You are on page 1of 18

Analysis of Cryogenic

Distillation for the Separation


of Methane and Carbon
Dioxide

David Belias



4/25/2014




2
Contents
Executive Summary ....................................................................................................................................... 3
Introduction .............................................................................................................................................. 3
Distillation Analysis ....................................................................................................................................... 6
Appendix A .................................................................................................................................................. 16














3
Executive Summary
This report provides an analysis and evaluation of four separation unit operations. Cryogenic
distillation, gas separation membranes, and absorption using water and MDEA were explored. Analysis
for distillation utilized Aspen HYSYS, while the other three analyses utilized Wolfram Mathematica.
Current market values for all prices were used in the cost analysis to find operational costs and
revenues. All calculations can be found in the following discussions. The investigation found that
absorption was the most profitable operation due to the highest purity product being sold. Specifically,
absorption using water is the recommended option because it produced the greatest profit with the
least uncertainty. Distillation had, by far, the greatest costs due to refrigeration and equipment prices.
Gas separation membranes had much lower costs, but they were still triple those of absorption. The
assumptions used in the analysis of absorption need to be explored further in subsequent examinations
to more accurately examine the revenues and costs. In the future, changes in market prices and
innovations in industry may yield different results in the capabilities of each unit operation.
Introduction
The separation of methane and carbon dioxide is a recurring industrial process that can be
completed using numerous unit operations. In this analysis, Cryogenic distillation, gas separation
membranes, and absorption using both water and methyl diethanolamine (MDEA) were investigated in
great detail. The feed stream originated from a natural gas well and its properties are outlined in Table
A below.
Pressure (psia) Temperature
(F)
Methane
Concentration
(mol%)
Carbon Dioxide
Concentration
(mol%)
Feed Flow (MMSCFD)
494.7 77 86 14 35


The feed flow was converted to a molar flow using the ideal gas law at standard temperature
and pressure (14.7 psia, 459.67 R). Equation A below shows how the molar flow was calculated.

( (



)
(


) (


Gases with varying energy levels could be sold for different prices, which are outlined in Table B
below.
Product Methane
Concentration (mol%)
Price ($/MMBtu) Specifications
Waste Gas <10% - -
Low Energy Gas 10-40% 2.50 -
Medium Energy Gas 40-70% 3.00 -
Table A. Properties of the feedstock from the natural gas well.
4
High Energy Gas 70-95% 3.50 -
Pipeline Quality Gas 95% 4.00 600 psia
Liquefied Natural Gas 99.995% 5.20 600 psia
High Purity Carbon
Dioxide
1% 0.003/SCF -


Since the feed was already a high energy gas, pipeline quality gas and liquefied natural gas were
explored as possible product streams. The revenue generated from selling the feedstock, pipeline
quality gas, and liquefied natural gas were calculated using Equations B, C, and D. Selling the high purity
carbon dioxide was also investigated for some unit operations. Calculating the revenue for this stream is
outlined in equation E below. The operating time for this analysis was 7,000 hours per year.

(

)
(

)
(

)
(

)
(


The combustion value of methane was found in Btu per standard cubic feet and converted to
MMBtu per pound mole using Equation F.
1
Equation A was used to calculate the number of moles in
one standard cubic foot.
(


In order to calculate the high purity carbon dioxide flow rate used in Equation E, the ideal gas
law was employed at standard temperature and pressure.
There were also operational costs to consider, before calculating the net profit of a given
system. Each of the aforementioned unit operations requires the pressurization of gas or liquid streams.
The cost associated with this compression can be seen in Equations G and H.
(

)
(

)
(



( (


Table B. Product stream market prices based on purity.
5
Each unit operation has its own unique operational costs, which will be discussed further in the
subsequent sections. The net profit was calculated by subtracting the total operational costs from the
revenue generated from the product and waste streams.
Distillation was the first operation considered for this separation. The simulations were
performed using Aspen HYSYS with the Peng-Robinson fluid package. The main factors investigated
were number of stages, reflux ratio, column pressure, feed quality, and product streams. Costs for
running the column included refrigeration, compression, heating, and equipment.
The next unit operation examined was gas separation membranes. The primary elements
explored were number of stages, recycling strategy, membrane area, feed entry stage, and outlet
stream purity. The main costs associated with gas separation membranes are the cost of membranes,
and the cost of gas compression. The calculations were performed using Wolfram Mathematica.
The third operation that was investigated was MDEA absorption. The main variables of concern
for the absorption involved Henrys Law constant, liquid flow rate, volume of the column, pressure of
the column, and temperature of the column. The more substantial costs included the capital for building
the column and the pressurization of the feed as well as heating and cooling the two feed streams.
These calculations were performed in both Excel and Wolfram Mathematica.














