DISTILLATION: McCABE-THIELE DIAGRAMS AND SHORTCUT METHODS ChE 3G4 Spreadsheet Distillation columns can typically be described by the schematic diagram shown to In designing a column, we can identify two practical limiting cases for the reflux with a vapour flow rate of V and liquid flow rate of L. A feed stream of molar flow F, mole fraction composition zi and quality q (q=0 is a saturated vapour; q=1 is a saturated liquid) is fed to the column at an optimized tray NFEED. The vapour from the top of the column (molar flow rate V) is totally condensed, with part of the condensate returned to the column (molar flow rate L) and part removed as a distillate product (molar flow rate D with mole fraction composition xi,DIST). Similarly, the liquid from the bottom of the column is partially reboiled back to the column, with the remaining liquid portion removed as a bottoms product (molar flow rate B with a mole fraction composition xi,BOT). Unlike in a flash drum, the product distillate and bottoms streams are NOT themselves in equilibrium, only the vapour and liquid compositions of each single tray. desired product compositions. As a result, it's important we find a way to design columns to meet specific product stream specifications. A variety of methods can be used, the most obvious of which is performing tray-by-tray balances within the column. Since each tray is at a constant pressure, this essentially amounts to performing a flash calculation on each single tray up and down the column. While computer simulation programs can do such calculations, they are very cumbersome and difficult to do with a spreadsheet. However, we can use shortcut methods to estimate the number of trays and external (L/D) and internal (L/V) reflux ratios required to produce product streams of specified compositions. (1) L=V; that is, we do not take any distillate (or bottoms) product In this case, we need a minimum number of trays to perform the separation since we number of trays (Nmin) can be predicted using the Fenske equation, expressed either and B or the fractional recoveries (FR) of components A and B in the distillate an (ie. a component present in both the distillate and bottoms but recovered primaril (ie. a component present in both the distillate and bottoms but recovered primaril two components whose recovery in the distillate and/or bottoms is specified in the The parameter aAB in these equations is the relative volatility of A with respect where PA* and PB* are the vapour pressures of N min = ln { ( FR A ) DIST ( FR B ) BOT [ 1−( FR A ) DIST ][ 1−( FR B ) BOT ] } ln α AB N min = ln { ( x A x B ) DIST / ( x A x B ) BOT } ln α AB α AB = y A x A y B x B ≈ P ¿ A P P ¿ B P = P A ¿ P B ¿
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
DISTILLATION: McCABE-THIELE DIAGRAMS AND SHORTCUT METHODSChE 3G4 Spreadsheet
Distillation columns can typically be described by the schematic diagram shown to the right.
In designing a column, we can identify two practical limiting cases for the reflux ratios L/V and L/D:
A column contains N trays, each of which is at a particular temperature and pressure. Vapour-liquid equilibrium is established across each of these trays, with a vapour flow rate of V and liquid flow rate of L. A feed stream of molar flow F, mole fraction composition zi and quality q (q=0 is a saturated vapour; q=1 is a saturated liquid) is fed to the column at an optimized tray NFEED. The vapour from the top of the column (molar flow rate V) is totally condensed, with part of the condensate returned to the column (molar flow rate L) and part removed as a distillate product (molar flow rate D with mole fraction composition xi,DIST). Similarly, the liquid from the bottom of the column is partially reboiled back to the column, with the remaining liquid portion removed as a bottoms product (molar flow rate B with a mole fraction composition xi,BOT). Unlike in a flash drum, the product distillate and bottoms streams are NOT themselves in equilibrium, only the vapour and liquid compositions of each single tray.
Distillation processes are frequently used in industry to do perform well-defined separations, often using a cascade of columns in sequence to achieve the desired product compositions. As a result, it's important we find a way to design columns to meet specific product stream specifications. A variety of methods can be used, the most obvious of which is performing tray-by-tray balances within the column. Since each tray is at a constant pressure, this essentially amounts to performing a flash calculation on each single tray up and down the column. While computer simulation programs can do such calculations, they are very cumbersome and difficult to do with a spreadsheet. However, we can use shortcut methods to estimate the number of trays and external (L/D) and internal (L/V) reflux ratios required to produce product streams of specified compositions.
