142
C5 H12 n-Pentane 188 C10 H22 2,4-Dimethyloctane 862

Outlet Conditions
Product Specifications Required Values Product Sale Price
H2 S Content < 200 wppm
Refinery Fuel Gas $4.00/MMBtu of Higher Heating Value
Pressure > 50 psig
Propane Content > 90 weight percent
Commercial Grade Propane Propylene Content < 10 weight percent $1.00/gallon
H2 S Content < 50 wppm
Propylene Content > 70 weight percent
Refinery Grade Propylene $0.35/lb
H2 S Content < 50 wppm
Propylene Content > 92 weight percent
Chemical Grade Propylene $0.50/lb
H2 S Content < 50 wppm
Butane Content > 90 weight percent
Commercial Grade Butane Butylene Content < 10 weight percent $1.20/gallon
H2 S Content < 50 wppm
Propane and Lighter Content < 5 weight percent
Mixed Butane and Butylene $0.40/gallon
Pentane and Heavier Content < 10 weight percent
Butylene Content > 70 weight percent
Refinery Grade Butylene $0.45/lb
H2 S Content < 50 wppm
Butylene Content > 90 weight percent
Chemical Grade Propylene $0.60/lb
H2 S Content < 50 wppm
Gasoline Blending Component Reid Vapor Pressure < 9 psia $1.30/gallon

Available Utilities
Utility Supply Parameters Return Requirements Cost
85◦ F supply temperature
Cooling Water < 130◦ F $0.20/klb
16,000 gpm available
300 psig saturated vapor < 310◦ F
High Pressure Steam > 50 psig $6.00/klb
Unlimited supply
Must be returned as a liquid to the
condensate header
50 psig saturated vapor < 310◦ F
Low Pressure Steam > 50 psig $4.00/klb
Unlimited supply
Must be returned as a liquid to the
condensate header
Electricity Unlimited supply N/A $50/MWH
20 weight percent MEA (remainder water)
< 0.4 mol H2 S + CO2 /mol MEA
100◦ F
MEA Amine Solution $1.00/gallon
300 psig
> 80 psig
Unlimited supply
 Coker Gas Plant Design — 19/28

Appendix B
R
Aspen Plus Initial Model

Figure 12. Model iteration #1 (Visio PFD top, Aspen Plus

R
screenshot bottom): This model achieved subpar separations and
overused utilities as a result of the erroneous inclusion of a fractionating unit based on a similar process [21]. Further
discussion of this model can be found in 2. Aspen Plus R
Design.
 Coker Gas Plant Design — 20/28

Appendix C With the above parameters determined, the column was speci-
Process Unit Sizing fied in Aspen Plus R
. The remainder of the method consists of
calculations used for determining column height and diame-
FUGK Method ter for use in cost analysis. Liquid and vapor flow rates and
In the FUGK method, distillate and bottoms flowrates are densities are used to determine a liquid/gas average flow rate:
found according to desired mass balances (optimal separation
r
L ρG
between light and heavy keys). K values for each species FLG =
G ρL
in each stream are generated in Aspen by determining equi-
librium properties for the ideal top and bottoms streams and This flow rate is combined with a number of empirical factors
equilibrium vapor density to calculate the diameter of the col- [65] to determine an empirical capacity parameter:
umn. These K values are used to calculate relative volatility C = CSB FST FF FHA
data (α) for each stream:
which is used directly in the equation for vapor flooding ve-
  KLK locity:
α LK = in the distillate
HK D KHK r
ρL − ρG
UF =
  KLK ρG
α LK = in the bottoms
HK B KHK The vapor flooding velocity is typically modified by a conser-
From here, a geometric mean of the two volatilities is found: vative factor ( f = 0.75) and used to calculate column diame-
r ter:
     v
α LK = α LK × α LK
u 4G
DT = t
u
HK avg HK D HK B  
ρG ( fU f )π 1 − AAdt
and this average is combined with the key component mole
fractions in each stream to find a minimum number of column where Ad /At is the ratio of downcomer area to tower cross
stages: h    i sectional area. Note the column diameter is determined based
ln xxHK
LK
× ln[ xxHK
LK
on flow rates at the top of the column to provide an overes-
Nmin = D   B
timate of necessary diameter. Column height is dependent
ln α LK on the number of stages and estimates for overhead vapor
HK avg
disengaging space and bottom fluid reservoir [65]:
α is also used to find the minimum reflux ratio, first by finding
H = Nhtray + htop + hbot
θ , which depends on per-component properties:
n
αi xF,i Column height and diameter are the primary inputs for costing
∑ αi − θ = 1−q calculations. Other factors will be discussed in further detail
i=1 in Appendix D: Process Unit Costing.
where q is the number of moles saturated liquid on the feed
tray per mole feed (which is assumed to be 1 for the columns), Decanter Sizing
xF,i is the mole fraction in the feed for each component ranging The process decanters are assumed to be horizontal vessels
i-n, and αi is the per-component relative volatility. θ is solved with design shown in the decanter schematic.
and implemented in the minimum reflux ratio equation:
n
αi xD,i
Rmin = ∑ −1
i=1 i − θ
α
With Rmin determined, the design reflux ratio (R) is determined
by multiplying by a scaling factor of 1.5. The actual reflux
ratio is used in an iterative calculation to determine the number
of designed stages, N:
" #
R − Rmin 0.566
 
