PB84-232628
Process Design Manual for Stripping of Organics
Water General Corporation
Waltham, MA
Aug 84
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
-------
-------
EPA-600/2-34-139
Auoust 1984
PROCESS DESIGN MANUAL FOR STRIPPING OF ORGANICS
-by
Harish M. Shukla and R. Edwin Hicks
Water General .Corporation
Waltham, MA 02154
Contract No. 68-03-3002
Project Officer
Kenneth A. Dostal
Organic & Inorganic Chemicals & Products Branch
Industrial Pollution Control Division
Industrial Environmental Research Laboratory
Cincinnati, OH 45268
INDUSTRIAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OH 45268
-------
TECHNICAL REPORT DATA
(Pleae nad Instructions on the.reverse before completing)
1. REPORT NO. 12. . 3. RECIPIENT'S ACCESSION NO.
FpA-f\nn/2-84-139 PAA 4 232628
4. TITLE AND SUBTITLE 5. REPORT DATE
. 1QA4
Process Design Manual for Stripping of Organics 6. PERFORMING ORGANIZATION CODE
.
7. AUTHOR(S) . B. PERFORMING DRGANIZATION REPORT NO.
Harish M. Shukla & R. Edwin Hicks
9. PERFORMING ORGANIZATION NAME AND ADDRESS 10. PROGRAM ELEMENT NO.
Water General Corporation
Waltham; MA 02154 11. CONTRACT/GRANT NO.
12. SPONSORING AGENCY NAME AND ADDRESS 13. TYPE OF.REPORT AND PERIOD COVERED
Industrial Environmental Research Laboratorj- ,
Office of Research and Development 14. SPONSORING AGENCY CODE
united States Environmental Protection Aqency EPA/600-l2 .
Cincinnati, Ohio 45268
15. SUPPLEMENTARY NOTES
16. ABSTRACT
Procedures and correlations for designing and costing stripping towers for the removal
of organics from aqueous streams are presented. The emphasis is on practical methods
suitable for engineering estimates. . The designs cover steam strippers with and with-
out condensers and reflux, as well as air stripping. 'Steam stripping is treated as
an isothermal process and simplified equations for the determination of tower height
may be used. Determination of the height of adiabatic'ai~ strippers involves a tedi-
.ous, iterative solution of heat and. material balances. A BASIC computer program for
carrying out these calculations is provided.
"
Capital costs are determined, essentially, by estimating the quantity of materials
required in conjunction with ~ateria~ costs. Cost factors for a rang~ of materials
and installation factors are suggested., Methods for costing ancillary equipment such
as heat exchangers, pumps, compressors, and storage vessels are included. It is rec-
commended that vendor quotes be obtained wherever possible. Typical operating costs
for energy and maintenance are also given.
17. KEV WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS b.IDENTIFIERS/OPEN ENDED TERMS c. COSATI Field/Group
'..
.
18. DISTRIBUTION STATEMENT ,19. SECURITY CL.ASS (Tllis Report) 2"" NO. OF PAGES
. Unclassified - ...
148
Release Unlimited 20. SECURITY CLASS (Tllis page) 22. PRICE
Unclassified . "
EPA F...", 2220-1 (Rn.4-77)
PREVIOUS ECITION 18 OBSOLETE.
. 1
-------
NOTICE
This document. has. bee'n reviewed in accordance wi th U. s.
Environmental Protection Agency policy and approved for
publication. Mention of trade names or commercial products
does not constitute endorsement of recommendation fo~ use.
ii
-------
FOREWORD
When energy and material resources are extracted, pro-
cessed, converted, and used, the related pollutional impacts
on our environment and even on our health often require that
new and increasingly more efficient pollution control methods
be used. The Industrial Environmental Research Laboratory-
Cincinnati (IERL-Ci) assists in developing and demonstrating
new and improved methodologies that will meet these needs both
efficiently and economically. . .
Stripping is one of several processes that are used com-
mercially for separating organics from water and wastewater.
The selection of the optimum process for any given application
is dependent on the characteristics of the organics as well as
on other factors including cost and the need. to recover the
separated organics.
In the case of a mixture of organics having different
properties, two or more treatments in serie~ may have to be
provided. The use of two different processes, for example one
with cost dependent on throughput, followed by a polishing
step having cost dependent on feed concentration, can result
in a lower overall treatment cost. In a wastewater treatment
train, $tripping is typically the first process that separates
dissolved substances. It follows clarification or filtration
. .
steps that are used for removal of suspended solids, and may
precede polishing steps such as carbon or resin adsorption.
Requests for further information regarding Stripping of
Organics should be directed to the Organic and Inorganic
Chemicals and Products Branch of the Industrial Pollution
Control Division, IERL, Cincinnati. '
Industrial
David .G. Stephan
Director
Environmental Research
Cincinnati
'Laboratory
iit
-------
ABSTRACT
Procedures and correlations for designing and costing
stripping towers for the removal of organics from aqueous
streams are presented. The emphasis is on practical methods
suitable for engineering estimates. The designs cover steam
strippers with and without condensers~and reflux, as well as
air stripping. Steam stripping is treated as an isothermal
process and simplified equations for the determination of tower
height may be used. Determination of the height of adiabatic
air strippers invovles a tedious, iterative solution of heat
and material balances. A BASIC computer program for carrying
out these calculations is provided.
Capital costs are determined; essentia!ly, by estimating
the quantity of materials required in conjunction with material
costs. Cost factors for a range of materials and installation
factors are suggested. Methods for costing ancillary equipment
such as heat exchangers, pumps, compressors and storage vessels
are included. It is recommended that vendor quotes be obtained
wherever possible. Typical operating costs for energy and
maintenance are also given.
This report was submitted in fulfillment
68-03-3002 by Water General Corporation under
of the u.S. Environmental Protection A~ency.
covers a period from March" 1983 to June.1984,
completed as of June 1984.
of Contract No.
the sponsorship
This report
and work was
iv
-------
CONTENTS
Foreword. . . . . . . . . . . . . . . . . . . . . . . . . i i i
Abstract. '. . . . . . . . . . . . . . . . . . . . . .. iv
F i gu re s .... . . . . . . . . . . . . . . . . . . . . . vii
Tables. . . . . . . . . . . . . . . . ., . . . . . vii i
1.
Introduction. . . . . . . . .
1.1 Stripping Applications. .
1.2 purpose. . . . . . . .
1.3 Scope
. . . . . . . . . . . . 1-1
. . . . . . . . . . . . 1-1
. . . . . . . 1-3
. . . . . . . . . . 1-4
2.
