Skip to main content

Oxygen Purification Experiment


Mass Transfer in Packed Beds





By:



Brett Koehn







Submitted June 5, 2015







Based on experimental work for Laboratory #7,

conducted between May 14, 2015 and March 28, 2015 with group members



Gary Banh and E.J. Crowley



in Section 5 of ECH 145B











Table of Contents:

Abstract………………………………………………………………………………………………………………………………………………..3

Introduction…………………………………………………………………………………………………………………………………………....3

Theory…………………………………………………………………………………………………………………………………………………5

Experimental Methods………………………………………………………………………………………………………………………………...7

                -Figure 1…………………………………………………………………………………………………………………………………...8

                -Figure 2…...………………………………………………………………………………………………………………………………9

Results………………………………………………………………………………………………………………………………………………..11

                -Figure 3………………………………………………………………………………………………………………………………….11

                -Figure 4………………………………………………………………………………………………………………………………….12

                -Figure 5……………………………………………………………………………………………………………………………….…13

                -Figure 6……………………………………………………………………………………………………………………….…………14

Discussion……………………………………………………………………………………………………………………………………………15

Design………………………………………………………………………………………………………………………………………………..17

                Table 1………………………………………………………………………………………………………………………………...…17

                Figure 7…………………………………………………………………………………………………………………………………..18

Conclusion………………………………………………………………………………………………………………………………………...…19

Nomenclature Table…………………………………………………………………………….……………………………………………………19

                -Table 2…………………………………………………………………………………………………………………………………..19

References……………………………………………………………………………………………………………………………………………20

Appendix A……………………………………………………………………………………………………………………………..……………21

                Figure 8……………………………………………………………………………………………………………………………..……21

                Figure 9………………………………………………………………………………………………………………………………..…22

                Figure 10…………………………………………………………………………………………………………………………………23

                Figure 11…………………………………………………………………………………………………………………………………24

                Figure 12…………………………………………………………………………………………………………………………………25

                Figure 13…………………………………………………………………………………………………………………………………26

                Figure 14…………………………………………………………………………………………………………………………………27

                Table 3…………………………………………………………………………………………………………………………………...28

                Table 4…………………………………………………………………………………………………………………………………...28

                Table 5…………………………………………………………………………………………………………………………………...29

                Table 6………………………………………………………………………………………………………………………………...…36

Appendix B………………………………………………………………………………………………………………………………………..…38

                Figure 15…………………………………………………………………………………………………………………………………38

                Figure 16…………………………………………………………………………………………………………………………………39

Appendix C………………………………………………………………………………………………………………………………...………...40

Figure 17…………………………………………………………………………………………………………………………………40

Abstract:
             During this lab, we analyzed synthesis gas orsyngas’, which is a fuel that can be created with methane and oxygen. Methane is a greenhouse gas that accounts for 10% of all greenhouse gas emissions. Therefore reducing methane emissions is good for the environment. Converting landfill methane emissions into syngas is a useful process because it produces fuel, stops methane emissions, and only requires oxygen. However, this reaction requires a catalyst that does not work in the presence of water. Therefore to create syngas air must be scrubbed of water moisture. Another solution is to purchase pure oxygen. We found that column with a packed bed full of Drierite can be used to remove the water moisture. 5 of these columns were then used to create a design schematic that would cost $420 a year, without initial costs. Purchasing the oxygen would cost $12.8 million a year. Therefore we should purify air ourselves.