N − Nmin
= 0.75 1 −
N +1 R+1

The feed stage is calculated as a ratio of stages below the feed, Figure 13. Individual decanter design schematic
NB , to stages above the feed, ND :
Mathematically, the design is based on the residence time
  "    2 # necessary for complete aqueous - organic phase partition:
NB B xHK (xLK )B 100µ
ln = 0.206 ln t=
ND D xLK F (xHK )D ρA − ρB
 Coker Gas Plant Design — 21/28

where µ is the viscosity of the continuous phase and ρA , ρB

are fluid densities. The overall drum volume is determined CPL = 300.9(Di )0.63316 (L)0.80161
based on the total volumetric flow rate through the drum:
Fractionating columns F-3 and F-4 do exceed the 170 ft maxi-
Vdrum mum for this correlation, but this oversight was deemed negli-
V̇ = ρV ṁV + ρaq ṁaq + ρorg ṁorg =
t gible given the scope of the project and cursory nature of the
The drum diameter is then determined based on the Sounders- present analysis.
Brown equation for vapor rising velocity [65]. With necessary
drum diameter determined, the vessel length can be calculated Decanters
for use in costing operations. Decanter sizing is very similar to column sizing, however
there is no necessary platforming or tray costs:
Appendix D CP = FMCV
Process Unit Costing where,
For the purposes of unit costing, equipment can be subdivided
into five categories: columns, decanters, compressors, pumps, CV = exp{8.9552 − 0.2330 ln(W ) + 0.04333[ln(W )]2 }
and heat exchangers. In each case, a f.o.b. purchase cost is
calculated based on correlations in literature and then adjusted Compressors
to a modern value using today’s CE index. Compressor f.o.b costs are determined base on the equation:
CP = FD FMCB
Columns
Each column (A-1, E-1, S-1, F-1, F-2, F-3, F-4) is costed as a Each compressor is costed as a centrifugal reciprocating com-
vertical vessel with pricing included for trays and necessary pressor:
ladders and platforms: CB = exp{7.5800 + 0.80 ln Pc }
CP = FMCV +CT +CPL
with electric drive (FD = 1.15) made from stainless steel
where CP is the f.o.b. purchase cost, CV is the cost of the (FM = 2.5).
empty vessel, CT is the cost of trays, and CPL is the cost of
platforming equipment. Note the material cost factor FM is Pumps
selected independently for each column based on process- Each pump in the process is costed as a simple centrifugal
ing requirements. Carbon steel (FM = 1) can be used for pump. In order to determine approximate costs for centrifugal
[H2 S] < 0.2 wt% and in the present design stainless steel type pumps, a size factor must be determined:
316 (FM = 2.6) is used for [H2 S] > 0.2 wt%. CV is calculated S = QH 0.5
based on the weight of the vessel:
where Q is the flow rate through the pump (gal/min) and H
CV = exp{7.0132 + 0.18255 ln(W ) + 0.02297[ln(W )] } 2 is the generated pump head (ft). This size factor can then be
used to calculate a base cost:
W represents the weight of the vessel, and is determined based
on material density, vessel height, inner diameter, and shell CB = exp{9.7171 − 0.6019 ln(S) + 0.0519[ln(S)]2 }
wall thickness:
W = π(Di + ts )(L + 0.8Di )ts ρ Similarly to previously described costing methods, the base
cost is adjusted with a material factor (FM ) and a pump type
The cost of trays is calculated as follows: factor (FT ):
CT = NT FNT FT T FT MCBT CP = FT FMCB
where NT is the number of trays, FN T is an additional ad-
justing the cost of trays in columns where NT > 20, FT T is a Heat Exchangers
costing factor for tray type, FT M accounts for tray material, The f.o.b. cost for most heat exchangers is approximated by:
and CBT is base cost per tray, calculated based on column CP = FP FM FLCB
diameter:
the pressure factor,FP , is calculated base on shell side pressure:
CBT = 468 exp(0.1739Di )
P 2
   