Background to Design . -. . . . . . . 2-1
2.1 Tray and Packed Towers. . . . . . . . . . . . . . 2-5
2.1.1 Selection Between Tray and Packed Towers 2-11
. 2'.2 Design Approach. . /.. '0 " . . . . . . . .. 2-14
2.2.1 Equilibrium Stage.' . . . .. 2-16
2.3 Vapor-Liquid Equilibrium. . . . . . . . . .. 2-17
2.3.1 Gas Phase. . . . . . . . . . . . . . .. 2-17
2.3.2 Liquid Phase. . . . . . . . . . . . . . 2-18
2.3.3 Henry's Law. . . . .. . . . 2-19
2.4 Stage Efficiency. . . . . . . . . . . . . . 2-22
2.4.1 Types of Stage Efficiencies. . . . . . . 2-23
2.4.2 Tray Towers. . . . . . . . . 2-23
2.4.3 Packed Towers. . . . . 2-26
Development of Des ign Procedures. . . ., . . . .. 3-1
3.1 Tray Tower Design. . . . . . . . . '. . . 3-1
3.1.1 Theoretical Number of Stages. . . . . . 3-3
Simplified Equations for Steam Stripping 3-10
Adiabatic Stripping with Air 3-10
3.1.2 Tower Diameter and Height. . . . . . . . 3-13
3.1.3 Stage Efficiency 3-17
3.1.4 Pressure Drop. . . . . . . . . . . . .. 3-18
3.2 Packed Tower Design. . . . . . . . . . . . .. 3-19
3.2.1 Selection of Packing. . . . 3-19
3.2.2 Tower Diameter. . . . . . . . . 3-20
3.2.3 Height of Packing. . . . . ... . . . .. 3-23.
3.2.4 Pressure Drop. . . . .:. . . . 3-27
3.3 Ancillary Equipment. . . . . . . . 3-29
3.3.1 Heat Exchangers. . . . . . . . . . . .. 3-29
3.3.2 Storage Vessels. . . . . . . . . . ... 3-30
3.
4.
An a 1 y s is. . . . . . e.. . . . . . . . . . . . . . . 4 - 1
Capital Costs. . . . . . . . . . . . . . . . . . 4-2
4.1.1 Mass Transfer Equipment. . . . . . . . . . 4-4
Shell Weight. . . . . . . . . . . . . . . 4-5
Shell Cost. . . . . . . . . . 4-8
Platforms and Ladders. . . . . . . . .. 4-10
Cost of Interna1s. . . . . . . . . 4-10
Total Purchases Cost. . . . . . . 4-12
Cost
4.1
v
-------
4.1.2 Heat Transfer Equipment. . . . . . . . . 4-13
4.1.3 Centrifugal Pumps. . . . . . . . . 4-15
4.1.4 Electric Motors. . . . . . . 4-17
4.1.5 Air Compressors. . . . . . . 4-19
4.1.6 Storage Vessels. . . . . . . . . . . . . 4-19
4.1.7 Installation Costs.. . . . . 4-21
Operating Costs. . . . . . . . . . . . . 4~22
4.2.1 Fixed Costs. . . . . . . . . . . . . . . 4-23
4.2.2 Variable Costs. . . . . . . . . . . . . 4-23
4.2
5.
Summary of Design Procedures . . . . . . . . .
5.1 Process Desig~. . . . . . . . . . . . . . . . .
5.2 Cost Estimation . . . .
6.
References
. . .
.......
...........
Appendices
A.
B.
C.
D.
E.
Estimation of Tray Efficiency. . . . . . . . . . . .
Design Example. . . . . . . . . . . . . . . . . . .
Adiabatic Air Stripper: Basic Program . . . .
Antoine Coefficients: Basic Program. . . . . . . .
Toxic organic List: Ease of Stripping,
Henry's Law Constant, and Activity Coefficients. . .
S ymbo 1 s . . . '. . . . . . . . . . . . . . . . . . .
F.
vi
5-1
5-1
5-5
6-1
A-l
B-1
C-1
0-1
E-1
F-l
-------
Number
2-1
2-2
2-3
2-4
2-5
3-1
3-2
3-3
3-4
3-5.
3-6
A-1
FIGURES
Page
Simplified Diagram of a Stripping Tower. . . . . . 2-2
Modes of Flow in Stripping Equipment. . . . 2-4
Pictorial of Packed and Plate Towers. . ~ . . . . 2-6
Methods for Dispersing Vapor into Liquid on
a Tr a y . . . . . . . . . . . . . . . . . . . . . . 2 - 7
Packing Materials. . . . . . . . . . . . .. ~-10
Process Flow Diagram for Steam, Stripping . . . . . 3-2
Stripping Tower Nomenclature. . . . . . . . . . . 3-4
Effect of Temperature on Air Stripper Design. . 3-14
Flooding and Pressure Drop in Packed Towers. .. 3-21
Packing Parameters for Determination of HTU. .. 3-26
Dependency of NTU on Stripping Factor and
Removal Efficiency. . . . . . . . . . . . 3-28
Entrainment CO,rrelation. . . . . . . . A-4
vii
-------
Number
1-1
2-1
2-2
3-1
3-2
3-3
3-4
3-5
,3-6
, 4-1,
,4-2 '
4-3
4-4
4-5
4;"'6
4-7 '
4-'8
,4-9
4-10
4-11
A-1
TABLES
Page
Common Processes for the Separation of
Organics from Water. . . . . . . . . .
Relative Performance Ratings of Trays
and ];)ackings . . . . . . . . .~. . . '. .
Selection Guide for Tower Internals. . .
Summary of Stream Rates and Compositions.
Design Equations for Tray Towerf? . . . . .
Recommended Values for Tray spacing' and
Flooding Constants for Tray Towers. . . . . . . . 3-16
Characteristics of Commercial Packings . . . . . . 3-22
Determination of HTU for Packed Towers. . . . . . 3-25
Typical Mean Values of the Overall Heat-
Transfer Coefficieant. . . . . . . . . . . . 3-31
AACE Classification of Capital Cost Estimates. . . ' 4-1
Correlations for Cost of Stripping Towers. . . . . 4-9
Correlations for Cost of Platforms and Ladders. . 4-11
Correlations for Cost of Tower Trays'. . . . . . . 4-11
Cost of Tower Packing per Unit Volume. . . . 4-12
Correlatlons for Costs of Heat Transfer'
1-2
. . . . . 2-13
. . 2-13
. . . . 3- 8'
. 3-11
. 4-14
4-16
Equ i pme n t . . . . . . . . . . . . . . . . ~
,Correlations for Cost of Centrifugal Pumps.