(1) L=V; that is, we do not take any distillate (or bottoms) product
In this case, we need a minimum number of trays to perform the separation since we have no incoming or outgoing flow. This minimum number of trays (Nmin) can be predicted using the Fenske equation, expressed either in terms of the mole fractions (x) of components A and B or the fractional recoveries (FR) of components A and B in the distillate and bottom stream. Here, A is the light key component (ie. a component present in both the distillate and bottoms but recovered primarily in the distillate) and B is the heavy key component (ie. a component present in both the distillate and bottoms but recovered primarily in the bottoms). The light and heavy keys are the two components whose recovery in the distillate and/or bottoms is specified in the problem in a multi-component distillation problem:
The parameter aAB in these equations is the relative volatility of A with respect to B, which (assuming ideal conditions) can be estimated as:
where PA* and PB* are the vapour pressures of components A and B (from Antoine's equation)
A key assumption to this approach is that the relative volatility is constant throught the entire column, despite the fact that a range of temperatures are present on the different trays. Several approaches are taken to get the "best" estimate of this average relative volatility; in this spreadsheet, the dew and bubble temperatures are calculated and the partial pressure ratio at the midpoint of these temperatures is used.
(1) L/Dmin - the external reflux ratio at which the specified separation is just achieved with an infinite number of trays
As we continue to take more and more product off the top, we reduce the amount of product returned to the column and consequently reduce the total time an average molecule stays within the column. Eventually, we reach a point where an "infinite" number of stages is required to separate the components according to the column specifications. This minimum external reflux ratio L/Dmin can be predicted using the Underwood equations given below:
To solve these equations, we first find the value of the parameter f which satisfies equation (1). We can then substitute this value into (2) to calculate Vmin. D can be calculated based on the feed flow and composition and the specified mole fractions and/or fractional recoveries of the components. Lmin can be calculated using equation 3. These equations can be used for any number of different components within the column; however, if solved in this fashion (ie. with a single value of f), it is required to assume that any light or heavy non-key components (ie. components with relative volatilities higher and lower than the light and heavy key components respectively) either do not distribute at all (ie. all light non-key ends up in the distillate and all heavy non-key in the bottoms) or distribute according to the Fenske equation prediction:
We can then use these extreme results (ie. the minimum possible number of trays and the minimum possible external reflux ratio) to predict the number of trays and reflux ratio required in a real column using the Gilliland correlation. A design value for L/D is first specified, usually as some factor of (L/D)min (typical values are between 1.05 and 1.25). The following values are then calculated and fit to the correlation developed by Gilliland:
We can therefore solve for N as well as the optimum feed location in a real column, based on the N/Nmin scaling and the Fenske optimum feed prediction:
Practically, the Fenske-Underwood-Gilliland approach gives rough, first-pass estimates of the number of stages required to perform a given separation. However, the assumption of constant relative volatility can be inaccurate in some cases, particularly in columns with highly non-ideal components and/or a large temperature range.
ΔV FEED=∑i
αi Fziα i−φ
=F (1−q ) V min=∑i
αiDxi ,DISTαi−φ
Lmin=Vmin−D
N F ,min=
ln {( xAxB )DIST /(zAzB )}
lnα AB
N F=N F ,min( NNmin )
x=Abscissa=
LD
−(LD)min
LD
+1
α AB=
yA
xA
yBxB
≈
P¿A
P
P¿B
P
=PA
¿
PB¿
(FRC )DIST=αCB
N min
(FRB)BOT1−(FRB )BOT
+αCB
Nmin
y=N−Nmin
N+1
Feed Line:
Top Operating Line:
Bottom Operating Line:
We also plot the simple y=x line on the graph. A typical McCabe-Thiele plot is shown below (with the polynomial y vs x VLE fit)
This spreadsheet will allow you to use both the Fenske-Underwood-Gilliland and the McCabe-Thiele approach to design distillation columns.
Stage-by-stage calculations can be performed graphically using the McCabe-Thiele method. In this approach, the y vs x VLE relationship is plotted directly on the graph, eliminating the uncertainty regarding the constant relative volatility estimate. This means we need to perform dew or bubble point calculations to generate the y vs x equilibrium data; in this spreadsheet, we use an ideal bubble point calculation to produce this data. The curve is then fit to a fourth-order polynomial expression in order to give us an algebraic expression for the y vs x equilibrium line, allowing us to calculate its intercepts with the other lines. On the same graph, we plot three additional lines:
Starting at the specified mole fraction of component A (the lighter component) in the distillate on the y=x line, we can step down the curves, using the y vs. x VLE equilibrium data curve and either the top or bottom operating lines as our step limits. The top operating line is used when the component A mole fraction is greater than the x value of the feed line - y vs x VLE curve intercept; the bottom operating line is used at mole fractions below the intercept. This method can be used to design any column with any specifications (ie. we are not limited to total reflux or minimum reflux). However, by setting L/V equal to one, we can use the McCabe-Thiele diagram to check that the Fenske calculation of Nmin is accurate. In the case of minimum reflux, we can obviously not plot an infinite number of stages using the McCabe-Thiele method; however a "pinch point" will be visible on the graph in which the operating and equilibrium lines touch.