P
The remaining component of the f.o.b. cost equation is the cost 0.9803 + .018 + 0.0017
100 100
of platforming, which can be calculated for smaller towers
(3 < Di < 21 ft; 12 < L < 40 ft) as: Heat exchangers were assumed to be constructed from carbon
0.73960 0.70684 steel (FM = 1). The tube length factor FL was determined [65].
CPL = 361.8(Di ) (L)
For costing purposes, in-line heat exchangers are assumed to
and for larger towers (3 < Di < 24 ft; 27 < L < 170 ft) be U-tube exchangers:
 Coker Gas Plant Design — 22/28

CB = exp{11.147 − 0.9186 ln(A) + 0.09790[ln(A)]2 }

Fractionating and stripping column re-boilers are assumed

to be included in column f.o.b. cost. For a more in-depth
analysis, these units could be costed as kettle vaporizers:

CB = exp{11.967 − 0.8709 ln(A) + 0.09005[ln(A)]2 }

Itemized Total Cost

Label Description Purchased Cost ($)
Major Unit Processes
C-1 Compressor 1 7,422,536
C-2 Compressor 2 5,599,037
D-1 Decanter 1 332,656
D-2 Decanter 2 1,166,476
A-1 Lean Oil Absorber 1,901,437
E-1 Enriching Column 1,687,710
S-1 Stripping Column 2,888,102
F-1 Debutanizer 1,090,398
F-2 C3/C4 Splitter 1,078,110
F-3 C3 Separator 5,327,649
F-4 C4 Separator 4,293,596
Minor Unit Processes
HX-1 12,541
HX-2 54,209
HX-3 48,147
HX-4 20,046
HX-5 16,215
HX-6 8,740
HX-7 9,729
HX-8 9,033
HX-9 8,308
HX-10 11,962
HX-11 18,561
P-1 7,227
P-2 13,516
P-3 6,544
P-4 8,532
C-3 137,495

Total 33,970,830

Equipment Materials
Type 316 SS Type 410 SS Carbon Steel
A-1 P-1 HX-1
C-1 D-1 HX-2
E-1 D-2 HX-3
C-2 S-1 HX-4
C-3 HX-5
HX-6
HX-7
HX-8
HX-9
HX-10
HX-11
P-2
P-3
P-4
F-1
 Coker Gas Plant Design — 23/28