Capacity, Head, and Hor~epower Limits for
Centrifugal pumps. . . . . . . . . . . . .
Correlations for Cost of Electric Motors. .
Correlations for Costs of Storage Vessels. . .
Fixed Operating and Maintenance Cost Basis
and Unit'Cost Factors for Stripping. . . . .
Summary of AIChE Procedure for Prediction ,of
Tray Efficiency. . . . . . . . . . . . .J . . . .
. 4-17
4-18
. . 4-20
4-24
,A-1
viii
-------
SECTION I
INTRODUCTION
Stripping is one of several processes that. are used
commercially for separating organics from water and wastewater.
The selection of the optimum process for any given application is
dependent on the characteristics of the organics as well as on
other factors including cost and the need to recover the.
separa.ted organics. Excellent descriptions of all the important
separation processes can be found in the U.S. Environmental
Protection Agency's (EPA's) review "Control of. Organic Substances
in Water and Wastewater".l Commonly used separation processes are
listed together with some selection criteria in Table 1-1.
In the case of a mixture of organics having different
properties, two or more treatments in series may have to be
prov~ded. The use of two different processes, for example one
with cost dependent on throughput, followed ~y a polishing step
having cost dependent on feed concentratio~, can result in a
lower overall treatment cost.2 In a wastewater treatment train,
stripping is typically the firs~ process that separates dissolved
substances. It follows clarifica~ion or filtration steps that are
used for removal of suspended solids, and may preceed polishing
steps such as carbon or resin adsorption.
1.1
STRIPPING APPLICATIONS
The
essential
characteristic
that
determines
the
effectiveness of stripping in separating "dissolved organics is
the relative volatility or vapor pressure of the organic above
. 1:"1
-------
TABLE 1-1
~
COMMON PROCESSES FOR THE SEPARATION OF ORGANICS FROM WATER
Process
Biochem.ica~.
Aerobic,
Anaerobic
Solvent
ertraction
Membrane
Ultra-
filtration,
Revers e
os mos is
Adsorption
Carbon,
Res in
Stripping
Required
Characteristic
of Organic
Biodegradable
More soluble
in solvent
than in water
High molecular
weight,
. ( ionizable)
Ads orbs on
selected
ads orben t
Volatile
Recovery
of Organics
No, organics
are destroyed
Yes, by
fractionating
the solvent
Concentrated
aqueous stream
recovered
Not with carbon,
possible with
res ins
Not us ually
with air:
possible with
steam stripping
.. . .
.In add~t~on to removal efficiency.
1-2
Primary
Cost
*
Dependence
Concentration
of organics
Water
throughput
Water
throughput
Concentration
of organics
Water
t11;roughput
-------
the aqueous phase. It has been shown ~hat at least half of the
186 organics on the EPA's toxic pollutant list are sufficiently
volatile to be effectively removed from aqueous waste streams by
stripping.3 Sixty eight of these can be very easily stripped by
air at ambient temperatures. Others can be stripped at
eleva~ed temperatures with steam. The remaining substances on the
list have relatively low vapor pressures and are not easily
stripped. The classification of - the compounds on the toxic
organics list by ease of stripping that is given in Refere.nce 3
is reproduced in the ,Appendix.
Stripping is emerging as a cos t effecti ve al ternati ve for
treating a wide range of aqueous streams containing organics. It
may be used both as an "in-plant" process for the recovery of
organics from relatively concentrated aqueous 'streams, and as an
~end-of-pipe" treatment for removal of dilute and even trace
quantities of organics from wastewaters prior to discharge or
recycle. Steam stripping is typically ,used for in-plant
separation, whereas air or steam may be used for end-of-pipe
. ,
treatment, depending on the volatility of the organics and post-
treatments provided. In'addition, air stripping is being
increasingly used for the removal of trihalomethanes (THM's) and
trichloroethylene (TCE) from drinking water 'supplies. References
4-9 contain useful data on pilot and commercial scale strippers.-
, ,
1.2 PURPOSE
The purpose of this manual. is to provide, within a single
document, both data and procedures for designing and costing
stripping systems for organics separation. A ma~or objective was
to develop and summarize simplified and practical engineering
procedures of study grade accuracy. The des igns and cos ts
obtained are suitable for eva 1 uating the feas ibi 1 i ty and
viability of stripping relative to other control technologies,
and for checking commercial designs. They"are not intended for
detailed or definitive designs.
1-3
-------
1.3
SCOPE
The design procedures cover:
1. Tray and packed towers
2. Air and steam stripping
3. Live and reboil steam
4. Refluxed and non-refluxed steam stripping
5. Isothermal and adiabatic operation
6. Continuous operation
7. Ancillar~ equipment including heat exchangers
The design and cost procedures are summarized in a stepwise
fashion to facilitate their routine use. Procedures for using.
. .
the simpli£ied analytical equations appropriate to most stripping
applications are demonstrat~d by means of a worked example. A
BASIC program suitable for desk-top computers is provided for the
case of adiabatic air stripping where the usual simplified
equations are less reliable. A comprehensive review of the theory
of stripping and t?e development of the design equations is also
included. Although this manual is not intended as a text, the
reader may find the background material useful as a refresher
course in s tripping. We recommend that all us ers s can the
background sections, particularly with reference to the
limitations on the procedures and data.
The process design is oriented towards sin~le component
relatively dilute systems. Other systems can nevertheless be
. .
handled as well. For example, multicomponent systems- can be sized
by designing for the least volatile organic, and then determining
the dis tribution of the other components separate lYe Us e of the
simplified .desigri correlations for concentrated ~treams may
result in errors due to thermal effe~ts and deviations from
vapor-liquid equilibrium correlations. The more rigorous design
equations may, however, be used without difficulty provided the
necessary enthalpy and equilibrium constant data...are available.
1-4
-------
The.design of ancillary systems such as decanters and other
vapor handling equipment is specifically not handled. Multiple
towers and batch stripping are also not treated~ Design
procedures for these applications as well as more detailed
methods for handling concentrated, multicomponent systems may be
found- in References 10-12. Maintenance and operational problems
are not discussed, nor are controls, instrumentation.and civil
des ign covered. The cos t of thes e i terns is neverthe 1 ess inc 1 uded
in the budget es timates .
. ,
, 1-5
-------
SECTION 2
BACKGROUND TO DESIGN
Stripping operations are characterized by the transfer of
mass from one phase to another. In our case, we are specifically
interested in the transfer of an organic solute from the water in
which it is dissolved to the gas phase.