yA=LV
xA+(1− LV ) xA ,DIST
yA=q
q−1xA+
11−q
zA ,FEED
yA=LV
xA+(1− LV ) xA ,BOT
Cells highlighed in YELLOW require input from youCells highlighted in BLUE require you to perform a manual GoalSeek procedure on that cell to get a converged solution
Cells with a red triangle in the upper right-hand corner have comments which will give you more information about what the variable in the cell means or how to select a value for that variable.
where
In this case, we need a minimum number of trays to perform the separation since we have no incoming or outgoing flow. This minimum number of trays (Nmin) can be , expressed either in terms of the mole fractions (x) of components A and B or the fractional recoveries (FR) of components A and
B in the distillate and bottom stream. Here, A is the light key component (ie. a component present in both the distillate and bottoms but recovered primarily in the distillate) and B is the heavy key component (ie. a component present in both the distillate and bottoms but recovered primarily in the bottoms). The light and heavy keys are the two components whose recovery in the distillate and/or bottoms is specified in the problem in a multi-component distillation problem:
in these equations is the relative volatility of A with respect to B, which (assuming ideal conditions) can be estimated as:
* are the vapour pressures of components A and B (from Antoine's equation)
A key assumption to this approach is that the relative volatility is constant throught the entire column, despite the fact that a range of temperatures are present on the different trays. Several approaches are taken to get the "best" estimate of this average relative volatility; in this spreadsheet, the dew and bubble temperatures are
- the external reflux ratio at which the specified separation is just achieved with an infinite number of trays
As we continue to take more and more product off the top, we reduce the amount of product returned to the column and consequently reduce the total time an average molecule stays within the column. Eventually, we reach a point where an "infinite" number of stages is required to separate the components according to the column
Underwood equations given below:
which satisfies equation (1). We can then substitute this value into (2) to calculate Vmin. D can be calculated based on the feed flow and composition and the specified mole fractions and/or fractional recoveries of the components. Lmin can be calculated using equation 3. These equations can be used for any number of different components within the column; however, if solved in this fashion (ie. with a single value of f), it is required to assume that any light or heavy non-key components (ie. components with relative volatilities higher and lower than the light and heavy key components respectively) either do not distribute at all (ie. all light non-key ends up in the distillate and all heavy non-key in the bottoms) or distribute according to the Fenske
We can then use these extreme results (ie. the minimum possible number of trays and the minimum possible external reflux ratio) to predict the number of trays and reflux ratio required in a real column using the Gilliland correlation. A design value for L/D is first specified, usually as some factor of (L/D)min (typical values are between 1.05 and 1.25). The following values are then calculated and fit to the correlation developed by Gilliland:
We can therefore solve for N as well as the optimum feed location in a real column, based on the N/Nmin scaling and the Fenske optimum feed prediction:
Practically, the Fenske-Underwood-Gilliland approach gives rough, first-pass estimates of the number of stages required to perform a given separation. However, the assumption of constant relative volatility can be inaccurate in some cases, particularly in columns with highly non-ideal components and/or a large temperature range.
Lmin=Vmin−D
We also plot the simple y=x line on the graph. A typical McCabe-Thiele plot is shown below (with the polynomial y vs x VLE fit)
This spreadsheet will allow you to use both the Fenske-Underwood-Gilliland and the McCabe-Thiele approach to design distillation columns.
Stage-by-stage calculations can be performed graphically using the McCabe-Thiele method. In this approach, the y vs x VLE relationship is plotted directly on the graph, eliminating the uncertainty regarding the constant relative volatility estimate. This means we need to perform dew or bubble point calculations to generate the y vs x equilibrium data; in this spreadsheet, we use an ideal bubble point calculation to produce this data. The curve is then fit to a fourth-order polynomial expression in order to give us an algebraic expression for the y vs x equilibrium line, allowing us to calculate its intercepts with the other lines. On the same graph, we plot three additional
Starting at the specified mole fraction of component A (the lighter component) in the distillate on the y=x line, we can step down the curves, using the y vs. x VLE equilibrium data curve and either the top or bottom operating lines as our step limits. The top operating line is used when the component A mole fraction is greater than the x value of the feed line - y vs x VLE curve intercept; the bottom operating line is used at mole fractions below the intercept. This method can be used to design any column with any specifications (ie. we are not limited to total reflux or minimum reflux). However, by setting L/V equal to one, we can use the McCabe-Thiele diagram to
is accurate. In the case of minimum reflux, we can obviously not plot an infinite number of stages using the McCabe-Thiele method; however a "pinch point" will be visible on the graph in which the operating and equilibrium lines touch.
