Air Emission ControlCHE 435Section 7Team 1

Team Leader: Avery CarlsonExperimental Engineer: Ahmad HazamuddinAnalyst: Nina Bragg

AbstractIn this experiment, the team analyzed a packed-bed absorption column with the goal of determining the height of gas transfer units (HOG) within the column. The height of transfer units governs the effectiveness of mass transfer, and the extent of control the column has over ammonia emission to the atmosphere. The equipment operated using standard absorption techniques, i.e. liquid water flowed down the packing, while a vapor mixture of air and ammonia flowed counter-current to the water in the packing. Depending on the quality of mass transfer and the extent of vapor-liquid equilibrium (VLE), some ammonia would be absorbed into the water and exit the system. An ammonia analyzer was used to measure concentrations of unabsorbed ammonia in the effluent air stream, and titration methods were used to measure absorbed ammonia in the effluent water stream. Using a combination of the HTU-NTU method, mass balance, and Henrys law equations, the team tested the effect of air flow rate on HOG. A linear relationship was found to exist between the flow rate of inlet air and the height of the gas transfer units. This conclusion was validated with the logic that as air flow rate increases, contact time between water and ammonia decreases, and thus a higher HOG is needed for effective mass transfer to occur. Air flow rate was found to be more significant than water flow rate.

Table of ContentsIntroduction1

Theory 1

Apparatus Figure 1: Schematic diagram of the experimental setup33

Procedures 3

Calibrations Table 1: Water Calibration Data Figure 2: Water Calibration Chart Table 2: Range of Non-Flooding Combination of Air and Water Flow Rates Figure 3: The relationship of Air Flow Rates and Pressure Drop Table 3: Range of Non-Flooding Air and Water Flow Rates Combination Figure 4: Range of Non-Flooding Air and Water Flow Rates Combination5556677

Results Figure 5: Comparison of HOG values for multiple air and water flow rates Table 4: Analysis of error calculations for each air flow rate Table 5: Analysis of error calculations for each water flow rate Table 6: Error propagation for each water flow rate. Table 7: Error propagation for each air flow rate.888999

Discussion9

Conclusions and Recommendations10

Notation 10

References11

Appendix Table A-1: Calculation for the amount of ammonia and the mole fraction of ammonia Table A-2: Calculation of NOG and HOG Table A-3: Error propagation of NOG for each water flow rate. Table A-4: Error propagation of NOG for each air flow rate. Table A-5: Error propagation of HOG for each water flow rate. Table A-6: Error propagation of HOG for each air flow rate Figure A-1: Comparison of NOG values for multiple air and water flow rates Figure A-2: Graph of HOG at various water flow ratesIIIIIIIIIVIVIVVV