Introduction:
            The concepts of mass transfer are very powerful and useful. One important concept of mass transfer is scrubbing. Scrubbing is used to purify or filter a stream of multiple components. Typically the stream being scrubbed contains a desirable and undesirable component. This stream is run through a column of a third component that removes the undesirable component from the stream. However, the scrubbing component eventually becomes ineffective, and needs to be replaced. Therefore, if we understand the mass transfer concepts of scrubbing, we can also determine when the scrubbing component needs to be replaced.
            One practical use of scrubbing is eliminating hazardous materials such as methanol, ethanol, isopropanol, butanol, and acetone from gaseous streams.[1] These hazardous materials are also water soluble. Therefore, if these materials are released into the environment, they will contaminate drinking water. In this specific scrubbing example, a mixture of water and amphiphilic block copolymer is used to scrub the gaseous streams entering. However, the scrubber does become ineffective. Therefore, it is very useful to know when the mixture of water and amphiphilic block copolymer will become ineffective.
Scrubbers are also used in the creation of high fructose corn syrup.[2] High fructose corn syrup must be purified of unwanted proteins before processing. Therefore, high fructose corn syrup is scrubbed with activated carbons and ion-exchange resins. The syrup will not process correctly and will taste poorly, if high fructose corn syrup is not purified of these unwanted proteins. The process of scrubbing streams to separate different components accounts for 40%-70% of both capital and operating costs in the process industry.[3]
 During this lab, we examined the mass transfer between water vapor and solid grains of calcium sulfate.  The calcium sulfate was placed inside a column or packed bed and an air stream saturated with water vapor was allowed to flow through the packed bed. The grains of calcium sulfate rapidly scrubbed the water vapor from inflowing the air stream, because calcium sulfate absorbs atmospheric moisture rapidly.[4] Therefore, the exiting air stream was very pure oxygen with almost no water vapor left. Removing water from the air stream is important because water can make reaction catalysts ineffective. During this lab report, we wanted to create “syngas” which cannot be created in the presence of water, but requires oxygen as a reactant. Therefore we analyzed the system of the packed bed to determine the cheapest method of obtaining pure oxygen. We found that creating oxygen ourselves is much cheaper.

Theory:
            The mass transfer concepts that are involved in this packed bed experiment are very complicated. However, several assumptions can be made to make this system easier to understand. First off, if the packed bed is treated as a single slab we know that the gas stream interacts with a, which is the contact area per unit of packed volume of column. We also know the driving force across a is kc which has units of length per time. Furthermore, we know that driving force for the mass transfer coefficient will be defined by the difference of the mole fraction of water inside the airstream, xw, and the mole fraction of water at the surface the slab, xws. xws is approximated to be equal to the concentration at the surface of the packed bed, Cws, and the equilibrium absorption constant, Keq, which has units inversely proportional to concentration. Dimensionless equations for the depletion of water versus time from the inflowing air stream, equation 1, and the absorption of water at the surface of the slab, equation 2, can now be solved for
 where X is the normalized water vapor mole fraction in the air stream is the intial mole fraction of water in the air stream, Y is the normalized water vapor mole fraction on the surface of the slab, τ is the dimensionless time, t is the time, seconds, NA is the flux of air coming in, moles per seconds, ε is the porosity, C is the total molar concentration of the air stream, moles per cubic meter,, ξ is the dimensionless length within the slab, z is the length traveled in the slab, meters, and L is the total length of the slab, meters.
            These two equations can be solved with a finite difference approximation to find the best solution for β. This was done by creating a running approximate α and β values into a finite difference scheme. This scheme then gave values for approximated values for X and Y. X was plotted against τ which gave a clear representation of the break through curve. Y was also plotted against ξ which gave a representation of how much water was being scrubbed out by Drierite. The best fitted values of α and β were then used found by evaluating the solution graphically. β  and α were then used to calculate the Reynolds number and the Sherwood number.
            The Reynolds number is defined by

where ρ is density, kilograms per cubic meters, D is diameter, meters, v is velocity, meters per second, and μ is the kinematic viscosity, Pascal seconds. The Sherwood number can be assumed to be

where D is the diffusivity of water vapor, meters cubed per second, and d is the diameter of the Drierite grains, meters. There is also one important relationship which was developed by Gupta and Thodos between the Sherwood number and the Reynolds numbers which is

where Sc is Schmidt number.


Experimental Methods:
            This lab was begun by weighing approximately 5 grams of Calcium Sulfate, which is commercially sold as Drierite. This Drierite was then placed inside of the slab or packed bed as shown in figure 2 below. The packed bed was then filled with glass wool at the entrance and exit of the column. Cooling water was then allowed to flow through the outer shell of the packed bed to create a counter current heat exchanger. However, cooling water was only used for half of the trials. Next the wet air was allowed to flow through the glass wool and Drierite which were both inside the shell of the packed bed. The Drierite removed the moisture from the wet air, which allowed dry air to flow out the end of the packed bed. Thermocouples were then placed on the bottom and top of the packed bed. These thermocouples measured the temperatures of the dry air coming out and the wet air coming in. Once the packed bed was set up with wet air flowing through it, we started taking pictures of the packed bed. We were able to take pictures of the packed bed by using the automated camera program. The lens of the camera was also position horizontally in order to make the data analysis easier. Figure 1 shows some experimental photos that we took at the beginning middle and end of our experiment.