Stripping occurs in nature'. Many poll.utants have a half life
in a.river or iake which is controlled by the' rate at which they
vaporize into the atmosphere. Bubbling air through the water will
. greatly enhance the rate of pollutant vaporization and can be
used to clean the water. The air that bubbles through the water
is called the stripping medium. Steam is used as .the stripping
medium in industrial applications in cases where improved removal
is obtained at elevated temperatures.
In industry, stripping is carried out in. a stripping tower
similar to those shown in Figure 2-1. The stripping medium is
genera lly introduced at the bas e of the toweJ:' whi 1 e the 1 iquid
stream, called the feed, is introduced at or near the top. When
steam is used it may be introduced directly into the base of the
tower as live steam or may be introduced indi~ectly as reboil
steam through a heat exchanger or reboiler. The steam leaving the
top of the tower may be condensed and some of the condensate
returned to the tower as reflux. The tower may be fitted with
trays or filled with packing for enhancing contact between the
two phas es .
2-1
-------
CONDENSER
OVERHEAD VAPORS
WITH MOST. OF ORGANICS
FEED
DIRTY WATER CONTAINING
ORGANICS
STRIPPING MEDIUM
'LIVE I STEAM
OR AIR
Figure 2-1.
REFUX
BOTTOMS
CLEANED WATER
COOLING WATER
LIQUID OVERHEAD
-,
Simplified Diagram of a Stripping Tower
2-2
-------
. The basic design of a stripping tower involves specifying
the size of the tower (height and diameter) and the stripping
medium flow rate required to strip a given quantity of .feed down
to a desired purity. More completely, a design requires selecting
the type of stripping medium (steam or air), the operating
temperature, the type of contacting device (trays or packing),
and the combination of tower height and stripping medium flow
rate that will achieve the desired,.separation at minimum overall
cos t .
The applicable design procedure depends on the selections
. made amongst the following five types of operation:
1.
Batch and continuous. Continuous operation is more
effective in separating components of comparable volatility.
It provides higher purity of separated products and uses
less stripping medium for the same separation. Batch
stripping is of less commercial interest for the types of
separation under consideration here. We will restrict our
design to continuous systems.
2.
Mode of flow.
Three modes, concurrent, countercurrent and
---
cross flow, as shown in Figure 2-2, are possible. Cocurrent
flow is not generally used and is not addressed here,.
Cross flow operation is often preferred to couQterflow
because it provides greater transfer efficiency over a wider
operating range. Both are qiscussed.
3.
Isothermal or adiabatic.
In isothermal
operation,
the
temperature is constant along the length of the tower. If
, .
the stripper operates adiabatically, the temperature may
change significantly along the tower length. Steam stripping
is treated as being isothermal, and the feed is assumed to
enter the tower pre-heated to the boiling point. Heat
requirements are s at is fied by s tea:in condens ing at the
saturation temperature. Air strippers are treated as being
2:"3
-------
, r
, I
CO CURRENT
FLOW
COUNTERCURRENT
FLOW' .
CROSSFLOW
Figure 2-2.
Modes of Flow in Stripping Equl.prnem:
LIQUID
VAPOR
J ,
, ,
LIQUID
VAPOR
~ / LI:;
t \
VAPOR
2-4 ,
-------
adiabatic. Water usuall¥ evaporates into the air, and the
water stream is cooled. Saturating the air with water vapor
before it enters the tower avoids this problem and can
improve separation as the stripping operation then occurs at
a. higher temperature. This possibility is discussed in
tteference 3. The selection between air and steam stripping
is essentially dependent on process economics.
4.
Reflux. Reflux involve~ condensing some or all of the vapor
1 ea v ing the top of the s tripping tower, and returning some
or all of the condensate back tOo the tower. Reflux enhances
the separation, increases the concentration of the stripped
organics in the vapor stream, and is useful if the organics
are to be recovered. It is used with steam rather than air
stripping. Condensate may be taken to a decanter where an
organic phase separates arid can be incinerated or recovered.
The organic-saturated aqueous phase is returned to the tower
as reflux. The design procedures given here for steam
strippers are for both refluxed and non-refluxed towers.
5.
Mechanism of trans fer. This includes differential (staged)
contact and integral (continuous) contact. Both are widely
used and are discussed in detail under tray and packed
towers below.
. ,
2.1
TRAY AND PACKED TOWERS
In general, stripping is carried out either in tray towers
which prov ides for staged contact between the 1 iquid and vapor
streams, or in packed towers which provides c~ntinuous contact
between the two phases. The two types of towe~are illustrated in
Figure 2-3.
In tray towers the tower is fitted with regularly spaced
trays or plates of the type shown in Figure. 2-4. .T~e vapor passes
through openings in each tray and contacts the liquid flowing
2-5
-------
FHD
II.)
I
0\
CLEANED
WATER
OVERHEAD
VAPOR
. II \'-,
./ I . \ """'-
I I
FEED
PACKING
. ~ SUPPORT
, GR I 0
PACKED TOWER
Figure 2-3.
AIR OR
STEAM
ClEANEO
WATER
OVERHEAD
VAPOR
TRAY TOWER
3
Pictorial of Packed and Plate Towers
AIR OR
STEAM
-------
SLOTS (APPROXIMATELY 30
ALL AROUND PERIPHERY) CAP
. ~.
TOP VIEW
BASE OF INLET
DOWNCOMER
DOWNCOf-tER
y
VAPOR
FLOW
1t .
----CHIMNEY
PLATE
BUBBLE-CAP
SlOE VIEW
VAPOR 10
PLATE ABOVE
t t t t
WEIR
N
I
.......
., LIQUID FROM
f'V PLA TE ABOVE
DOWNCOMER
OPEN REGION (EXCEPT
FOR THIN SUPPORTS)
RISER
DOWNCOMER"
Y"
VAPOR.
FLOW
PLATE
t t t t.
VAPOR FROM
PLATE BELOW
J
LIQUID TO
PLATE BELOW
VALVE-CAP (PARTLY OPEN)
SIEVE PLATE
. 10
Methods for Dispersing Vapor into Liquid on a Tray
Figure 2-4.
-------
across the tray. A quantity of liquid is retained on each tray by
a weir. To reach the next "stage, the liquid flows over the weir
and through a "downcomer" which provides sufficient volume and a
long enough residence time for the liquid to be freed of
entrained vapor before entering the next tray. The liquid flows
down the tower under the force of gravity, while the vapor flows
upward under the force of a pressure drop from tray to tray.