6
Distillation Analysis
The first analysis was done on the output streams that could be produced. Since the gas that
comes out of the well is already high energy gas, there were only two different products that could be
produced. Pipeline quality gas needed a methane concentration of 95%, while liquefied natural gas
needed a methane concentration of 99.995%. The first setup used a cooler to cool the gas to a
saturated vapor that entered the column. The condenser was operated at a pressure of 480 psia. Since
the specifications called for a product pressure of 600 psia, a pump was added at the outlet of the
condenser to pressurize the liquid methane rich stream. At the outlet of the reboiler and condenser,
heaters were added to bring the liquid streams back up to a vapor. No heater was required for the
liquefied natural gas because the specifications required a liquid. This setup is displayed in Figure 1
below. The condenser and all heat exchangers were assumed to have a pressure drop of 5 psia. The
reboiler was assumed to have no pressure drop and the distillation column was assumed to have a 3
psia pressure drop, on average, regardless of the number of stages.








When producing the pipeline quality gas, there were two options considered for the bottoms
stream. High purity carbon dioxide and low energy gas were both considered as the bottoms stream for
distillation. To perform the comparison of the different scenarios, the distillate methane concentration
was specified to be 95% and the bottoms stream was specified to be the two different options. This was
also done for the liquefied natural gas, but changing the methane distillate concentration to 99.995%.
The number of stages was changed during each simulation to find the optimum number of stages. The
Fenske Equation was used as a starting point to estimate the minimum number of stages for each
different scenario that would be expected. Table 1 below gives the values for minimum stages. The
number of stages was increased until the reflux ratio did not change drastically.



Figure 1. Setup 1 for the methane/carbon dioxide cryogenic distillation.
7
Product Stream
High Purity Carbon Dioxide Low Energy Gas
Pipeline Quality Gas 2.3061 1.5727
Liquefied Natural Gas 4.4346 3.7012


Equations 1-4 below show the Fenske Equation and calculations of vapor pressures. The vapor
pressures were calculated using the Antoine Equation at the temperature of the feed entering the
column.
2,3
The alpha value was assumed to be constant at each stage.




There were four factors that contributed to the total cost of distillation. This included
equipment costs, compression costs, refrigeration costs, and heating costs. Equations 5-8 below show
how each of the costs were calculated. Energy and power values were calculated in HYSYS. The
refrigeration costs were obtained from Figure A1 of Appendix A.
Table 2 below shows the optimal number of stages for each scenario, which was determined by
minimizing total cost, which maximizes profit.
Table 1. Minimum stages of different scenarios estimated using Fenske Equation.
(


#
(

) (

) +
(

)

(

) (
(


) ( + (
(

) (
(


) (
(

+
(

)
8
Product Stream
High Purity Carbon Dioxide Low Energy Gas
Pipeline Quality Gas 4 2
Liquefied Natural Gas 26 26


After performing the analysis, it was determined that producing pipeline quality gas with high
purity carbon dioxide in the bottoms was the most profitable scenario. The profit vs. number of stages
plots are pictured below in Figure 2. There are several key points that can be learned from these plots.
Appendix A provides plots of number of stages versus reflux ratio and total cost versus reflux ratio for
these product streams.