V

IntroductionResearchers expanding knowledge of pollutants has led to air emission control becoming an increasingly important function on the industrial scale. In todays society, industry understands the detrimental effects pollutants, greenhouse gases, and other waste products can have to ecology, the environment, and general public health. Absorption methods, such as packed-bed absorption, have been particularly effective in reducing the amount of undesired chemicals that are released to the atmosphere. In this study, the design team looked at one such type of emission control using absorption methodthe process of absorbing ammonia into water. Ammonia is a chemical that is commonplace in industrial processes. Industries ranging from agricultural to pharmaceutical to plastics utilize ammonia to design products (1). However, the use of ammonia has certain drawbacks, especially when unused ammonia is released to the atmosphere. Ammonia pollution is destructive to water ecosystems, can toxify soil and groundwater, and poses a health risk to humans (1). Absorption into water allows for easier disposal of excess ammonia through means that are environmentally friendly, and is a relatively simple process to understand. The objective of this experiment was to calculate the height of gas transfer units (HOG) within a packed-bed absorption column as a function of the air flow rate entering the column. In order to complete this task, the design team took measurements of the concentrations of ammonia entering and exiting the column in order to understand the extent of mass transfer within the packing. The results of experiments demonstrated a positive linear relationship between the HOG and the air flow rate entering the column, e.g as air flow rate increased so did HOG. This is shown from the resulting data from the design of experiments. For every trial, as air flow rate increased from two to four SCFM, HOG increased. At a water flow rate of 400 cubic centimeters per minute, HOG increased from 4.85 inches to 6.71 inches. The next two trials showed nearly the exact same data. As air flow rate increases from two to four SCFM, HOG increased from 4.24 to 5.30 inches. In the opposite manner, HOG is shown to not be heavily dependent on water flow rate. When the water flow rate increases from 500-600 cc/min, the data varies little. HOG for both cases increases about 4.24 to 6.71 inches, which shows that water flow rate is not as significant a variable as air flow rate. This is concurrent with results found in a University of Florida (2, pg. 21). TheoryThe theory behind the calculation of HOG in packed-bed absorption columns is based off the HTU-NTU method described in Separation Process Engineering by Philip Wankat. The fundamentals of this method revolve around mass transfer between solute and solventin this case ammonia and water, respectively. The underlying assumptions are that the molecules of water are not counter-diffusing into the ammonia vapor, i.e. all water fed to the system ends up as liquid at the bottom of the column, and that the system is isothermal (3).A diagram of a general absorption column is shown in Figure 1. Pure water enters the column through the top of the column, while an air-ammonia mixture is fed through the bottom. As the water contacts the ammonia in the packing, mass transfer of ammonia occurs into the film of water, which falls to the bottom of the column for removal. Some unabsorbed ammonia exits the column through the top. The first step in the analysis is performing the mass balance of ammonia around the column. Mass is conserved so the mass balance becomes (3, pg. 5): [1]where L = liquid flow rate of water [mol/min] V = vapor flow rate of mixed air and ammonia [mol/min] xB = mole fraction of ammonia in water feed stream [unitless] xA = mole fraction of absorbed ammonia in liquid exit stream [unitless] yA = mole fraction of mixed ammonia in air inlet stream [unitless] yB = mole fraction of unabsorbed ammonia in exit air stream [unitless]In this case, there is no ammonia entering with the pure water, therefore xB is equal to 0 (3, pg. 5). The equation is simplified to: [2]For this system, the team assumedand later verifiedthat the amount of ammonia entering the column in the inlet was dilute compared to the overall air flow rate. The team determined that it was applicable for this dilute mass transfer problem to employ Henrys Law to estimate the equilibrium between ammonia in the liquid and vapor (3, pg. 11). Henrys Law is described below: [3]where y* = the vapor phase mole fraction that is in equilibrium [unitless] H = Henrys Law Constant Using partial pressure data of ammonia in Perrys Chemical Handbook, the Henry constant can be found. Calculating the number of gas transfer units, NOG, is the first step of the HTU-NTU method (4, pg. 686). Taking the integral of the mole fractions in the inlet and outlet streams allows for this calculation: [4]where NOG = number of gas transfer units within the column [unitless] Essentially, this simplifies to the log mean difference between the inlet and outlet mole fractions, and the mole fraction in equilibrium within the column or (4, pg. 686): [5]The final step in the method is to relate NOG to HOG. According to Wankat, HOG is inversely proportional to the found value of NOG (4, pg. 687). The equation is below: [6] where HOG = height of the gas transfer units within the column [inches] ZT = height of packing within the column [inches]

Apparatus

Figure 1: Schematic diagram of the experimental setup.Figure 1 shows the equipment set up of the experiment. The equipment set up was composed of a packed column absorber which was packed with approximately 5 L, 4.3 kg of ceramics Raschigrings. Water was fed to the top of the columnwhileair to the bottom. Ammonia gas was mixed with air before entering the column. The flow rates of both water and air were controlled using rotameters. The set up also came with an ammonia analyzer with infrared continuously reading the amount of ammonia in air. The overhead product, which was the unabsorbed ammonia, was sent through water trap to filter the water out before it was sent to the ammonia analyzer.