Cells highlighed in YELLOW require input from youCells highlighted in BLUE require you to perform a manual GoalSeek procedure on that cell to get a converged solution
Cells with a red triangle in the upper right-hand corner have comments which will give you more information about what the variable in the cell means or how to select a
A118
: Sample Comment
ANTOINE EQUATION COEFFICIENTS
Use the numbers in column B to choose your components in the subsequent spreadsheets
No. Substance Formula Range (ºC) A B C
1 Acetaldehyde -0.2 to 34.4 8.00552 1600.017 291.809
2 Acetic Acid 29.8 to 126.5 7.38782 1533.313 222.309
3 Acetic Acid 0 to 36 7.18807 1416.700 225.000
4 Acetic Anhydride 62.8 to 139.4 7.14948 1444.718 199.817
5 Acetone -12.9 to 55.3 7.11714 1210.595 229.664
6 Acrylic Acid 5.65204 648.629 154.683
7 Ammonia 7.55466 1002.711 247.885
8 Aniline 7.32010 1731.515 206.049
9 Benzene 6.89272 1203.531 219.888
10 n-Butane 6.82485 943.453 239.711
11 i-Butane 6.78866 899.617 241.942
12 1-Butanol 7.36366 1305.198 173.427
13 2-Butanol 7.20131 1157.000 168.279
14 1-Butene 6.53101 810.261 228.066
15 Butyric Acid 8.71019 2433.014 255.189
16 Carbon disulfide 6.94279 1169.110 241.593
17 Carbon tetrachloride 6.87926 1212.021 226.409
18 Cholorobenzene 0 to 42 7.10690 1500.000 224.000
19 Cholorobenzene 42 to 230 6.94504 1413.120 216.000
: Type the number in this column (corresponding to the chemicals you want to analyze) in cells C5 and G5 (Ideal T - 2 cpts) or cells C5, G5, and K5 (Ideal T - 3 cpts)
: Mole fraction of component A in the feed to the column
G12
: Mole fraction of component B in the feed
D14
: Input any temperature here - compare P* values for the two components to determine which is A and which is B
B15
: Antoine Equation
F15
: Antoine Equation
C22
: Set as a constant over the entire column
B26
: Bubble Point - the temperature at which the first bubble of vapour forms (or, the bottom line on a two component phase diagram at a given composition)
C27
: GoalSeek in this cell to solve the bubble T equation -- set this cell (C21) to the pressure at which you want to evaluate the bubble point (P=760 mm Hg is atmospheric pressure) by changing the temperature (cell C20)
B28
: Calculate the vapour pressure of component 1 at the calculated bubble point pressure
B29
: Calculate the vapour pressure of component 2 at the calculated bubble point pressure
F29
: Under VLE conditions, the sum of the partial pressures should equal the total pressure of the system (cell C21) - this is a check that your bubble T calculation is correct
C30
: Mole fraction of component 1 in the first bubble of vapour produced - ratio of the vapour pressure of component 1 and the total pressure
C31
: Mole fraction of component 2 in the first bubble of vapour produced - ratio of the vapour pressure of component 2 and the total pressure
B40
: Dew Point - the temperature at which the first drop of liquid forms (or, the top line on a two component phase diagram at a given composition)
C41
: GoalSeek in this cell to solve the dew T equation -- set this cell (C39) to 1 (the sum of the total liquid mole fractions) by changing the dew point temperature (cell C38)
C42
: Mole fraction of component 1 in the first drop of liquid produced - calculated by Raoult's Law (x=yP/P*)
C43
: Mole fraction of component 2 in the first drop of liquid produced - calculated by Raoult's Law (x=yP/P*)
C48
: Average of dew and bubble point - good estimate of average column temperature
B54
: Antoine Equation
F54
: Antoine Equation
B63
: GoalSeek in this column to solve the bubble T equation -- for example, for x1=0 (row 54), set pressure (cell C54) to the pressure at which you want to evaluate the bubble point (P=760 mm Hg is atmospheric pressure) by changing the temperature (cell D54)
C63
: Bubble point temperature
D63
: Ratio of partial pressure of component 1 to the total pressure at the bubble point temperature
COLUMN DESIGN EQUATIONS - TWO COMPONENTS
Parameters for Chosen Set of Components
Component A 9 Benzene Component B 78 Toluene
2.46 1
1 0.406504065
0.5 Benzene 0.5 Toluene
Quality of Feed (q) 0Basis Flow 10 mol/hr
Fenske Equation
0.95 Benzene 0.979269497 Benzene
0.95 Toluene 0.99462004 Toluene
6.5420248922 10.09337884
1.567154753
Underwood Equation
10f 1.7299967254
9.9998936924D 5
15.664310222
10.664310222
2.1328620444
0.95 Benzene 0.979269497 Benzene
0.95 Toluene 0.99462004 Toluene
3.2710124461
Gilliland Correlation
L/D Scaling 1.5 Gilliland Correlation under different Abscissa Ranges:L/D (actual) 3.1992930666 0<x<0.01 -3.716323298854Abscissa 0.2539548932 0.01<x<0.9 0.406433619887186
0.4064336199 0.9<x<1 0.123806185475335
11.70628719
3.2710124461
5.8531435951
aAB aBB
aAA aBA
zA, FEED zB, FEED