The three common types of trays, name iy bubb 1 e-cap, sieve.
and valve trays, are shown in Figure 2-4. Bu};)ble-caps were used
widely in the past, "but have recently been replaced by sieve. and
valve types which are less expensive and just "as efficient.
Bubble-cap assemblies. are in the form of round bell caps and
common ly ha ve diameters ranging from 4 to 7 inches. The caps ha ve
slots or notches around the lower periphery which aids vapor
flow. The slots can be of a saw-tooth type or in the form of
punched holes, usually rectangular or triangular. The number of
bubble-caps to be used per tray is set by the allowable gas
velocity through the slots. Davies13 has recommended two
empirical equations for use in prel.iminary estimates of minimum
and maximum linear slot velocities in distillation towers.
The sieve tray consists of a flat plate perforated with many
small holes of 1/8- to 1/2-inch in diameter. The plates are
connected with one or more downcomers for handl ing 1 iq~id
overflow and may contain weirs and paffles for directing vapor
and liquid flows.
-
Valve trays are sieve trays with lift valves fitted over the
holes. Whereas with sieve trays, only the vapor flow prevents
liquid from "short-circuiting" and flowing down t"hrough the.
holes, in valve trays, the valve serves as a liquid seal while
allowing the passage of vapor. The major advantage of valve trays
is that high efficiencies can be maintained over a wider range of
opera ting conditions than with sieve trays. However, va love uni ts
2-8
-------
are more complex mechanically and are more expensive to fabricate
than sieve trays. Design information 'on valve trays is not
readily available.
Further information on tray configurations,
hydraulics is presented by Smithll.
des ign and
A packed tower is filled with divided solids of the type
shown in Figure 2-5. The packing materials may be made of
ceramics, metal or plastic, and a~e s~aped to provide a large
surface area. The liquid and gas compositions change continuously
with height of packing, as opposed to stepwise changes in tray
towers.
The packed tower is a simple device compared with tray
towers. A typical tower consists of a cylindrical shell
containing a support tray for the packing material and a liquid-
distributor. designed to provide effective irrigation of the
packing. Devices may be added to the packed bed to provide
redistribution of liquid that might channel down the wall.
Severa 1 beds in series may be us ed in the same tower she 11.
Because of its simplicity, the cost for a packed tower is
often considerably less than that for an equivalent tray tower.
. , .
Packed towers are particularly useful in cases:'where the' pressure
drop must be low and where liquid hold-up must' be small.
The packings may be either random or stacked. Random
packings are simply dumped into the tower. Pall rings, Intalox
saddles, Ras~hig rings and Berl saddles are mo~t often used for
random packings in industrial operations. Stacked packings are
those in which the individual pieces are arranged in a particular
pattern. The larger sizes of Raschig rings (3 inches or larger)
may be stacked. Rings are also available with internal spirals
... .....
which may be stacked one upon the other to provide continuous
pass ages for the gas. Wood grids and drip-point grids are other
2-9
-------
RASCHIG
RING
DRIP.POINT GRIDS.
. LESSING
RING
TELLERETTE
Figure 2-5.
PARTITION
RING
.~~
BERL
SADDLE
PALL RING
Tower packings12
2-10 ,
INTALOX
SADDLE
.. '. ,"
-------
examples of stacked packings. Stacked packings gi ve lower
pressure drops for equivalent throughputs than random packings'.
However, this advantage is gained at the expense o~ higher
initial costs due to the extra installation labor.
2.1.1-Selection Between Tray and Packed Towers
The choice between a tray an~ a packed tower is often made
on the basis of cos~s, but there are distinct advantages and
disadvantages associated with each type of equipment.that must
. . .
also be considered. The following summary of the relative merits
of tray and packed towers is based on that given by Peters and
Timmerhaus 14 :
1.
If the operation involves liquids that contain dispersed
solids, use of a tray tower is preferred because the trays
are more accessible for cleaning.
2.
Tray towers are preferred if interstage cooling is required
to remove heats of reaction or solution. Cooling coils can
be installed on the trays or the liquid-delivery line from
tray to tray can be passed through an external cooler.
3.
When large temperature changes are invol v.ed, tray towers are
often preferred because thermal expansion or contraction of
the equipment may crush the packing.
4.
Random packed towers are seldom designed with diameters
larger than 4 feet, and diameters of commercial tray towers
are seldom less than 2 feet.
5.
Packed towers prove to be cheaper and easier to construct
than tray towers if highly corrosive fluids must be handled.
It is eas ier and cheaper to replace packings periodica lly
than trays.
2-11
-------
6.
7.
8.
9.
Packed towers are usually preferred for liquids that have a
tendency to foam.
The liquid holdup is considerably less in packed towers.
The pressure drop through packed towers may be less than the
pressure drop through tray towers designed for the same
duty. This advantage, plus the fact that the packing serves
to lessen the possibility of tower-wall collaps"e, makes
packed towers partic~larly desirable for vacuum operations.
Tray towers can operate efficiently over a wider range of
liquid flow rates than can packed towers.
Other considerations, while not directly related
performance of the two types of tower, may nevertheless
bearing on the selection made. These include:
10.
11.
12.
to the
have a
Design information for tray towers is generally more readily
available and more reliable than that for packed towers.
Because of liquid dispersion difficulties in packed towers,
,
the design of tray towers requires less safety margin when
the ratio of 1 iquid to gas flow is low.
Reliable design data for packed towers
obtained from experiment.
must
often. be
13.
The total weight of a dry tray ~ower is usually less than
-
that of a packed tower designed for the same duty. However,
if liquid hold-up during operation is taken into account,
both types of towers have about the same weight.
Tables 2-1 and 2-2 provide useful information to help the
designer select amongst the various trays and packings available
commercially.
2-12
-------
TABLE 2-1
*
RELATIVE PERFORMANCE RATINGS
OF TRAYS AND PACKINGS10
. Bubble
3
4
3
5
3
3
Trays
.£!E Sieve
4
4
4
3
4
5
Valve
4
4
4
5
4
4
Parameter
Vapor capacity
Liquid capac i ty
Efficiency
Flexibil i ty
Pressure drop
Cost
Packings
High-void Normal
5 2
5 3
5 2
.2 2
5 2
1 3
*
5= excellent; 4= very good; 3= good; 2"= fair; 1= poor
TABLE 2-2
SELECTION GUIDE* FOR TOWER INTERNALSIO
Trays
Bubble cap Sieve ~
.1 2
2 3
2 3
3 2
3 1
1 2
3 2
1 2.
valve
pressure, low
mo. derate
" high
High turndown ratio
Low liquid flow rates
Foaming systems
Internal tower cooling
Suspended solids
Dirty or
polymerized solution
Multiple feeds'
or sidestreams 3
High liquid flow rates 1
Small-diameter columns 1
Column diameter 1 to 3m 2
Larger-diameter comuns 1
Corrosive fluids 1 .