ProceduresThe Material Safety Data Sheet (MSDS) of ammonia was studied to be aware of safety issues when handling aqueous ammonia. Goggles were worn while being in the lab. Calibration of water flow rotameter was executed. The water flow rate was first set at a low value at 1 SCFM. The team allowed a few minutes for a water seal to form at the bottom of the column. This was to ensure a consistent flow rate of effluent water. The time needed to collect 1000 mL of water in a flask was taken. This experiment was repeated with higher water flow rates over the same increment. These flow rates were then compared to the flow rates set on the rotameter.The water flow rate was calibrated.Next, the relationship of pressure drop and air flow rate was investigated. The water rotameter was opened to form a water seal at the bottom of the column.It was adjusted accordingly after changing the air flow rate tohave the water seal at a constant height. This was to ensure that no air exits through the bottomof the column. The air flow rate was set starting from a low value at and then increased over the same increment which. The team allowed the column to achieve a steady state. The pressure drop across the packed column measured by a computer software was recorded. The experiment was repeated and data was taken for multiple air flow rates in a descending order.The range of flow rates that causes flooding was also obtained. In this experiment, the water flow rate was held constant and the air flow rate was increased by 0.5 SCFM until flooding occurs. Flooding occurs when there is a big increase in pressure drop across the column after increasing the air flow rate. The experiment was then continued by following the same procedure with four larger water flow rates. The pressure drop of all combinations of water and air flow rates were recorded. This data was to be used for the ammonia absorption experiment as a reference to avoid flooding.Gloveswere worn throughout the whole experiment when handling ammonia. Any water left in the column was drained out of the column to assure that the initial water condition was not contaminated by ammonia from a previous trial. The water rotameter was first opened at a flow rate of 400 cc/min. The team allowed a few minutes to pass in order for a water seal to form at the bottom of the column at a constant height of 8.5 cm. The rotameter of air flow rate was set at 4 SCFM and the ammonia rotameter was kept constant at 78 mm throughout the experiment. 8-10 minutes were allowed for the system to achieve equilibrium. The water seal was drained out and formed again to ensure the concentration of ammonia in the water was at equilibrium and not contaminated by accumulation of ammonia. A sample of 50 mL effluent water was collected in a beaker for titration. These steps were repeated with increased air flow rates of 5 and 6 SCFM. Then, the whole experiment was then repeated using higher water flow rates at 500 and 600 cc/min. All combinations of water and air flow rates used in this experiment should not cause flooding and this can be referred to the data obtained in the previous flooding experiment.For titration, 0.5 M HCl was prepared in the burette. Three drops of methyl orange were added into the beaker. Hydrochloric acid was added drop by drop into the beaker until the color of methyl orange changes from yellow to red, which was the equivalence point. The mixture was stirred throughout the process. The amount of HCl solution needed was recorded.

Calibrations

Table 1: Water Calibration DataRotameter Reading (cc/min)Time Required to Collect 1000ml (min)Effluent Water flow Rate (cc/min)

2750.580.00058

5000.860.00086

7250.870.00087

9750.910.00091

11500.900.00090

12500.900.00090

Figure 2: Water Calibration Chart

Table 2: Range of Non-Flooding Combination of Air and Water Flow RatesAir Flow Rate (SCFM)Comment by Hazamuddin, Ahmad Hafizul Aimran B: I changed the values of pressure dropPressure Drop (cm H2O)

12.08

22.69

33.44

45.05

510.92

Figure 3: The relationship of Air Flow Rates and Pressure Drop

Table 3: Range of Non-Flooding Air and Water Flow Rates CombinationH2O Flow Rate (cc/min)Maximum Air Flow Rate (SCFM)

23605

27605.5

30006

33405

36105

Figure 4: Range of Non-Flooding Air and Water Flow Rates Combination

ResultsResults for the experiments are shown in Figure 5 and Figure A-1.

Figure 5: Comparison of HOG values for multiple air and water flow rates.

Linear trendlines fit the data for all three water flow rates used in this experiment. HOG as a function of just water flow rate is shown in the Appendix (Figure A-2).