Problem A: Given: fractional recovery of A in distillate and B in bottoms
Problem B: Given: mole fraction of A or B in distillate and bottoms product
FRA, DISTILLATE xA, DISTILLATE
FRB, BOTTOM xB, BOTTOM
Nmin Nmin
Nmin, feed
DVFEED
DVFEED, TEST
VMIN
LMIN
(L/D)MIN
Problem A: Given: fractional recovery of A in distillate and B in bottoms - calculate mole fractions
Problem B: Given: mole fraction of A or B in distillate and bottoms product - calculate fractional recoveries
: You need to put a guess value in here to start the iterations - the phi parameter for any combination of components always lies between the relative volatility of these components (ie. aAB>f>aBB)
B34
: Change in vapour flow rate at the feed stage
C34
: Goal seek here - set value of C29 to the DVFEED value in C27 by changing cell C28
C35
: Total flow in distillate stream - distillate flows of each component
C36
: Minimum vapour stream flow rate - from tray 1 of the column
C37
: Minimum liquid condensed in partial condenser to be returned to column (tray 1)
C38
: Minimum reflux ratio
B53
: Ratio between the actual L/D value and the minimum L/D value
C53
: Input any number greater than 1 - columns usually operate at a (L/D)/(L/D)min scaling ratio of 1.05-1.25
F53
: The Gilliland correlation has different regimes described by different equations - all values are calculated here and the spreadsheet (in cell C46) selects the appropriate value based on the abscissa calculated in cell C45
B54
: Reflux ratio of a "real" column
B55
: x-axis of the Gilliland correlation
C56
: This variable is the y-axis value in the Gilliland correlation and is evaluated by one of three functions depending on the abscissa value (see to the right) - the spreadsheet automatically selects the correct value based on the abscissa value given in C45
B57
: Rearrangement of y-axis value to solve for N
C58
: Choose the appropriate value of Nmin,feed depending on whether you were solving Problem A (Nmin,feed=C38) or Problem B (Nmin,feed=G22)
B59
: Scaled feed stage by the value N/Nmin
McCABE-THIELE DIAGRAMS FOR SOLVING DISTILLATION PROBLEMS
Parameters for Chosen Set of Components
Component A 9 Benzene Component B 78 Toluene
0.5 Benzene 0.5 Toluene
0.95 0.05
0.05 0.95
L/D T Type "T" for total refluxL/V 1.00
Quality of Feed (q) 0Basis Flow 10 mol/hr
y vs. x Equilibrium Relationship
Use Bubble T calculation approach:
0.0 0.000 0.0000.1 0.209 0.2080.2 0.376 0.3760.3 0.511 0.5120.4 0.622 0.622 A = -0.48110.5 0.714 0.713 B = 1.63160.6 0.791 0.790 C = -2.45940.7 0.856 0.855 D = 2.30810.8 0.911 0.911 E = 0.00040.9 0.959 0.9591.0 1.000 1.000
Feed and Operating Lines
FEED LINE TOP OPERATING LINE BOTTOM OPERATING LINE
If q is not 1: y difference 0.000 If q is equal to 1: 0.500
0.291 0.71348125
0.50023653175
Appropriate Intercept Value: 0.290748918732101
0.500236531746407
zA, FEED zB, FEED
xA,DIST xB,DIST
xA,BOT xB,BOT
The y vs x values are automatically entered into this table if you've solved the bubble T calculation using the Ideal T - 2 cpts worksheet
xA yA yA
Polynomial Fit: Input these coefficients yourself from graph:
x1
y1
x2
y2
xA yA xA yA
xA yA
xA
xA yA
yA
xA
yA
yA=LV
xA+(1− LV ) xA ,DISTyA=
qq−1
xA+11−q
zA ,FEED
yA=LV
xA+(1− LV ) xA ,BOT
PTOTAL=x1P1¿(T bp )+x2P2¿(T bp)
yA=Ax4+Bx3+Cx2+Dx+E
A3
: Read automatically from the Ideal T - 2 cpts sheet
B11
: External Reflux Ratio
C11
: If you want to do a total reflux calculation, type "T" in this cell - this will set L/V to 1, as must be the case in total reflux
B12
: Internal reflux ratio
C14
: q=1 - saturated liquid q=0 - saturated vapour
C15
: Flow rate to be used as basis for calculations - choose any value
B22
: Read automatically from solved Ideal T - 2 cpts worksheet
C22
: Predictions from polynomial fit of data (4th order) - copy the polynomial fit equation coefficients displayed on the graph on the next worksheet into cells F27-F31 to calculate - these values should be the same as the yA values in column B (from the bubble point calculation)
F27
chen lu: Input from the fit equation in the "McCabe-Thiele Graph" worksheet
A36
: Read automatically from the Ideal T - 2 cpts sheet
D38
: Applies to trays above the feed tray (rectifying section)
G38
: Applies to trays below the feed tray (stripping section)
H39
: The bottom operating line passes through the intercept of the top operating line and the feed line - here, those two equations are simply added to solve for a common value of y and x
H41
: The bottom operating line intercepts with the y=x line when y=x=xB
H43
: Calculated from the two known points (x1,y1) and (x2,y2)
H44
: b=y-mx (at any point on the curve
A49
: When q=0, the feed line is vertical. Since it's impossible to express a vertical line in an equation as a function of x, just insert steps just before and just after the x= zA,FEED (the x-intercept of the vertical line) in this case - the feed line will appear vertical on the graph