Figure 1: This figure above shows the progression of Drierite absorbing water moisture for Trial 3. The top picture was taken at a time of 40 seconds. The middle picture was taken at a time of 455 seconds. The third picture was taken at a time of 1025 seconds.

            We used flow rates of 12.944 liters per second, 11218 liters per seconds, 5000 liters per seconds, and 2000 liters per seconds for trials 1, 2, 3, and 4 respectively. Trials 1 and 4 were performed without cooling water. Trials 2 and 3 were performed with cooling water.


Figure 2: This figure above shows a detailed sketch of the packed bed full of Drierite. Wet air comes in through the bottom of the column, and dry air comes out the top of the column. The temperatures at the top and bottom of the column were measured with thermocouples. Cooling water was also used for half of the trials.

           We then analyzed the pictures of the packed bed by using Matlab. We assumed that blue and red Drierite corresponded to Y values of 1 and 0 respectively.  This allowed us to calculate the Y versus ξ, which was obtained by equation 3d. We then plotted Y versus ξ for specific times. We then fit this data numerically by using a finite differences solution on equations 1 and 2. Finally, we calculated the best fit curves.

           ξ was then calculated for when Y reached values .25 and .75. We then plotted this against τ, which was calculated with equation 3c and numerically fit this by using a finite differences approach on equations 1 and 2. This approach best fit the data in figures and simultaneously found values of α and β. This was done until the best values of α and β were found.  The temperature of the thermocouples versus time was calculated and plotted for a cooling and non-cooling trial. The Reynolds and Sherwood numbers were then calculated with equations 4 and 5 respectively. The Reynolds and Sherwood number was then plotted against the Reynolds numbers. Theoretical curves were then superimposed over this plot with data obtained from equation 6.




Results:



Figure 3: This figure above shows Y versus ξ. This data was obtained from the experimental values obtain during trial 1 without cooling.

            As shown figure 3, the variable Y was plotted versus ξ. These values were plotted for different τ values.  The experimental values of Y decreased as the values of ξ increased for all of the trials that we analyzed. Furthermore, the numerically fitted data values of Y also decreased as the values of ξ increased. The break through curve is the section in the best fit lines where the slope initially decreases at an increasing rate, eventually reached a slope which is completely vertical, and finally the slope increases at a decreasing rate until the curve is level. This region of the curve represents the where the Drierite is completely saturated with water vapors. Therefore, the water moisture is “breaking through” the packed bed. The experimental data sets also have a break through region. However, this region is much harder to determine. Therefore, the break through regions had to be estimated.


Figure 4: This figure above shows ξ versus τ. This data was obtained from the experimental values obtain during trial 1 with cooling. The blue data represents the dimensionless distance for when Y=.25. The red data represents the dimensionless distance for when Y=.75. The error bars are 2 standard deviations of ξ.

As shown in figure 4, the variables ξ25 and ξ75 were plotted against τ. ξ25 and ξ75 were the dimensionless distances when Y was equal to .25 and .75 respectively. The values of ξ25 increased logarithmically as the experiment preceded for all of the trials. The values of ξ75 increased exponentially as the experiment preceded for all of the trials. The experiments had τ values that ranged from 0 to 3.5E6.




Figure 5: This figure above shows temperature versus time for trial 1 and trial 2. Trial 1 is the trial without cooling, and trial 2 was done with cooling.
As shown in figure 5, the temperatures at the top and bottom of the packed bed were plotted against dimensional time. The top of the column without cooling had the highest temperatures, which is represented in figure 5. The temperatures of the top of the column with cooling water, the bottom of the column with cooling water, and the bottom of the column without cooling water all had temperatures of about 24 degrees Celsius throughout the trials. Furthermore, initially the top of the temperature at the top of the column is less than the temperature at the bottom of the column for trials 1 and 2. Figure 14 represents the data from obtained trials 3, which had cooling, and trial 4, which had no cooling. As shown in figure 14 the temperatures at the bottom of the column with and without cooling water remain constant at 22 degrees Celsius. The temperature at top of the column without cooling water was the highest temperature for the full duration of the trials. The temperature at the bottom of the column with cooling water was the lowest temperature for the full duration of the trials.