The two main contributors to the changes in total cost are the reflux ratio and number of
stages. At a very large reflux ratio, the amount of energy needed for the condenser is very large. This is
because a higher reflux results in more liquid going back into the column, so there is more methane still
in the column to allow the desired purities to be reached. This liquid eventually enters the reboiler at
the bottom, which results in a higher vapor flow rate moving up the column. Finally, this vapor enters
the condenser, so a higher vapor flow rate means more energy needs to be removed to fully condense
Table 2. Optimal stages of different scenarios from HYSYS analysis.
9
all of the vapor. The refrigeration costs increase due to this increased reflux ratio. Increasing the
number of stages decreases the reflux ratio. Each stage is assumed to be 100% efficient in order to
analyze the best profit that a distillation column can make. This means that the mixture reaches
equilibrium at each stage. Going up the column, the temperature and pressure at each stage decrease,
allowing the methane to transfer into the vapor phase more easily. The vapor phase goes to the
condenser and becomes the distillate stream. More stages gives more opportunity for the methane to
transfer into the vapor phase. However, increasing the number of stages also increases the equipment
costs. The optimal number of stages allows the desired purities to be reached, while minimizing cost.
By comparing all four figures, it is apparent that the pipeline quality gas with high purity carbon
dioxide bottoms is the most profitable option, with a maximum profit of $31.52 million per year. For the
pipeline quality gas with high purity carbon dioxide in the bottoms, seen in Figure 2a, there is a clear
maximum and then a linear decline in profit as the number of stages is increased. The optimum number
of stages is clearly 4 stages because that maximizes the profit. The minimum number of stages
determined from the Fenske Equation was 2.3061, so at 3 stages, the reflux is beginning to grow very
large. This is the reason for the decrease in profit. Above 4 stages, the reflux ratio begins to decrease
due to the increase in stages. Refrigeration costs begin to decrease, but equipment costs increase. The
profit decreases because the decrease in refrigeration costs is not enough to counteract the increase in
equipment costs.
The profit versus number of stages for pipeline quality gas with low energy gas bottoms plot is
displayed in Figure 2b. The maximum profit for this scenario is $29.8 million per year. The optimal
number of stages that gives the maximum profit is two. There is no maximum in the plot, so the
maximum profit is at two stages and the profit decreases linearly with an increase in the number of
stages. This could be because the reflux ratio is never large enough for the condenser refrigeration
costs to outweigh the equipment costs. There is no minimum in the plot of total cost vs. reflux ratio
seen in Figure A4 of Appendix A, so the equipment costs always outweigh the refrigeration costs.
When simulating the production of liquefied natural gas, it is clear that this option is not as
profitable as pipeline quality. Looking at Figure 2c, it is seen that the maximum profit for liquefied
natural gas with high purity carbon dioxide bottoms is only $19.25 million per year. Figure 2d shows
that the maximum profit for liquefied natural gas with low energy gas bottoms is $16.25 million per
year. This is a little over half the maximum profit that can be achieved by the pipeline quality gas. Even
though these result in a more pure stream that can be sold for $5.20 per million Btu versus $4.00 per
million Btu for the pipeline quality, it requires many more stages. The optimal number of stages for
both scenarios is 26, which is over six times as many as the pipeline quality requires. As a result,
equipment costs are six times as high, so profits are much lower. Also, the refrigeration costs are higher
because a higher reflux ratio is also needed to get to the desired purity. The required reflux ratio is very
high at a lower number of stages, so the refrigeration costs greatly outweigh any money earned from
selling the products. Below fifteen stages, this is the case, and money is actually lost from producing this
liquefied natural gas. A heater is not needed for the distillate stream because the gas is sold as a liquid.
Even though this is the case, heating costs are almost negligible, so it does not change the profit
10
drastically. Above the optimal number of stages, the profit begins to decrease again because equipment
costs begin to dominate.
After analyzing the data, it was determined that producing pipeline quality gas is much more
profitable than liquefied natural gas. It was also determined that producing high purity carbon dioxide
in the bottoms is more profitable than producing low energy gas. Generating the very pure liquefied
natural gas requires many more stages and a higher reflux ratio in order to achieve the desired purity.
The refrigeration and equipment costs outweigh the revenue from selling the product streams due to
the high number of stages and reflux ratio. Producing the high purity carbon dioxide in the bottoms is
more profitable than the low energy gas because of the flow rate that can be produced. Since the
distillate is very pure with methane, the flow rate of methane in the bottoms is very low. As a result,
not much money can be made from this stream. On the other hand, there is a large amount of carbon
dioxide in the bottoms, so more money can be generated from this stream. Even though it takes several
more stages to produce this, selling the carbon dioxide outweighs the cost. In the end, the best option
was producing pipeline quality gas with high purity carbon dioxide in the bottoms.
The next investigation was to determine the best pressure for operating the column to produce
the pipeline quality gas with high purity carbon dioxide in the bottoms. Figure 3 below shows the xy
diagram for the methane/carbon dioxide mixture at feed pressure (494.7 psia) and at 600 psia, which is
required for the pipeline quality gas. It can be seen that the equilibrium curve at 494.7 psia has more
separation from the y=x line than for 600 psia. This means that the separation is easier at 494.7 psia, so
fewer stages are required. However, the operating costs may be lower for a setup with the condenser
operating at a pressure of 600 psia, so this was investigated.
11

Setup 2 for the distillation column is outlined in Figure 4 and it uses a compressor to pressurize
the feed that enters the column. At the outlet of the distillate, there is no pump required because the
pressure of the stream is at the required pressure specification.