Error AnalysisAnalysis of error was performed for HOG as a function of water flow rate and air flow rate. The tables below show the results. An explanation for how the percent error values are calculated is included in the Appendix.Table 4: Analysis of error calculations for each air flow rate. Calculated values correspond to Figure 5.Water Flow Rate (cc/min)Air Flow Rate (SCFM)HOG,calcHOG,expHOG,calc - HOG,exp% Error

40046.2856.2680.01730.275

35.5765.6100.03480.625

24.8664.8490.01740.357

50046.2855.2831.00215.939

35.5764.7760.79914.338

24.8664.2390.62712.882

60046.2855.2980.98715.704

35.5764.8670.70912.710

24.8664.2420.62412.830

Table 5: Analysis of error calculations for each water flow rate. Values correspond to Figure A-2.Water Flow Rate (cc/min)HOG,calcHOG,expHOG,calc - HOG,exp% Error

4005.2285.5760.3486.650

5004.2184.7660.54812.998

6004.0084.8020.79419.817

As the water flow rate increases, the percent error increases.

Propagation of ErrorBelow are the tables for the calculated propagation of error. See Appendix for further details on how error propagation was calculated.Table 6: Error propagation for each water flow rate.Water Flow Rate (cc/min)(NOG)(HOG)

4000.01780.251

5000.01780.214

6000.01780.220

Table 7: Error propagation for each air flow rate.Air Flow Rate (SCFM)(NOG)(HOG)

40.002020.184

30.002070.166

20.002290.146

DiscussionThe value for the water flow rate used in the calculations was the rotameter value. The trendline for the calibration curve did not fit the data well (see Figure 2). Instead of the effluent water flow rate continually increasing with increasing rotameter flow rate, the effluent water flow rate started to decrease after around 900 cc/min. The increasing pressure from an increased water flow rate should have made the effluent flow rate larger. Since this was not the case for the calibration data, it was decided to use the value from the rotameter instead of the calibrated value for the NOG and HOG calculations.Comment by Hazamuddin, Ahmad Hafizul Aimran B: I moved this paragraph up as it is part of calibration In the second experiment of calibration, it can be concluded that changing air flow rate has a significant effect on the pressure drop. The pressure drop increases exponentially with the air flow rates. The value of pressure drop escalated significantly from 5.05 cm H2O to 10.92 cm H2O when increasing the air flow rate from 4 SCFM to 5 SCFM, as shown in Figure 3, indicating flooding occurred. This data proves that increasing air flow rates above 4 SCFM would cause flooding.

Figure 4 revealed the range of allowable air and water flow rates that limit the occurrence of flooding. The air flow rates indicated in the chart are the maximum air flow rate for respective water flow rates before flooding. The maximum air flow rate before flooding is 6 SCFM whereas the larger flow rates of water max out at 5 SCFM. Any larger or smaller water flow rates than 3000 cc/min would require smaller air flow rates than 5 SCFM to avoid flooding.As the air flow rate increases, the number of gas-phase transfer units (NOG) decreases (see Figure A-1). This is consistent at all water flow rates. This occurs because there is less contact time between the water and ammonia at higher air flow rates. The opposite trend is shown in Figure 5 for the height of a gas-phase transfer unit (HOG). As the air flow rate becomes larger, the height of a gas-phase transfer unit increases as well. This indicates that the separation occurs more efficiently at smaller air flow rates and higher air flow rates induce resistance to mass transfer. If there is a smaller number of transfer units, then the transfer unit has to be larger in order for the desired absorption to occur. There is an inverse relationship between the height of a gas-phase transfer unit and the number of gas phase transfer units. This is shown in the results and by Equation 6. It is observed that the air flow rate has more of an impact on NOG and HOG than the water flow rate.A significant difference of HOG can be observed when changing the water flow rate from 400 to 500 cc/min compared to that when increasing it from 500 to 600 cc/min. This concludes increasing the water flow rate above 500 cc/min does not result in a significant decrease to HOG. However, at the air flow rate of 3 SCFM, HOG increases by a small amount when increasing the water flow rate from 500 to 600 cc/min. This might be the result of few errors during the experiment. The rotameter readings kept fluctuating that needed to be adjusted accordingly most of the time. Moreover, since titration is part of this process, the amount of HCl added to achieve the equivalence point might have exceeded the actual amount required.The error analysis calculations show a wide range of percent error. This large fluctuation of error could be due to fluctuations in the ammonia rotameter or the water rotameter. The ammonia and water flowrate are assumed to be constant for each trial. Fluctuations in flow rates would change the amount of ammonia coming out of the bottom and top of the absorber, and in turn, affect the error associated with the team's calculation method.