Viscous fluids 1
Low pressure drop 0
Expanded column capacity 0
Low cost 1
Reliability" of design 2
1
2
3
2
1
3
'3
2
2
1
2
2
3
*
0= do not use; 1= evaluate carefully;
2= usua 11 y appl icabl e; 3= best se 1 ection
. 2;..13
packings
Random Stacked
2 3
2 1
2 0
1 2
1 2
3' 0
1 0
1 0 "
1
1
3
3
2
2
3
3
2
2
2
2
o
o
o
2
2
1
1
o
2
3
1
1
-------
2.2
DESIGN APPROACH
The tower height, liquid and gas flow rates, and the degree
of removal of the organics are re lated by three important
concepts, namely:
1. Vapor-liquid equilibrium,
2. Equilibrium or theoretical
3. Stage efficiency.
stages, and
Vapor-liquid equilibrium relates the concentrations in the vapor
.' . .
and liquid when these phases are in equilibrium, and an .
equilibrium stage can be loosely interpreted as the tower height
required for the two phases to reach equilibrium. The stage
efficiency relates an equilibrium stage to actual conditions in a
stripping tower.
Stripping occurs because the dissolved organic, by virtue of
its vapor pressure above the solution, tends to vaporize into the
stripping medium until its concentrations in the vapor and liquid
.phases are in equilibrium. In air stripping, some water will also
strip (evaporate) into the gas phase until the air becomes
saturated. Evaporation or condensation of the liquid or vapor
streams may occur in steam s tr ipping as dictated by the .therma 1
effects of the stripping process.
The flow rate of the liquid and gas streams may consequently
change due to both mass tran~fer and thermal effects. These
changes are particularly significant .for concentrated feeds and
complicate the design procedure. When stripping dilute
wastewaters with s.team or with nearly saturated air,. the amount
of mass transferred is small relative to the flow, and the liquid
and gas ra tes may be as s umed cons tan t. This as s umption
considerably simplifies the calculations.
2-14
-------
Thermal effects associated with heats of vaporization and
solution of the components as they pass from the liquid to the
gas phase tend to cool the water. In steam stripping, the tower
operates at the boiling point of the liquid at the tower
'pressure. The boiling point is not necessarily constant. This
temperature tends to increase down the tower with increasing
pressure, but to decrease with decreasing organic .concentration.
In air stripping, evaporation of w~ter causes the liquid to cool
towards the wet bulb temperature as it passes down the tower. Air
. . .
s~rippers are said to operate ~dia~atically. Changes in
t em pe rat u rea n d f low rat e can not, i n 9 e n era 1, be n e 9 1 e c t e d for
air stripping.
The rate of mass transfer from the 1 iquid to the vapor. phase
in a stripping system depends upon the difference between the
concentration of the solute in the water and the equilibrium
vapor concentration at the system temperature. The ease with
which a particular compound is stripped oc naturally volatilizes
depends on the relative volatility, which in turn depends on two
properties of .the solute - its vapor pressure. and its solubirity
in water. For aqueous mixtures containing low concentrations of
organic contaminants, the distribution of a contaminant between
the vapor phase and water under equilibrium c~nditions can often
be expressed by Henry's Law.. That '6ompounds with high vapot
pressures are easily stripped is expected. The effect of
solubility can also be explained quite simply. A compound which
is sparingly soluble in water is not compatible with water
molecules. In dilute solutions, each molecule of organic is
. surrounded by water molecules which want to push the organic
molecule away. The apparent vapor pressure of a poorly soluble
organic can be thousands of times higher over an aqueous solution
than over the pure organic.3
2-15
-------
2.2.1
Equilibrium Stage
A stage is a separate physical unit into which process
fluids are introduced, mixed, separated, and then removed. An
equilibrium stage is a stage in which the fluids leaving the
stage are in thermodynamic equi 1 ibrium. This. condi tion is
achieved if the streams entering the stage are mixed thoroughly
and for a sufficient length of time for the completion of the
required heat and mass transfer processe_so An equilibrium stage
is somj!times .called an ideal or theoretical stage. In tray.
stripping towers, an equilibrium stage is called a theoretical
plate, while in packed towers it is referred to as a trans fer
unit.
Process systems like stripping towers typically consist of
several stages, interconnected so that the materials being
processed pass through each stage in turn. The two principal
reasons for staging are to increase product purity and to reduce
con~umption of the separating agent. Multistage systems are
called cascades.
The equilibrium stage concept makes it possible to design
separation processes despite our poor knowledge of the complex
heat and mass transfer operations that occur on an actual stage.
. .
Based on this concept, the design of a staged pr9cess can be
divided into three steps: .
1.
Determine the equilibrium phase compositions,
. 20
Calculate the number of equilibrium stages required -to
accomplish a specified separation, and
3.
Estimate the actual number of trays or the height of
packing that is equivalent to an equilibrium stage.
.. '. ..
2-16
-------
The second step involves the relatively straightforward
application of.equilibrium, material-balance and enthalpy-balance
relationships. These are developed in Section 3. In the .remainder
of this section, we discuss vapor-liquid equilibrium and the
concept of stage efficiency in more detail. This is because some
empiricism and simplification is required to obtain practical
models suitable for design work.
2.3
VAPOR-LIQUID EQUILIBRIUM
Vapor-liquid equilibrium relationships relate the
concentration of. a species in one phase to the concentration in
the other phase when the two phases are in equilibrium. For a
detailed description of the theory and estimation of vapor-liquid
equilibria, the reader is referred to one of the many texts, for
example References 15-17. Here we will provide a brief
description suitable for the design procedures presented. in.. this
manual.
If we express concentration in terms of fugacity the
equilibrium "concentrations" in the two phases are identical:
fV = fL
(Equilibrium)
(2.l)
This is the fundamental vapor-liquid equilibrium relationship. By
writing the fugacities in terms of commonly used concentration
terms, we can convert Eq. (2.L) to a form sqitable for design
work.
2.3.1
Gas Phas e
The mole fraction of a species i in the gas phase is the
ratio of the number of moles of the substance ni to the total
number of moles N present:
Yi = (ni/N}v
(2.2)
2-17
-------
According to Dalton's law of partial pressures, one can write:
Pi = (ni/N)Vp
(2.3)
where
Pi is the partial pressure of species i, and
P is the total pressure of the system.