In order to find the effect of column pressure on maximum profit, the column pressure was
increased to 605 psia to find the maximum profit at the highest logical pressure choice. Since the heater
has a pressure drop of 5 psia, the distillate stream had to be at a pressure of 605 psia to maintain
specifications for pipeline quality gas. There is no point in operating the column above 605 psia because
the specifications only require a pressure of 600 psia. The analysis with setup 2 was performed, and the
comparison with setup one is summarized below in Figure 5.
Figure 4. Setup 2 for the methane/carbon dioxide cryogenic distillation.
12

Comparing the results from Figure 5, there is not a drastic increase in the profit due to
increasing the column pressure. The column at 605 psia results in a maximum profit of $32.1 million,
while operating at 480 psia gives a maximum profit of $31.52 million. This is not a very drastic increase
in maximum profit. This can be attributed to an increase in equipment costs because the optimal
number of stages is 6 versus 4 for a column pressure of 480 psia. However, the required condenser
energy decreases, so refrigeration costs decrease, outweighing the increase in equipment costs.
Looking at both plots in Figure 5, it can be seen that there is a much smoother maximum in the profit for
a pressure of 605 psia. This setup is better because there is more allowance in the variability of the
number of stages in the column, and thus the reflux ratio. If five or seven stages are used, the profit will
still be near the maximum profit. Other operating pressures were explored between 480 and 605 psia,
but they did not considerably change the maximum profit. The problem with these operating pressures
is that a compressor before the column and a pump after the column would be necessary to achieve the
required pressure specifications. Adding an extra piece of equipment adds more complications and
increases the chances of an equipment failure.
In the first two analyses, the feed entered the column as a vapor, so the final analysis was for
the effect of feed quality on profit. Setup two was used with the optimal range of stages for varying
vapor fraction feeds until a maximum profit was found. The optimal number of stages was between 4
and 8, giving an unbiased comparison of the feed qualities. The maximum profit was plotted against the
varying vapor fractions of the feed. The results are displayed below in Figure 6.










Looking at Figure 6, it is apparent that the best feed quality is a saturated vapor. Again, the
costs are dominated by the cost of refrigeration, which is determined by the reflux ratio. The higher the
vapor fraction in the feed, the higher the required reflux ratio. This increases refrigeration costs in the
Figure 6. Profit at varying feed qualities for producing
pipeline quality gas with high purity carbon dioxide bottoms.
29
29.5
30
30.5
31
31.5
32
32.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
P
r
o
f
i
t

(
$
M
/
y
r
)

Vapor Fraction of Feed
Pipeline Quality Gas with High
Purity CO2 Bottoms
13
condenser. On the other hand, the higher the liquid fraction in the feed, the higher the refrigeration
costs of the cooler because the feed needs to be cooled more. Judging by the plot, the energy
requirements for the cooler are much higher than the condenser. The flow rate into the cooler is higher
than the flow rate into the condenser, so more energy needs to be removed from the feed. The
minimum in the plot occurs because once some vapor is introduced into the column from the feed, the
condenser energy requirements increase due to the increase in reflux ratio. Table 3 below shows
energy requirements of the condenser and cooler for a low and high vapor fraction in the feed.
Vapor Fraction Number of Stages Reflux Ratio Cooler Energy
(Btu/hr)
Condenser Energy
(Btu/hr)
0.1 4 2.02 1.78E+07 1.55E+07
0.1 5 0.748 1.78E+07 8.96E+06
0.1 6 0.323 1.78E+07 6.78E+06
0.1 7 0.296 1.78E+07 6.65E+06
0.1 8 0.207 1.78E+07 6.19E+06
0.5 4 2.18 1.52E+07 1.63E+07
0.5 5 0.808 1.52E+07 9.27E+06
0.5 6 0.46 1.52E+07 7.49E+06
0.5 7 0.411 1.52E+07 7.23E+06
0.5 8 0.348 1.52E+07 6.91E+06
0.9 4 2.41 1.19E+07 1.75E+07
0.9 5 1.05 1.19E+07 1.05E+07
0.9 6 0.726 1.19E+07 8.85E+06
0.9 7 0.63 1.19E+07 8.35E+06
0.9 8 0.564 1.19E+07 8.02E+06