Conclusion & RecommendationsThe height of a gas-phase transfer unit in a packed-bed absorber is affected by the air flow rate and the water flow rate. 1. As the water flow rate increases, the height of a gas-phase transfer unit decreases.2. As the air flow rate increases, the height of a gas-phase transfer unit increases.3. At water flow rates of 500 cc/min and 600 cc/min the values for HOG varied little.4. NOG and HOG have an inverse relationship. When NOG decreases, HOG increases.It is recommended that the knobs on the rotameters either be tightened or changed. The location of the level-indicator piece fluctuated from the set value during the experiment. This occurred most often with the ammonia rotameter. The flow rate used for calculations may not be the real flow rate of ammonia being mixed with air, therefore making the team's calculations not as accurate. It is also recommended that a more dilute acid be used for the titrations. Adding one drop of acid when the sample was near equivalence brought the sample past its equivalence point and by the looks of the indicator into the acidic range. If a more dilute acid was used there would be less moles going into the sample at any one point in time, making the titration more accurate.

Notation cc/min = Cubic centimeters per min H = Henrys Law Constant HOG = height of the gas transfer units within the column [inches] L = liquid flow rate of water [mol/min] NOG = number of gas transfer units within the column [unitless] SCFM = Standard cubic feet per min V = vapor flow rate of mixed air and ammonia [mol/min] xB = mole fraction of ammonia in water feed stream [unitless] xA = mole fraction of absorbed ammonia in liquid exit stream [unitless] yA = mole fraction of mixed ammonia in air inlet stream [unitless] yB = mole fraction of unabsorbed ammonia in exit air stream [unitless] y* = the vapor phase mole fraction that is in equilibrium [unitless] ZT = height of packing within the column [inches]References[1] Scottish Environment Protection Agency. (n.d.). Scottish Pollutant Release Inventory. [Online]. Available: http://apps.sepa.org.uk/spripa/Pages/SubstanceInformation.aspx?pid=1. [2] McCabe, W.L., Smith, J.C., & Harriott, P. Unit Operations of Chemical Engineering, 7th ed. McGraw-Hill, New York, 2005.[3] Wankat, P.C. Separation Process Engineering, 3rd ed. Prentice Hall, Massachusetts, 2012.[4] Crisalle, O.D. University of Florida. (2013). Ammonia Gas Absorption. [PDF]. Retrieved from http://www.che.ufl.edu/unit-ops-lab/experiments/GA/GA-NH3/GA-NH3-Manual.pdf.[5] Barton, S.C., & Odian, M. Columbia University. (n.d.). Packed Bed Fluid Dynamics and Ammonia Absorption. [Online]. Retrieved from http://www.columbia.edu/itc/seas/E3810-lab/pbd.html.[6] Nagy, Gabriella. Fundamentals Laboratory Booklet: CHE 34800, Purdue University, 2015.

Appendix

Sample CalculationsThe following is a sample calculation for how to determine the height of a gas-phase transfer unit.