Figure 6: This figure above shows the Sherwood number plotted against the Reynolds number. The error bars are two standard deviations of the Reynolds number and Sherwood number for horizontal and vertical directions respectively. The experimental data increases at a logarithmic rate which agrees with the theoretical data.
As shown in figure 6, the Sherwood number was plotted against the Reynolds number. These were found from our values of α and β. α had values of .7±.32, .89±.69, .80±.56, .20±.17 for trials 1-4. β had values of 15.3±7.6, 4.21±5.76, 6.5±3.98, and 53±23.4 for trials 1-4. The values of the Reynolds numbers were 347±3.12, 300±4.62, 133±5.23, and 53.4±2.5 for trials 1-4. The Sherwood numbers were 35.6±29, 33.7±43, 21.9±48.6, and 31.3±23.3.


Discussion:
            From figure 3, we know that the experimental values Y decreases as ξ increases. This is because as water moisture flows through the Drierite, and the Drierite becomes saturated. Once the Drierite is saturated, it can no longer absorb water. Therefore as the dimensionless distance increases in the packed bed, the relative mole fraction on the surface of the packed bed will eventually be very small compared to the mole fraction of water flowing in the air stream. The mole fraction in the air stream will eventually reach zero, and likewise Y eventually reached 0. Therefore, as the time increased during this experiment, every value of Y became 0. This happened because the packed bed became fully saturated with water after a certain amount of time has passed during this experiment.

           From figure 4, we know that the experimental values of Y increases as τ increases. This agrees with figure 3, because it shows that the packed bed is becoming more saturated with water over time. Figure 4 also shows that Y approaches 1 at higher values of τ. Furthermore, there are more values of ξ25 when compared to ξ75. This shows that the mass transfer decreases with time and distance.

            From figure 5, we see that the reaction between water and Drierite is an exothermic reaction, because the temperature at the top of the packed bed is much higher without cooling. We also know that the absorption process takes longer when cooling is not used because the cooling trials took less time. The mass transfer is more effective when cooling is used because the Drierite absorbs water the best at low temperatures. Drierite releases water moisture at high temperatures, which is bad for the mass transfer between Drierite and water.
            The mass transfer coefficient increases as the flow rate increases. kc is in units of meters per seconds. Therefore increasing the velocity will increases the value of kc. In addition, the mass transfer coefficient will decrease if we were to scale up this process. As the process gets larger, the value of a get smaller. a gets smaller because the volume will become much larger compared to the interaction area.
            We used to limiting cases while performing our experiments. The thing we limited was flow rate. The flow rates were limited to values of 5 liters per second and 2 liters per second. However, these flowrate are only valid if the Reynolds number is between 1900 and 10,400. These values are acceptable theoretically. This is represented in figure 6 as the experimental data agrees with literature values. This is because the calculated mass transfer coefficients were within the except error bounds. We also limited our time. The time was limited to 450 seconds.
            The error in the mass transfer coefficient is 2 standard deviations vertically for the Sherwood number and 2 standard deviations for the Reynolds number.
            The error bounds for the calculated mass transfer coefficient α were .32, .69, .56, and .17 for trials 1-4. β had error bounds values of 7.6, 5.76, 3.98, 23.4 for trials 1-4. The error in these results is large because β and α were based on a lot of uncertain quantities such as the mole fraction of water, length, time, flow rate, and temperature.
            The temperature graph shows us that this reaction is going to produce a lot of heat if we scale up this backed bed. Therefore, we need to make sure we have enough cooling water so none of the equipment is damaged. Furthermore, sufficient cooling is also important for the reaction to proceed at maximum efficiency because Drierite absorbs more water at lower temperatures.

Design:
           We need to design a packed bed which can provide enough pure oxygen to react 10 million cubic feet of methane each year and create the valuable fuel, syngas. This is equal to about 3.3 billion moles of methane. Therefore, we need about 1.65 billion moles of oxygen each year because the reaction requires 1 mole of oxygen to react with two moles of methane. This is equal to incoming air flow rate of 14.8 million cubic feet each year or an incoming flow rate of 2.1 cubic feet per second.
 