A51
: When q=0, the feed line is vertical. Since it's impossible to express a vertical line in an equation as a function of x, just insert steps just before and just after the x= zA,FEED (the x-intercept of the vertical line) in this case - the feed line will appear vertical on the graph
A63
: We need to find this intercept so we can determine when we should be stepping off the top feed line (when the horizontal step lines end at an x value greater than the feed line-VLE equilibrium intersection x value) or the bottom feed line (when the horizontal step lines end at an x value less than the feed line-VLE equilibrium intersection x value)
C65
: Goal Seek to find the intercept of the feed line and the y vs x VLE equilibrium line, set y=mx+b equal to the polynomial describing the VLE curve -- both sides of the equation are equal at the intercept point or, as done here, the difference between the right hand side and the left hand side of the equation is zero; set cell C65 to zero by changing cell C66
F65
: x-coordinate of feed-VLE intercept
C66
: x-coordinate of feed-VLE intercept
F66
: y coordinate of feed-VLE intercept
C67
: y coordinate of feed-VLE intercept
D69
: Correct x and y values are automatically selected from spreadsheet according to the value of q
Stepping off Stages
Keep copying the four-row block at the end (ie. A134-->F137) until you have found the total number of equilibrium trays
: We need to find this intercept so we can determine when we should be stepping off the top feed line (when the horizontal step lines end at an x value greater than the feed line-VLE equilibrium intersection x value) or the bottom feed line (when the horizontal step lines end at an x value less than the feed line-VLE equilibrium intersection x value)
H77
: This table just displays the results in a columnar format for plotting on the graph
C79
: Starting point is the xA,DIST value on the y=x line (top operating line also intersects here)
C81
: Goal Seek - set this cell (C78) to the starting point yA value given in celll C76 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C77
C82
: Vertical step - same x-value as step 1
C83
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C85
: Goal Seek - set this cell (C82) to the step down yA value given in celll C80 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C81
C86
: Vertical step - same x-value as step 2
C87
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C89
: Goal Seek - set this cell (C82) to the step down yA value given in celll C80 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C81
C90
: Vertical step - same x-value as step 2
C91
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C93
: Goal Seek - set this cell (C91) to the step down yA value given in celll C89 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C90
C94
: Vertical step
C95
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C97
: Goal Seek - set this cell (C95) to the step down yA value given in celll C93 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C94
C98
: Vertical step
C99
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C101
: Goal Seek - set this cell (C99) to the step down yA value given in celll C97 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C98
C102
: Vertical step
C103
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C105
: Goal Seek - set this cell (C103) to the step down yA value given in celll C101 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C102
C106
: Vertical step
C107
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C109
: Goal Seek - set this cell (C107) to the step down yA value given in celll C105 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C106
C110
: Vertical step
C111
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C113
: Goal Seek - set this cell (C111) to the step down yA value given in celll C109 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C110
C114
: Vertical step
C115
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C117
: Goal Seek - set this cell (C115) to the step down yA value given in celll C113 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C114
C118
: Vertical step
C119
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C121
: Goal Seek - set this cell (C119) to the step down yA value given in celll C117 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C118
C122
: Vertical step
C123
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C125
: Goal Seek - set this cell (C115) to the step down yA value given in celll C113 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C114
C126
: Vertical step