Appendix E
Material Balances Table 6. Absorber-stripper-enriching column scheme
Note that the bolded streams in the material balance tables material balance. See Tables 9-11 for individual balances.
are inlets to the process while non-bolded text indicates outlet (lb/hr) Stripper Feed Lean Oil Enriched Gas Bottoms
H2 O 300 545 - 845
streams. As with any accurate material balance, the sum of CO2 2,755 - 2,755 -
the feed and outlet streams is zero: In a steady-state system N2 1,570 - 1,570 -
H2 1,590 - 1,590 -
with no accumulation or generation terms, in = out. Although H2 S 37,455 - 37,565 -
some components share chemical formulas, the mass flow CH4 42,720 - 42,720 -
C2 H4 2,670 - 2,670 -
rates in these tables are included in the same order of compo- C2 H6 32,530 - 32,535 -
nents as shown in the process feed table. C3 H6 9,540 - 485 9,010
C3 H8 27,095 - 110 26,975
C4 H8 14,270 265 - 14,530
Table 5. Pressurization scheme material balance. C4 H10 4,400 40 - 4,420
C4 H10 14,250 1,420 - 15,670
(lb/hr) Feed Water Condensed Oil Vapor C5 H12 190 1,060 - 1,245
H2 O 16,080 15,781 131 168 C5 H12 1,795 9,705 - 11,500
CO2 2,756 - 57 2,699 C6 H6 395 2,335 - 2,730
N2 1,568 - 2 1,566 C6 H14 12,365 73,315 - 85,680
H2 1,592 - 1 1,591 C6 H14 260 1,540 - 1,800
H2 S 37,454 - 1,495 35,959 C7 H8 650 3,855 - 4,505
CH4 42,722 - 175 42,547 C7 H16 1,315 7,805 - 9,120
C2 H4 2,670 - 44 2,626 C7 H16 505 3,025 - 3,505
C2 H6 32,532 - 970 31,562 C8 H10 1,070 6,360 - 7,430
C3 H6 9,538 - 797 8,741 C8 H10 920 5,475 - 6,390
C3 H8 27,094 - 2,598 24,496 C8 H18 1,040 6,155 - 7,190
C4 H8 14,270 - 3,597 10,673 C9 H12 850 5,040 - 5,890
C4 H10 4,400 - 972 3,428 C9 H20 1,685 9,990 - 11,670
C4 H10 14,252 - 3,989 10,263 C9 H20 905 5,375 - 6,280
C5 H12 188 - 110 78 C10 H14 135 795 - 930
C5 H12 1,796 - 926 870 C10 H22 140 840 - 985
C6 H6 394 - 333 61 C10 H22 860 5,110 - 5,975
C6 H14 12,366 - 10,328 2,038
C6 H14 260 - 205 55 Total 216,225 150,050 122,000 244,275
C7 H8 650 - 622 28 Sum 0
C7 H16 1,316 - 1,241 75
C7 H16 506 - 470 36
C8 H10 1,072 - 1,060 12
C8 H10 922 - 907 15
C8 H18 1,038 - 1,015 23
C9 H12 850 - 845 5
C9 H20 1,684 - 1,676 8 Table 7. Debutanizer material balance.
C9 H20 906 - 896 10
C10 H14 134 - 134 0 (lb/hr) Feed Distillate Bottoms
C10 H22 142 - 142 0 H2 O 846 207 639
C10 H22 862 - 858 4 CO2 - - -
N2 - - -
Total 232,014 15,781 36,595 179,639 H2 - - -
Sum 0 H2 S - - -
CH4 - - -
C2 H4 - - -
C2 H6 - - -
C3 H6 9,009 9,009 -
C3 H8 26,977 26,977 -
C4 H8 14,532 14,225 307
C4 H10 4,419 4,397 22
C4 H10 15,671 14,013 1,658
C5 H12 1,247 9 1,237
C5 H12 11,502 159 11,342
C6 H6 2,730 4 2,730
C6 H14 85,682 - 85,678
C6 H14 1,801 - 1,800
C7 H8 4,505 - 4,505
C7 H16 9,121 - 9,121
C7 H16 3,507 - 3,507
C8 H10 7,430 - 7,430
C8 H10 6,390 - 6,390
C8 H18 7,194 - 7,194
C9 H12 5,891 - 5,891
C9 H20 11,672 - 11,672
C9 H20 6,279 - 6,279
C10 H14 929 - 929
C10 H22 984 - 984
C10 H22 5,974 - 5,974
Total 244,291 69,000 175,291
Sum 0
 Coker Gas Plant Design — 24/28

Table 8. C3 -C4 separator material balance. Table 10. Individual stripping column material balance.
(lb/hr) Feed Propylene Propane Butylene Butane (lb/hr) Feed Distillate Bottoms
H2 O 207 - - - 207 H2 O 976 131 846
CO2 - - - - - CO2 2,570 2,570 -
N2 - - - - - N2 129 129 -
H2 - - - - - H2 62 62 -
H2 S - - - - - H2 S 144,059 144,059 -
CH4 - - - - - CH4 8,695 8,695 -
C2 H4 - - - - - C2 H4 2,118 2,118 -
C2 H6 - - - - - C2 H6 46,335 46,335 -
C3 H6 9,009 8,786 196 27 - C3 H6 32,954 23,945 9,009
C3 H8 26,977 542 26,104 331 - C3 H8 62,310 35,333 26,977
C4 H8 14,225 - 145 13,098 982 C4 H8 18,896 4,364 14,532
C4 H10 4,397 - 88 4,309 0 C4 H10 5,955 1,536 4,419
C4 H10 14,013 - 19 235 13,759 C4 H10 20,414 4,743 15,671
C5 H12 9 - - - 9 C5 H12 1,367 120 1,247
C5 H12 159 - - - 159 C5 H12 13,098 1,596 11,502
C6 H6 4 - - - 4 C6 H6 2,834 4,151 85,682
C6 H14 - - - - - C6 H14 89,833 104 2,730
C6 H14 - - - - - C6 H14 1,893 92 1,801
C7 H8 - - - - - C7 H8 4,580 75 4,505
C7 H16 - - - - - C7 H16 9,352 231 9,121
C7 H16 - - - - - C7 H16 3,585 79 3,507
C8 H10 - - - - - C8 H10 7,479 49 7,430
C8 H10 - - - - - C8 H10 6,449 59 6,390
C8 H18 - - - - - C8 H18 7,270 76 7,194
C9 H12 - - - - - C9 H12 5,917 26 5,891
C9 H20 - - - - - C9 H20 11,731 60 11,672
C9 H20 - - - - - C9 H20 6,323 44 6,279
C10 H14 - - - - - C10 H14 930 2 929
C10 H22 - - - - - C10 H22 986 2 984
C10 H22 - - - - C10 H22 5,997 22 5,974
Total 69,000 9,329 26,551 18,000 15,120 Total 525,098 280,807 244,291
Sum 0 Sum 0