           2.1 cubic feet per second is an extremely large incoming flow rate. However, if we utilize a condenser, such as the industrial evaporator vacuum controller, we only need an inflow airstream of around 1.35 cubic feet per second.[5] The engine driven pump that we use meets this flow rate requirement of 1.35 cubic feet per seconds.[6] The flow rate will then enter 4 of the active columns and eventually react with the methane as pure oxygen. The piping, tubing, and columns can all be purchased from the same corporation.[7] The prices of the previously mentioned equipment is shown in table 1.
The design scheme costs $67,701.11 to produce the required amount of pure oxygen. However, this price is deceivingly expensive. The initial costs for this design project are $67,281.11. The annual cost of running this design is only $420 per year. The annual cost of purchasing oxygen is about $12,800,000. After the initial year, the design scheme costs approximately 30 thousand times less each year. Therefore, this scrubber will become more profitable the longer it is operated.


Figure 7: The schematic design is shown figure 7. This design requires a pump, a condenser, 5 columns, and a heater. The heater is not shown because the heater is not a part of the cyclic process this schematic undergoes. There is also cooling water running counter current through each of the columns.
             The design schematic is shown in figure 7. This design contains 5 columns which will all have cool water running through them. However, one of these columns is always inactive. The column goes inactive because there is valve which can be closed below each column. Therefore, an employee can close the valve below the column. The employee can then remove the inactive column and extract the saturated Drierite. The moisture is then removed from the Drierite by a heater. The Drierite is then placed inside the inactive column. The inactive column is then made active again by opening the closed valve below the column. The employee will continuously cycle through each column. Therefore, no wet air is ever allowed into the reactor.

Conclusion:
            We should scrub atmospheric air ourselves instead of purchasing pure oxygen. Syngas can be created once we have the pure oxygen and methane from the landfill. Therefore, we can stop methane emissions from the landfill, and create a valuable fuel. Annual cost of scrubbing oxygen ourselves is $420. The annual cost of buying pure oxygen is $12.8 million a year. However, the design will be completely ineffective if there is not a very efficient condenser and powerful pump. The values we received for the dimensionless mass transfer coefficients were .7±.32, .89±.69, .80±.56, .20±.17 for α and 15.3±7.6, 4.21±5.76, 6.5±3.98, and 53±23.4 for β. The values we received for the Reynolds number were 347±3.12, 300±4.62, 133±5.23, and 53.4±2.5. The values we received for the Sherwood number were 35.6±29, 33.7±43, 21.9±48.6, and 31.3±23.3.

Nomenclature Table:

Table 2: Nomenclature Table

References:
1-Jacobs, John. "Air Pollution Control Technology Fact Sheet." Packed Beds. 2009. Accessed June 4, 2015.
2-Parker, Ken. Production of Syrup. 2008.
3-Kovel, Nick. "For Absorption, Desorption, Rectification and Direct Heat Transfer." In Packed Bed Columns. Vol. 1. Elsevier, 2006.
4-ECH 145B, Resources, hammond_j-chem-edu_1935.pdf’
5-Science Lab Evaporators: Amazon.com: Industrial & Scientific. Accessed June 3, 2015.
6-Accessed June 6, 2015. http://www.amazon.com/AMT-5586-Y6-Engine-Driven-1000gpm/dp/B00JBKOYOI/ref=sr_1_1?s=industrial&ie=UTF8&qid=1433550200&sr=1-1.
7-Gavin, Jake. "Oxygen Element Facts." Chemicool. Accessed June 1, 2015.

Appendix A:
 
Figure 8: This figure above shows Y versus ξ. This data was obtained from the experimental values obtain during trial 2 with cooling.


Figure 9: This figure above shows ξ versus τ. This data was obtained from the experimental values obtain during trial 2 with cooling. The blue data represents the dimensionless distance for when Y=.25. The red data represents the dimensionless distance for when Y=.75. The error bars are 2 standard deviations of ξ.

Figure 10: This figure above shows Y versus ξ. This data was obtained from the experimental values obtain during trial 3 with cooling.