Combining Equations (2.2) and (2.3) we obtain:
Pi = Yi P
(2.4)
Fugacity is
of the species
Consequently,
defined as being equal to the partial press ure
provided the vapor behaves as an ideal gas.
fV,i = Pi = Yi P
(Ideal .gas)
( 2 . 5 )
A correction factor called the fugacity coefficient should be
introduced for real gases, but we will assume that pressures are
low enough and temperatures high enough for the gas phase to be
considered ideal.
2.3.2
Liquid Phase
. The mole fraction xi of a species i in the li~uid phase is
defined analogously to that in the vapor phase by: .
Xi = (ni/N)L
(2.6)
The fugacity is related to the mole fraction by the expression:
fL . = y. x. f9L .
,1. 1. 1. ,1.
(2.7)
Two new concepts have been introduced in Eq. (2.7). The standard
state fugacity fO, which can be thought of as the vapor pressure
.', ...
of the pure species, p., at th~ temperature of the solution,
2-18,
-------
again assuming the gas phase is ideal. That is:
~ = p*
(2.8)
The activity coefficient compensates for non-idealities in the
liquldphase. From Eqs. (2.7) & (2.8) we get, for a system
containing an ideal gas and a real liquid:
*
fL,i = Yi xi Pi
(2~9)
Subs ti tution of Eq. (2.9) for the 1 iquid and Eq. (2.5) for
the vapor phase into the basic vapor-liquid equilibrium
relationship - Eq. (2.1) - yields an equilibrium relationship
which can be applied to. design work:
. *
Y. = Y. X. P. /p
1. 1. 1. 1.
(2.10)
All terms in this equation are either known or can be easily
calculated. The vapor pressure is a function of temperature, and
the activity coefficient is a function of both temperature and
concentration, so the problem now becomes one of evaluating these
parameters at the conditions prevailing in the stripper.
2.3.3
Henry's Law
Many solubility data for gases and volatile organics
published in the literature are in terms of. their Henry's Law
constant, B. This constant is a product of the activity
coefficient and pure component vapor pressure:
*
Hi = Yi Pi
(2.11)
Substituting in Eq. (2.10) results in Henry's Law:
Yi = (Hi/P) xi
(2.12)
2-19
-------
When applying this equation .to stripping and distillation
calculations, it is customary to write it in terms of a K-factor,
Yi ::I Ki xi
(2.13)
where
K = a/p
Dimens ions .
Mole fractions are' dimensionless, so B has the
units of pressure in Eq. (2.12). Other concentration units may be
used, in which case the reported B value will have dimensions of
pressure divided by concentration. For di~ute solutionstheseB
values may be converted to pressure-mole fraction dimensions
suitable for use in Eq. (2.12) by multiplying by the following
factor:
concentration. Unit
Factor
mq/L or ppm
moles/L
mOles/m3
mass percent
106 M.Wt/18
103/18
106/18
102 M.Wt/18
Temperature and Concentration Dependence. Like the activity
coefficient, Henry's Law "constant" is dependent on both the
system temperature and the concentration of the substance 'in
solution. As it is' unlikely that tabulated B values will be found
for the conditions of interes t, it is necess ary that we be ab 1 e
to estimate these from the available data.
In cases where data is available in the range o! interest,
plotting or curve fitting and interpolation should be
satisfactory. Linearizing the data, as explained below, may
improve accuracy.
2-20
-------
Many relationships have been proposed between the activity
coefficient and concentration.17,18 At very low concentrations,
B may ' be assumed independent of concentration. At higher
concentrations we may extrapolate using the temperature
dependence of the activity coefficient. In Reference .3, the van
Laar equations for binary mixtures are used, and it is suggested
that the data be linearized by plotting
(Xl x2)/(Xl ln Yl ~ x2 ln Y2) against the concentration Xl.
Here, subscript! refers to the solute and 2 to the solvent
(water). The plot has intercept 11A and slope (lIB - l/A). The
values of A and B so obtained may be substituted in the van Laar
equation:
Log Yl = A/[l + Axl/(BX2)]2
(2.14)
to obtain values for the activity coefficient ~t. other
concentrations. It must be remembered that A and B are functions
of temperature, so isothermal plots must be used.
Compensating for the effect of temperature on Henry's Law
constant is more involved because both the. acti vi ty coefficient
. .
and the vapor pressure vary significantly with temperature. One
method of predicting vapor pressure as a function of temperature
uses the Antoine equation: .
log p* = A ~ B/(~ + T)
(2.15)
A, B, and C ar,e the Antoine constants (differen~ from van Laar's
coefficients), and T is the temperature. Reference, 16".lists
values for these constants for a number of chemicals; A:~~SIC
computer program to find values for A, B, and C from thr~e data
points in any (mixed) units is given in the Appendix.
2-21
-------
Reed, Prausnitz and Sherwood17 suggest using:
log y
(constant composition) = D + E/T
(2.16)
where
T is
D is
only
the temperature in Kelvin, and
a constant that may be assumed
one data point is available.
equal to zero if
The correlations suggested above may be used for a wide
, -
variety of substances, but are not universally applicable. Refer
to Reference 19 for more details.
Data. Values of Henry.s Law constant for the organics on
the, toxic pollutant list that were estimated in Reference 3 are
listed in the Appendix. The equations for B are ,derived from
correlations for the vapor pressure and the activity coefficient.
In most cases, the activity coefficient was calculated from
solubility data. For the few toxic pollutants which are miscible
with water, the activity coefficient was estimated' from vapor-
liquid equilibrium data or from the azeotrope data ~~ing the
technique of Van Laar. The reader is referred to the,original
reference for information on these determinations.3 " ,
2.4
STAGE EFFICIENCY
Stage efficiencies describe the extent tQ which the
performance of an actual stage in the equipment duplicates the
performance of an equilibrium stage used in the design
calculations. An equilibrium stage n?rmally requires,a liquid
depth of at least 60 em, the vapor must be well distributed and
in fine bubbles, and flow rates must be low. Such a s~age may be
used for measuring the Henry.s Law constant experimentally but
are not practical for trays installed in commercial; to'Wers. The
relationship between an actual stage or tray and equilibrium
stage is normally determined using empirical and simplified
mathematical models. . '.,
2-22
-------
2.4.1. Types of Stage Efficiencies
The overall tower efficiency, Eo, is defined as tbe number
of theoretical stages required to produce. a given separation
divided by the number of trays actually necessary to produce the
same separation. Although the overall tower efficiency has no
fundamental mass-transfer basis, it is widely used because of its
simplicity. The number of actual stages. required for a given
separation is equal to the number of theoretical stages divided
by the overall tower efficiency.
