Cyclic Separation: New trays and Experimental Results

EFCE WP Fluid SeparationsTechnical Meeting #58, May 12, 2016

Copenhagen, Denmark

Abildskov J., Bay Jørgensen S.CAPEC‐PROCESS

Department of Chemical and Biochemical EngineeringTechnical University of Denmark

Outline

Cycled Trays Operation‐ Theoretical Performance: Lewis Case II analogy

‐ Early realizations; Simultaneous Draining

‐ Simultaneous Draining; Problems

‐ Ideas to approach theoretical performance

New Principles of Design and Operation

New Trays and Sequential Tray Draining

New Realization

Results

Future Directions

Conclusions

Other Explorations

• AIChE Spring Meeting 2016

• C. Knoesche, J. Paschold, S. Buetehorn(BASF): 

• Paper 87b: “Cyclic Operation of Distillation Columns”

• Maleta CD

Theoretical Performance:

Lewis Case II Analogy

Theoretical Performance:

Lewis Case II Analogy

• Performance:

• (Convenience of) realization:

CASE ”I” ”II” ”III”

εtray/εpoint 1.7 2 1.5

”II” > ”I” > ”III”

”III” > ”I” > ”II”

“ … (Lewis’ case III) …. is probably the most commonin conventional operation because of designers’ understandablereluctance to install downcomers in the configuration of case II”

(Sommerfeld et al., 1966)

εpoint → 1

Theoretical Performance:Lewis Case II Analogy; 

Conventional Operation with Downcomers

yn-1(z)

yn (z)

z=0

xn(0) = xn+1(z0)

xn(z0)

z=z0

z: horizontal tray positionz0: length of liquid flow

x n(z)

z

xn(0) = xn+1(z0)

xn(z0) = xn‐1(0)

xn(0) = xn+1(z0)

xn(z0) = xn‐1(0)

z00

→ Liquid in down‐comer from n+1 enters plate n at z = 0: xn(0) = xn+1(z0)

→ No liquid back‐mixing; progressive depletion of volatile component by contact with vapor from plate n – 1

→ z = z0: xn(z0) is discharged to n – 1

Theoretical Performance:Lewis Case II Analogy; 

Cyclic Operation without downcomers

x n(t)

t

xnL = xn+1

V

xnV = xn‐1

L

τ0

yn-1(t)

yn (t)

t=0

xnL =

xn+1V

xnV =

xn-1L

t=τ

Independent variable: tTime length: VFP = τ

→ No liquid phaseenrichment on plate nby liquid flowing down→ No liquid back‐mixing in time domain→ Depletion of volatile component occurs monotonically with time→ Perfectly mixed plate→ Amount of material dropped during LFP is exactly equal to one plate holdup

Theo

reticalPerform

ance:

Lewis Case II analogy

‐Sommerfeld(1966)

z: horizontaltray position

t: time of vaporflow period ofduration τ

x n(z)

z

xn(0) = xn+1(z0)

xn(z0) = xn‐1(0)

xn(0) = xn+1(z0)

xn(z0) = xn‐1(0)

x n(t)

t

xnL = xn+1

V

xnV = xn‐1

L

zn

τ

0

0

Early realizations; Simultaneous Draining• Three 1961 papers (Cannon)

• Exactly one plate holdup must be dropped during LFP

• Basis of the generation: Simultaneous Draining

• TP = VFP + LFP

• Modeling:  McWhirter (1963)

Sommerfeld et al. (1966)

Chien et al. (1966)

Robinson/Engel (1967) Vapor Flow Period (VFP):

VF, VR, VT and VB are closed. VV is open.

Liquid Flow Period (LFP):

VT, VB, VF, VR are opened. VV is closed.

Simultaneous Draining; Problems

• Back‐Mixing

• Non‐uniform draining of traysSchrodt V. N. et al., 1967. Chem. Eng. Sci., 22, 759

Gerster J.A., Scull H.M., 1970. AIChE J. , 16, 108.

Larsen J., Kümmel M., 1979. Chem. Eng. Sci., 34, 455.

• Pressure Dynamics

• Liquid oscillations on trays

Approaches to Theoretical Performance‐ Improved Liquid Flow Control

• Matsubara• Collect liquid in a store, prior to transfer to the tray below. 

• Vapor transferred in tubes from tray to tray

• Furzer (TD) Time delay trays• Gutter channels to delay the liquid draining

• Baffles to avoid oscillations

• Hentrich/Vogelpohl (1980)• 1st approach: Liquid was led into a storage tube; when all trays were empty, the liquid was released into the next tray. This was accomplished using a central rod in the column (EXPENSIVE). 

• 2nd approach: Sieve tray with packing material below, to make the liquid drain more slowly during the LFP

• Baron: Stepwise Periodic Operations

• Maleta CD (21st century): Sluice chambers

New Principles of Design/Operation

Tray Design Principle: The periodic cycled separation tray should be constructed such that:

1) All the liquid leaves the tray and enters the tray below without being mixed with any other liquid during the liquid drainage

2) Vapor flow can continue during liquid drainage

Operational Principle for Sequential Tray Draining:To minimize mixing with liquid from other trays enforce a special 

operational principle for Sequential Tray Draining: 

The trays are emptied sequentially rather than simultaneously, while the vapor flow is maintained 

throughout

New Realization; Trays

Cyclically Operated Perforated Sheets (COPS)

Layer of structured packing for liquid oscillation damping 

Sulzer MellapakPacking

Pressure Control Valves

Sheets

Closed Open (for drainage)

When tray is closed: Free area due to perforation is 6%When tray is open: Free area near 100%

New Realization; Sequential Tray Draining

Sequential tray drainage operation for a column with three trays. When the top tray isempty (d) a load of liquid is fed to the top tray (e). The column remains in state (e) forthe remainder (if any) of the cycling period time until a new liquid drainage sequenceis initiated at the start of the next period (a) – (c).