Figure 11: This figure above shows ξ versus τ. This data was obtained from the experimental values obtain during trial 3 with cooling. The blue data represents the dimensionless distance for when Y=.25. The red data represents the dimensionless distance for when Y=.75. The error bars are 2 standard deviations of ξ.

Figure 12: This figure above shows Y versus ξ. This data was obtained from the experimental values obtain during trial 4 without cooling.

Figure 13: This figure above shows ξ versus τ. This data was obtained from the experimental values obtain during trial 4 without cooling. The blue data represents the dimensionless distance for when Y=.25. The red data represents the dimensionless distance for when Y=.75. The error bars are 2 standard deviations of ξ.

Figure 14: This figure above shows temperature versus time for trial 3 and trial 4. Trial 3 is the trial with cooling, and trial 4 was done without cooling.


Table 3: Trial 1 Raw Data

Photo
Temp C Top
Bottom
Flow Rate
reading
mL
mass (g)
1
22
25
steel
100
12944
10.633
2
24
24
3
33
25
4
38
25
5
40
24
6
43
24
7
43
25
8
42
24
9
41
24
10
40
24
11
39
24
12
37
24
13
36
24
14
35
24
15
34
24
16
33
24



Table 4: Trial 2 Raw Data
Photo
Temp C Top
Bottom
Flow Rate
reading
mL
mass (g)
1
17
25
steel
95
11218
12.537
2
17
25
3
25
25
4
28
24
5
28
24
6
28
24
7
27
24
8
26
24
9
25
24
10
24
24
11
24
24
12
23
24
13
23
24
14
23
24
15
22
24
Table 5: Trial 3 Raw Data
Photo
Temp C Top
Bottom
Flow Rate
reading
mL
mass (g)
1
22
23
glass
20
961
5.137
2
22
22
steel
40
4732
3
22
22
4
20
22
5
20
22
6
19
22
7
19
20
8
19
22
9
19
22
10
19
22
11
19
22
12
19
22
13
19
22
14
19
22
15
19
22
16
19
22
17
19
22
18
19
22
19
19
22
20
19
22
21
19
22
22
19
22
23
19
22
24
19
22
25
19
22
26
19
22
27
19
22
28
19
22
29
19
22
30
19
22
31
19
22
32
19
22
33
19
22
34
19
22
35
19
22
36
19
22
37
19
22
38
19
22
39
19
22
40
19
22
41
19
22
42
19
22
43
19
22
44
19
22
45
19
22
46
19
22
47
19
22
48
19
22
49
19
22
50
19
22
51
19
22
52
19
22
53
19
22
54
19
22
55
19
22
56
19
22
57
19
22
58
19
22
59
19
22
60
19
22
61
19
22
62
19
22
63
19
22
64
19
22
65
19
22
66
19
22
67
20
22
68
20
22
69
20
22
70
20
22
71
20
22
72
20
22
73
20
22
74
20
22
75
19
22
76
20
22
77
20
22
78
20
22
79
20
22
80
20
22
81
20
22
82
20
22
83
20
22
84
20
22
85
20
22
86
20
22
87
20
22
88
20
22
89
20
22
90
20
22
91
20
22
92
20
22
93
20
22
94
20
22
95
20
22
96
20
22
97
20
22
98
20
22
99
20
22
100
20
22
101
20
22
102
20
22
103
20
22
104
20
22
105
20
22
106
20
22
107
20
22
108
20
22
109
21
22
110
21
22
111
21
22
112
21
22
113
21
22
114
21
22
115
21
22
116
21
22
117
21
22
118
21
22
119
21
22
120
21
22
121
21
22
122
21
22
123
21
22
124
21
22
125
21
22
126
21
22
127
21
22
128
21
22
129
21
22
130
21
22
131
21
22
132
21
22
133
21
22
134
21
22
135
21
22
136
21
22
137
21
22
138
21
22
139
21
22
140
21
22
141
21
22
142
21
22
143
21
22
144
21
22
145
21
22
146
21
22
147
21
22
148
21
22
149
21
22
150
21
22
151
21
22
152
21
22
153
21
22
154
21
22
155
21
22
156
21
22
157
21
22
158
21
22
159
21
22
160
21
22
161
21
22
162
21
22
163
21
22
164
21
22
165
21
22
166
21
22
167
21
22
168
21
22
169
21
22
170
21
22
171
21
22
172
21
22
173
21
22
174
21
22
175
21
22
176
21
22
177
21
22
178
21
22
179
21
22
180
21
22
181
21
22
182
21
22
183
21
22
184
21
22
185
21
22
186
21
22
187
21
22
188
21
22
189
21
22
190
21
22
191
21
22
192
21
22
193
21
22
194
21
22
195
21
22
196
21
22
197
21
22
198
21
22
199
21
22
200
21
22
201
21
22
202
21
22
203
21
22
204
21
22
205
21
22
206
21
22
207
21
22
208
21
22
209
21
22
210
21
22
211
21
22
212
21
22