Calculation Tables

Table A-1: Calculation for the amount of ammonia and the mole fraction of ammonia.Water Flow Rate (mol/min)Air Flow Rate (mol/min)Ammonia Flow Rate (mol/min)Amount of NH3 (mol)Amount of H2O (mol)Mole Fraction of NH3 (Bottom)Mole Fraction of NH3 (Top)

Trial 122.24.790.06670.004252.780.001530.000129

22.23.590.06670.004752.780.001710.000103

22.22.400.06670.005052.780.001820.0000730

Trial 227.84.790.06670.004102.780.001480.0000610

27.83.590.06670.004052.780.001460.0000480

27.82.400.06670.004402.780.001590.0000380

Trial 333.34.790.06670.003302.780.001190.0000590

33.33.590.06670.003402.780.001230.0000500

33.32.400.06670.003502.780.001260.0000350

Table A-2: Calculation of NOG and HOG.yAyBxy*yA-y*yB-y*NOGHOG (in)

Trial 10.01370.0001290.001531.46E-050.01370.0001144.796.27

0.01820.0001030.001711.64E-050.01820.00008665.355.61

0.02710.00007300.001821.74E-050.02710.00005566.194.85

Trial 20.01370.00006100.001481.41E-050.01370.00004695.685.28

0.01820.00004800.001461.39E-050.01820.00003416.284.78

0.02710.00003800.001591.52E-050.02710.00002287.084.24

Trial 30.01370.00005900.001191.14E-050.01370.00004765.665.30

0.01820.00005000.001231.17E-050.01820.00003836.164.87

0.02710.00003500.001261.21E-050.02710.00002297.074.24

Error AnalysisAfter the data was plotted, a trendline was fitted to the data for each air and water flow rate. HOG was calculated using the trendline equations for a given flow rate. The experimental value for HOG was subtracted from the calculated value to find the difference between the two. This difference was divided by the calculated HOG value and multiplied by 100 percent in order to find the percent error.

Error PropagationThe error propagation for NOG and HOG are calculated. To determine the values for error propagation, the following equation was used (6, p. 11)

In order to find NOG, three variables were used: yA, yB, and x. It was assumed that the value for H remained constant throughout the experiment. The detailed equations used to find the error values for NOG and HOG are shown below.

In the above equations N represents the number of samples; in this experiment there were a total of nine samples. Below are detailed tables of the intermediate steps of the error propagation calculations.

Table A-3: Error propagation of NOG for each water flow rate.Water Flow Rate (cc/min)Avg yA(yA)dNOG/dyAAvg yB(yB)dNOG/dyB

4000.01970.0067954.9580.0001022.802E-05-12756.337

5000.01970.0067954.9530.00004901.153E-05-31489.532

6000.01970.0067954.9450.00004801.212E-05-30229.887

Table A-3 continuedWater Flow Rate (cc/min)Avg x(x)dNOG/dx2(NOG)(NOG)

4000.001690.000146121.4010.0003190.0178

5000.001516.821E-05300.4560.0003180.0178

6000.001233.603E-05288.4160.0003180.0178

Table A-4: Error propagation of NOG for each air flow rate.Air Flow Rate (SCFM)Avg yA(yA)dNOG/dyAAvg yB(yB)dNOG/dyB

40.0137072.9480.00008303.985E-05-17022.773

30.0182054.9500.00006703.119E-05-22341.666

20.02714.249E-1836.9594.87E-052.113E-05-35111.317

Table A-4 continuedAir Flow Rate (SCFM)Avg x(x)dNOG/dx2(NOG)(NOG)

40.001400.000184162.0094.065E-060.00202

30.001470.000243213.0194.293E-060.00207

20.001560.000280335.2455.255E-060.00229

Table A-5: Error propagation of HOG for each water flow rate.Water Flow Rate (SCFM)Avg NOG(NOG)dHOG/dNOG2(HOG)(HOG)

4005.4400.705-1.0140.06300.251

5006.3450.702-0.7450.04580.214

6006.3000.715-0.7560.04830.220

Table A-6: Error propagation of HOG for each air flow rate.Air Flow Rate (SCFM)Avg NOG(NOG)dHOG/dNOG2(HOG)(HOG)

45.3760.510-1.0380.03380.184

35.9310.509-0.8530.02760.166

26.7790.512-0.6530.02140.146

Plots Not Included in the Results Section

Figure A-1: Comparison of NOG values for multiple air and water flow rates.

Figure A-2: Graph of HOG at various water flow rates.A polynomial trendline provided the best fit for HOG as a function of water flow rate.