As expected, an increase in the vapor fraction of the feed decreases the cooler energy
requirements and increases the condenser energy requirements.
After all of the analysis, it was determined that the optimal design was a distillation column
operated with a condenser pressure of 605 psia and six stages. The feed is compressed to 620 psia and
fed to the column as a saturated vapor. This design results in a maximum profit of $32.1 million. This
analysis was not entirely comprehensive because there are other factors that could have been
considered. To cut down on the costs of cooling the feed stream and heating up the distillate, the
distillate could have been used to cool the feed. The distillate did not need to be sold at such a low
temperature, and the feed needed to be cooled to a saturated vapor. Another factor that could have
been considered was feeding the column with a superheated vapor. This would cut down on the
refrigeration costs for the cooler, which could have potentially saved some money. However, the
condenser would need to remove more energy since the vapor is superheated. In the end, these
changes may not have increased profits by too much. Using a flash separator instead of an entire
distillation column is another consideration that could be made for further analysis. This would cut
14
down on the equipment and heating costs of the column. However, refrigeration would be needed to
cool the gas to a saturated liquid. Also, since the feed was very rich in methane, it would be difficult to
reach pipeline quality gas and liquefied natural gas with a single flash separator. As a result, the profit
would most likely not be any better than what was achieved with a full distillation column.
The final question that needed to be considered was whether it was economically beneficial to
purify the feed using distillation versus just selling the feed that came from the natural gas well. Figure
7 below shows the maximum profit from each of the different options explored, as well as the costs
associated with each option.



It can be inferred from Figure 7 that the most profit could be generated from selling the
feedstock. The costs of purifying the stream into a higher energy gas outweighed the revenue
generated. Refrigeration and equipment costs are the highest, with compression and heating costs
being very marginal. Producing the liquefied natural gas almost doubled costs, so this was clearly not a
viable option.
Clearly, cryogenic distillation is a very expensive purification technique. As a result, it is more
profitable to sell the feed gas, rather than selling the more pure pipeline quality gas with high purity
Figure 7. Profit and costs in millions of dollars per year for producing varying product streams.
15
carbon dioxide in the bottoms. Refrigeration costs make up nearly 75% of the total costs, so it is the
dominant factor in driving the profit down for this purification technique. Equipment costs are the
second highest, so reflux ratio and number of stages determine the total cost. Higher reflux ratio results
in fewer stages, but this increases the refrigeration costs. On the other hand, a lower reflux ratio results
in more stages, so equipment costs increase and refrigeration costs decrease. Finding the optimum
balance between the two gave the maximum profit. When trying to produce liquefied natural gas, six
times as many stages were needed, as well as a higher reflux ratio. This drove refrigeration and
equipment costs very high, so a lower profit was generated. When analyzing the operating pressure of
the column, it was determined that increasing the pressure only increases profit minimally. Finally, the
higher the vapor fraction in the feed, the higher the profit generated. This is due to the large
refrigeration costs in the condenser for a more liquefied feed. After completing all of this analysis,
distillation was most likely not the best option for purification of the feedstock.

















16
Appendix A

























Figure A1. Effect of temperature level on cost of refrigeration for cooler and condenser.
Figure A2. Plots of a) Number of Stages vs. Reflux Ratio and b) Total Cost vs. Reflux Ratio for
producing pipeline quality gas with high purity carbon dioxide bottoms with a condenser pressure
of 605 psia.
17

























Figure A3. Plots of a) Number of Stages vs. Reflux Ratio and b) Total Cost vs. Reflux Ratio for
producing pipeline quality gas with high purity carbon dioxide bottoms with a condenser pressure
of 480 psia.
Figure A4. Plots of a) Number of Stages vs. Reflux Ratio and b) Total Cost vs. Reflux Ratio for
producing pipeline quality gas with low energy gas bottoms with a condenser pressure of 480 psia.
Figure A5. Plots of a) Number of Stages vs. Reflux Ratio and b) Total Cost vs. Reflux Ratio for
producing liquefied natural gas with high purity carbon dioxide bottoms with a condenser pressure
of 480 psia.
18









Profit ($M/yr)

Refrigeration
Cost ($M/yr)

Compression
Cost ($M/yr)

Equipment
Cost ($M/yr)

Heating
Cost
($M/yr)
Feed 35.38 0 0 0 0
600 Psia Pipeline
CO2
32.1 8.11 0.325 2.65 0.144
480 Psia Pipeline
CO2
31.52 9.83 0.0269 1.77 0.1799
480 Psia Pipeline
Low Energy
29.8 9.35 0.0274 0.884 0.192
480 Psia
Liquefied CO2
19.25 12.48 0.01899 11.49 0.0843
480 Psia
Liquefied Low
Energy
16.25 12.32 0.01899 11.49 0.0763








Figure A6. Plots of a) Number of Stages vs. Reflux Ratio and b) Total Cost vs. Reflux Ratio for
producing liquefied natural gas with low energy gas bottoms with a condenser pressure of 480 psia.
Table A1. Profit and different costs in millions of dollars per year for the different product streams
explored.

You might also like