PH3

PH2

PH1

P3

T6P4

On/off valveColumn

ProductTank

FeedTank

PumpsSide streamoutlet

Outlet

MeasurementsT1,T3: dry thermomentertemperatureT2, T4: wet thermometertemperatureT5: humidifier temperatureT6: column temperatureP1, P2: pressure difference across orificeP3, P4: pressure difference across top trayPH1, PH2, PH3: places for pH measurements

Outlet

Throttle

Humidifier

Hea

t Exchanger

PumpCentrifugal Blower

Orifice

P1 P2

T2T1

T4T3

drain

Ventilation

Steaminlet

Wateroutlet

Manual valve

Automatic valve

T5

WaterFeed

ProductReservoir

Control: G <‐ (P1,P2) <‐ Frequency ConverterT6  <‐ T5 <‐ T4  (Cascade)

Results; Point Efficiency, εt

• Strip NH3 from water (Gerster/Scull, 1970)

• Tray molar balance + Murphree expression:→ Slope of pH versus t gives εt

• Assume: Only NH3 transfer

• CO2 Controversy

y = 0.0295x + 0.7588

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

00 20 40 60 80 100

ε t

H/G (seconds)

Efficiency vs. Hold up/Vapor Flow

Efficiency vs. Hold up/VaporFlow

Linear (Efficiency vs. Holdup/Vapor Flow)

in out

t t

t

dxH = G(y ‐ y )dt

dx G dx G= ‐ε m x + ε m x = 0

dt H dt H

Hε = logc(t = 0) ‐ logc(t) ln10

Gmt

2H= pH(t = 0) ‐pH(t) ln10Gmt

G,yin(t)

G,yout (t)

H, x(t)

Results; Column Efficiency, 

Experimental Procedure

id idp c

N NCE = , CE =

3 3

f p

p f

x (1 ‐ x )S =

x (1 ‐ x )

f

pid

1 x 1ln ︵1- ︶ +λ x λ

N =lnλ

mG, λ =L

G

L

L ρFP =

G ρ

Measure(xf, xp) by 

titration

ObtainStationaryPeriodicState

Set (L,G)

16656.695‐

Tm = 0.073×10

(Gerster,1958)

Results; Separation Factors

Stream Limits

G (mol/s) 5‐10

L (mol/s) 0‐20

0

5

10

15

20

25

30

35

40

1.E‐03 1.E‐02 1.E‐01 1.E+00

Separation Factor, S

FP

Sieve Trays Periodic Trays Packings (Sulzer BX gauze)

Results; Column Efficiencies

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.E‐03 1.E‐02 1.E‐01 1.E+00

Column Efficiency

FP

Sieve Trays Periodic Trays

Results; Influence on Tray Efficiency of Flows

y = 0.0055x + 0.7921

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

0 10 20 30 40 50 60 70

Column  Efficiency, C

E p

H/G (seconds)

CEp vs. H/G

Column Efficiency vs. H/G

Linear (Column Efficiency vs. H/G)

No points at H/G > 70 …….!

Results; Influence on Column Efficiency of Period Lengths

• Maximum?

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0 50 100 150 200 250

CE p

= Nid / 3

TP (seconds)

CEp vs. TP

TP: Total Period Duration

f

pid

id

1 x 1ln ︵1- ︶ +λ x λ

N =ln ︵λ ︶

λ= mG / LExperimental data make N go thru extremum

6.9 < G(mol/s) < 8.42 < L (mol/s) < 8.8

Why optimum TP?

0.E+00

1.E‐04

2.E‐04

3.E‐04

4.E‐04

5.E‐04

6.E‐04

0.E+00 5.E‐04 1.E‐03

y

x

Henry's law

1.1666

Equilibrium Curve:y* = x∙mOperating Line:y = (x‐xp)∙(L/G)

TP↑ ~ L↓~ (L/G)↓ ~ Nid ↓

BUT if

xp ↓ Nid ↑

This may compensate, if xpgets sufficiently low. Thisseems to be the case at low TP values, but only up to 80 seconds.

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

0 50 100 150 200 250

CE p

= Nid / 3

TP (seconds)

TP/CEp

Blue: G = 7dP = 50

Red: G = 8.3dP = 60

Further Work

• Operation at H/G beyond 70 seconds?

Latest Results

0.5

0.7

0.9

1.1

1.3

1.5

1.7

1.9

2.1

2.3

2.5

00 20 40 60 80 100

CE p

= Nid/3

H/G

Spring 2016

Previous

Linear (Spring 2016)

Conclusions

• Presented:

– New Tray Design (Principle): COPS

– New Operational Principle: Sequential Draining

– Steady Operation: Avoiding Pressure Dynamics and Liquid (Back) Mixing

• Future Directions:

– Operation at higher (periodic) tray efficiencies