Table 6: Trial 4 Raw Data

mass of Drierite
Iteration
Temp Lower ©
Temp Upper ©
7.157
1
22
27
2.0 L/min
2
22
27
3
22
27
4
22
27
5
22
27
6
22
27
7
22
27
8
22
27
9
22
28
10
22
28
11
22
29
12
22
30
13
22
30
14
22
31
15
22
32
16
22
32
17
22
33
18
22
33
19
22
33
20
22
34
21
22
34
22
22
34
23
22
34
24
22
35
25
22
35
26
22
35
27
22
35
28
22
34
29
22
34
30
22
34
31
22
34
32
22
34
33
22
34
34
22
33
35
22
33
36
22
33
37
22
32
38
22
32
39
22
32
40
22
31
41
22
31
42
22
31
43
22
31
44
22
30
45
22
30
46
22
30
47
22
30

Comments

Popular posts from this blog

Synthesis of Methyl Acetate through Esterification Plant Design

ECH 158C Design Project Final Report TO: Jason White FROM: Bryan Eglan, Daron Fong, Brett Koehn, and Simar Singh DATE: May 27, 2016 SUBJECT: Revised Methyl Acetate Plant Design Report     Attached to this memo is our final conceptual design and economic analysis for the methyl acetate plant project commissioned by Ester Chemicals Corporation (ECC). The purpose of this memo is to outline the revisions and improvements carried out on the previous design report. In the updated report, a distillation column previously used to separate the acetic acid and sulfuric acid in the bottoms of the first column has been removed, and the entire stream is now simply recycled back into the reactor as excess reactants and unconsumed catalyst, thus saving on capital and operating costs. The economic analysis was revised to include an higher, much more realistic capital cost, and a heat exchanger network was designed and analyzed to optimize utility savings. Lastly, improvements to the design

Drying Water and Alcohol through a Porous Medium Experiment

Department of Chemical Engineering and Materials Science University of California Davis Drying Water and Ethanol through a Porous Medium! By Brett Koehn For Professors Powell and Tseregounis ECH 155: Chemical Engineering Laboratory Group #1 Ethan Jensen, Chris Gee, and Zheyu Li Experiment Performed: January 29, 2016 Report Submitted: February 9, 2016 TABLE OF CONTENTS : Abstract……………………………………………………………………………………………3 Introduction………………………………………………………………………………………..3 Theory……………………………………………………………………………………………..5 Experimental Methods…………………………………………………………………….………7             Figure 1……………………………………………………………………………………8 Results…………………………………………………………………………………………......9             Figure 2…………………………………………………………………………………..11             Figure3………………………………………………………………………………...…11 Discussion……………………………………………………………………………………..…12 Conclusion…………………………………………………………………………………….…14

Vodka Production Experiment

Vodka Distillation, Production, and Distribution By: Brett Koehn Submitted March 17, 2015 Based on experimental work for Laboratory #6, conducted between February 25, 2015 and March 13, 2015 with group members Michael Kagan and James Makel in Section 4 of ECH 145A Table of Contents: Abstract………………………………………………………………………………………………………………………………………………..2 Introduction……………………………………………………………………………………………………………………………………………3 Theory…………………………………………………………………………………………………………………………………………………5                 -Figure 1…………………………………………………………………………………………………………………………………...6                 -Figure 2. …………………………………………………………………………………………………………………………………7 Experimental Methods………………………………………………………………………………………………………………………………...9                 -6.1…………………………………………………………..……………………………………………………………………...........10                 -6.2………………………………………………………..………………………………………………………………………...........11