C127
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C129
: Goal Seek - set this cell (C115) to the step down yA value given in celll C113 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C114
C130
: Vertical step
C131
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C133
: Goal Seek - set this cell (C115) to the step down yA value given in celll C113 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C114
C134
: Vertical step
C135
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
C137
: Goal Seek - set this cell (C115) to the step down yA value given in celll C113 (we are taking a horizontal step, so the y value is the same as y value of starting point) by changing cell C114
C138
: Vertical step
C139
: If the x value is still above the feed line-VLE equilibrium intercept, take a vertical step down to the top operating line; if it is below the feed line-VLE equilibrium feed step, take a vertical step down to the top of the bottom operating line
Relative Volatility aAA Relative Volatility aBA Relative Volatility aCB
P¿=10(A− B
T−C )P¿=10
(A− BT−C )
PTOTAL=x1P1¿(T bp )+x2P2¿(T bp)+x3P3¿(T bp )
∑ x i=y1PTOTAL
P1¿(T dp )+y2PTOTAL
P2¿(T dp )+y3PTOTAL
P3¿(T dp)=1
P¿=10(A− B
T−C )P¿=10
(A− BT−C )
P¿=10(A− B
T−C )P¿=10
(A− BT−C )
α AB=
yA
xA
yBxB
≈
P¿A
P
P¿B
P
=PA
¿
PB¿
D5
: Highest vapour pressure of all three components
H5
: Vapour pressure 2nd highest of 3 components
L5
: Vapour pressure is either lower than both components A and B (heavy non-key) or higher than both components A and B (light non-key) - if the vapour pressure is intermediate to these values, a different solution is required for the Underwood equation requiring solving systems of equations, outside the scope of the spreadsheet
C13
: Mole fraction of component A in the feed to the column
G13
: Mole fraction of component B in the feed
K13
: Mole fraction of component A in the feed to the column
D15
: Input any temperature here - compare P* values for the two components to determine which is A and which is B
L15
: Input any temperature here - compare P* values for the two components to determine which is A and which is B
B16
: Antoine Equation
F16
: Antoine Equation
J16
: Antoine Equation
C23
: Set as a constant over the full column
C25
: x1=z1 as the first bubble is produced
B27
: Bubble Point - the temperature at which the first bubble of vapour forms (or, the bottom line on a two component phase diagram at a given composition)
C28
: GoalSeek in this cell to solve the bubble T equation -- set this cell (C21) to the pressure at which you want to evaluate the bubble point (P=760 mm Hg is atmospheric pressure) by changing the temperature (cell C20)
B29
: Calculate the vapour pressure of component 1 at the calculated bubble point pressure
B30
: Calculate the vapour pressure of component 2 at the calculated bubble point pressure
F30
: Under VLE conditions, the sum of the partial pressures should equal the total pressure of the system (cell C21) - this is a check that your bubble T calculation is correct
B31
: Calculate the vapour pressure of component 2 at the calculated bubble point pressure
C32
: Mole fraction of component 1 in the first bubble of vapour produced - ratio of the vapour pressure of component 1 and the total pressure
C33
: Mole fraction of component 2 in the first bubble of vapour produced - ratio of the vapour pressure of component 2 and the total pressure
B43
: Dew Point - the temperature at which the first drop of liquid forms (or, the top line on a two component phase diagram at a given composition)
C44
: GoalSeek in this cell to solve the dew T equation -- set this cell (C39) to 1 (the sum of the total liquid mole fractions) by changing the dew point temperature (cell C38)
C45
: Mole fraction of component 1 in the first drop of liquid produced - calculated by Raoult's Law (x=yP/P*)
C46
: Mole fraction of component 2 in the first drop of liquid produced - calculated by Raoult's Law (x=yP/P*)
C47
: Mole fraction of component 2 in the first drop of liquid produced - calculated by Raoult's Law (x=yP/P*)
C52
: Average of dew and bubble point - good estimate of average column temperature
B57
: Antoine Equation
F57
: Antoine Equation
J57
: Antoine Equation
B59
: Relative volatility with respect to the heavy key
C59
: Relative volatility is estimated as ratio of partial pressures at "average" VLE temperature (since (yA/xA)/(yB/xB) = (P*A/P*B) - this value is assumed to be approximately constant over the operating temperature range of the column, which is never correct but usually close enough for approximate design considerations. Try changing the temperature to the dew and bubble points (the extreme operating temperatures of the column) and see how the ratios change.