Table 9. Individual absorber material balance. Table 11. Individual enriching column material balance.
(lb/hr) Wet Gas Lean Oil Rich Oil Dry gas (lb/hr) Feed Distillate Bottoms
H2 O 66 547 198 415 H2 O 415 - 415
CO2 4,238 - 779 3,463 CO2 3,460 2,756 703
N2 1,645 - 45 1,600 N2 1,600 1,568 32
H2 1,630 - 23 1,607 H2 1,607 1,592 15
H2 S 122,164 - 38,511 83,704 H2 S 83,704 37,563 46,140
CH4 47,840 - 2,887 44,954 CH4 44,954 42,720 2,234
C2 H4 3,906 - 667 3,239 C2 H4 3,239 2,670 569
C2 H6 59,153 - 14,680 44,477 C2 H6 44,477 32,535 11,943
C3 H6 18,221 - 9,200 9,008 C3 H6 9,017 485 8,533
C3 H8 31,343 - 17,260 14,076 C3 H8 14,076 111 13,965
C4 H8 4,314 263 4,108 454 C4 H8 454 - 454
C4 H10 1,592 19 1,442 168 C4 H10 168 - 168
C4 H10 4,282 1,419 4,161 1,540 C4 H10 1,540 - 1,540
C5 H12 72 1,059 600 531 C5 H12 531 - 531
C5 H12 858 9,706 4,534 6,030 C5 H12 6,030 - 6,030
C6 H6 74 2,336 1,951 459 C6 H6 459 - 459
C6 H14 2,820 73,317 59,779 16,340 C6 H14 16,340 - 16,340
C6 H14 63 1,541 1,186 417 C6 H14 417 - 417
C7 H8 50 3,855 3,607 298 C7 H8 298 - 298
C7 H16 154 7,805 7,143 816 C7 H16 816 - 816
C7 H16 54 3,001 2,704 350 C7 H16 350 - 350
C8 H10 29 6,365 6,181 206 C8 H10 206 - 206
C8 H10 33 5,468 5,261 241 C8 H10 241 - 241
C8 H18 48 6,156 5,873 330 C8 H18 330 - 330
C9 H12 14 5,041 4,949 107 C9 H12 107 - 107
C9 H20 32 9,988 9,808 211 C9 H20 211 - 211
C9 H20 26 5,373 5,214 185 C9 H20 185 - 185
C10 H14 1 795 789 6 C10 H14 6 - 6
C10 H22 1 842 835 8 C10 H22 8 - 8
C10 H22 12 5,112 5,032 92 C10 H22 92 - 92
Total 304,733 150,007 219,407 235,333 Total 235,338 122,000 113,338
Sum 0 Sum 0
 Coker Gas Plant Design — 25/28

Appendix F
Energy Balances

Hin (btu/hr) -2.92E08 Hout (btu/hr) -3.27E08 Qin (btu/hr) 2.45E08

Qout (btu/hr) 3.14E08 Win (btu/hr) 3.69E07 Wout (btu/hr) 2.41E06
 Coker Gas Plant Design — 26/28

Appendix G
Failure Mode Effects Analysis

Figure 14. FMEA-mitigated design: This variation of our finalized model includes the process control units that enhance
worker safety and decrease the probability of a dangerous failure that could pose a threat to the surrounding environment or
community.
 Coker Gas Plant Design — 27/28

Appendix H
Environmental Impact

Figure 15. Biwer’s method applied to our finalized design on per-process and per-unit bases.
 Coker Gas Plant Design — 28/28

Appendix I
Economic Sensitivity Plots

View publication stats