B60
: Relative volatility with respect to the light key
COLUMN DESIGN EQUATIONS - MULTIPLE COMPONENTS
Parameters for Chosen Set of Components
LIGHT KEY COMPONENT HEAVY KEY COMPONENTComponent A 78 Toluene Component B
3.3809878333
1
0.3 Toluene
Quality of Feed (q) 0Basis Flow 100 mol/hr
Fenske Equation
0.95 Toluene
0.95 Cumene
4.9184225125
0.9992688854
Underwood Equation
100f 0.6010869816
99.999999282D 69.970755417
123.80952017
53.838764748
0.7694466699
0.4073130243 Toluene
0.9490748234 Cumene
0.5712494481 Benzene
2.4171043403
Gilliland Correlation
L/D Scaling 1.25 Gilliland Correlation under different Abscissa Ranges:L/D (actual) 0.9618083374Abscissa 0.0980532419
aAB aBB
aAA aBA
zA, FEED zB, FEED
In a multicomponent system, fractional recoveries are usually specified:
: This is the estimate calculated using ideal VLE assumptions - if the problem gives you a better estimate, input it here in place of the VLE prediction
C13
: q=1 - saturated liquid q=0 - saturated vapour
B31
: Change in vapour flow rate at the feed stage
B32
: Lumped parameter: L(min)/[(V(min)*K]
C32
: You need to put a guess value in here to start the iterations - the phi parameter for any combination of components always lies between the relative volatility of the 2 key components (ie. aAA>f>aBA)
B33
: Change in vapour flow rate at the feed stage
C33
: Goal seek here - set value of C33 to the DVFEED value in C31 by changing cell C32
C34
: Total flow in distillate stream - distillate flows of each component
C35
: Minimum vapour stream flow rate - from tray 1 of the column
C36
: Minimum liquid condensed in partial condenser to be returned to column (tray 1)
C37
: Minimum reflux ratio
B47
: Ratio between the actual L/D value and the minimum L/D value
C47
: Input any number greater than 1 - columns usually operate at a (L/D)/(L/D)min scaling ratio of 1.05-1.25
B48
: Reflux ratio of a "real" column
B49
: x-axis of the Gilliland correlation
0.4878361556
10.555721039
2.4171043403
5.1874923461
[N-Nmin]/(N+1)
[N-Nmin]/(N+1)
Nactual
Nmin, feed
Nfeed, actual
N F=N F ,min( NNmin )
Abscissa=
LD
−(LD)min
LD
+1
C50
: This variable is the y-axis value in the Gilliland correlation and is evaluated by one of three functions depending on the abscissa value (see to the right) - the spreadsheet automatically selects the correct value based on the abscissa value given in C45
B51
: Rearrangement of y-axis value to solve for N
C52
: Cell C42
B53
: Scaled feed stage by the value N/Nmin
Parameters for Chosen Set of Components
HEAVY KEY COMPONENT NON-KEY COMPONENT21 Cumene Component C 9 Benzene
1 7.89830348
0.295771546 2.33609343
0.3 Cumene 0.4 Benzene
Fenske Equation
Underwood Equation
Gilliland Correlation
Gilliland Correlation under different Abscissa Ranges:0<x<0.01 -0.82099578205674
: Must be either a light non-key or a heavy non-key - a sandwich component must be treated using a series of equations approach outside the scope of this spreadsheet
G9
: This is the estimate calculated using ideal VLE assumptions - if the problem gives you a better estimate, input it here in place of the VLE prediction
K9
: This is the estimate calculated using ideal VLE assumptions - if the problem gives you a better estimate, input it here in place of the VLE prediction
F47
: The Gilliland correlation has different regimes described by different equations - all values are calculated here and the spreadsheet (in cell C46) selects the appropriate value based on the abscissa calculated in cell C45