The Murphree plate efficiency, EM' applies to a single
stage. It is defined as the ratio of the actual change in average
vapor composition accomplished on a tray to the change in average
vapor compos i tion if the vapors 1 ea v ing the tray were in
equilibrium with the liquid leaving the tray. The Murphree plate
efficiency has a more fundamental basis than the overall tower
efficiency, but is less convenient to use because it must be
applied to each individual tray.
The point efficiency, Ep, is similar to the Murphree plate
efficiency, except that the point efficiency applies to a single
location on a given tray. This efficiency is of considerable
theoretical interest but is seldom used in design because i~
requires knowledge of the variations in liquid. compos i tic::m across
the tray and integration of the point efficiencies over the
entire tray.
2.4.2
Tray Towers
The approach to equilibrium depends largely on the
effectiveness of mixing the liquid and gas phases on the tray.
The bverall efficiency has been found to depend on the following
factors.16,28
. 2-23
-------
1.
Viscosity. Efficiency increases significantly as the liquid
viscosity decreases. Different authors report different
express ions for this dependence but the variation is
approximately:
Efficiency - {viscosity)-n
0.7 < n < 0.9
(2.17)
Since the viscosity of water decreases from 0.89 c~ntipoise"
at 2SoC to 0.28 at 1000C there" will be a significant
increase in efficiency with temperature.
2.
Liquid depth. The efficiency increases as the depth of
liquid on the tray is increased. The price is an increased
pressure drop.
3.
Vapor~. The efficiency is not much dependent on vapor
rate up to the point where frothing and entrainment occur.
Entrainment causes liquid to be mixed backwards up the tower
and the efficiency to decrease.
4.
Liquid~. Since the liquid on bubble-cap and valve trays
flows across the vapor, it is possible for there to exist
more than one equilibrium stage on a tray. Insofar as an
increased liquid rate causes more back mixing in the liquid,
the efficiency falls somewhat as the liquid rate increases.
The effect is dependent on tray design.
5.
Length of liquid path. The length of the liquid path across
a tray is an important factor in"determining the degree-of
liquid concentration gradient across the tray. In general,
" "
as the length of the liquid path is increased, the overall
tower efficiency increases. The effect is usua~ly negligible
if the length is less than 5 ft., but increas ing the length
to 10 to 15 ft. may increase the overall tower efficiency by
20 to 40 percent.
2-24
-------
6.
7.
8.
9.
Tray spacing. The effect of tray spacing is related to the
superficial vapor velocity. If the vapor velocity is greater
than the allowable value, too small a tray spacing can cause
entrainment carry-over and a decrease in efficiency.
Hul ticomponent separations. In mul ticomponent separation,
the assumption is usually made that the same efficiency
applies to all components being separated. This assumption
may not be correct for some mixtures where properties of the
components (such as liquiddiff,usivity and Henry's Law
,
constant- see 8 below) are significantly different.
Henry's Law Constant. The efficiency may decrease as the
Henry's Law cons tant increas es for certain organic
compounds. This is a most important consideration in the
design of strippers because we are mos.t often concerned with
compounds having a high Henry's Law constant. It has been
. .
observed that there is not necessarily' a direct correlation
between a compound's Henry's Law constant and the removal
efficiency. This means that for the same conditions, a
compound A with Henry's Law cons tant much higher than that
of compound B, may not be much better stripped than compound
B. This effect of Henry's Law constant results when the rate
of diffusion of the organic through the 'liquid (rather than
across phases) starts to control the overall rate of mass
trans fer. 3
Other factors. Design details of the tower, such as vapor
opening dimensions, tray layout, or the total number'of
trays can affect the efficiencies.
In general, the three types of trays discussed in Section
2.1 have stage efficiencies in the range of 80 to 90 percent.14
Sieve and val ve trays achieve higher eff~cienc;i~ than bubbl e-
caps.
2-25
-------
2.4.3
Packed Towers
Packing efficiency, which is defined as the ability of a
gi ven packing to achieve effective mass trans fer between a gas
phase and a liquid phase, is inversely related to the height of
packing equivalent to one transfer unit (HTU). A packed tower is
more efficient the smaller the value of HTU. Apart from
geometrical considerations related to the shape and arrangement
of the packing, the packing efficiency is affecte~ by the
following factors:
Packing efficiency
Factors
Increas es
Viscosity decreases
Liquid flow rate decreases
Little change
Vapor flow rate alters
below the. flooding rate
Decreas es
Henry's Law cons tant or
relative volatility increases
The height of
actual height
efficiency.
a transfer unit determined in Section 3 is the
and. incorporates the factors related to packing
2-26
-------
SECTION 3
DEVELOPMENT OF DESIGN PROCEDURES
A complete process design for a steam stripping system to
recover organics from an aqueous feedco~ld involve specification
of all the units shown in Figure 3-1. The system shown in this
figure has two stripping towers operating in series, with a
common decanter for separating the condensed overheads into
'organic-rich and organic-lean streams. In the system shown,. the
feed is preheated by heat interchange with the bottoms.
The design procedures presented here will be broken down
into towers, heat exchangers (heat exchanger, reboilers and
condensers) and storage vessels (feed tanks, decanters,
accumulators, etc.). From these basic components, the'designer
may synthetize any desired flow sheet. Note that while the design
o~ pumps, flow controllers and instrumentation is not given, the
costs for these components are included in the costing procedures
given in Section 4.
3.1
TRAY TOWER DESIGN
The des ign of a tray tower cons is ts
determinations :
of the following
1. The number of stages theoretically necessary for the
required separation,
2. The stage efficiency of the.. tray:; ..relating
theoretical pla~e to the actual trays,
the
3-1
-------
(.oJ
I
N
TREATED
WATER
--- TOWER 1
F££D
F££D
TANK
,.------
I
ACCUMULATOR/
DECANTER
-10-0
g-- --]
t
I
FEED PUMPS
BOTTOMS PUMPS
------ ------...1
STRIPPED
ORGANICS
WATER RECYCLE PUMPS REFLUX PUt4PS
Figure 3-1.
Process Flow Diagram" for Steam Stripping
-------
3. The diameter of tower necessary to avoid flooding or
excessive entrainment, and
4. The pressure drop across the tower.
3.l.L Number of Equilibrium Stages
The number of equilibrium stages required to.effect a .
desired separation can be determined from a material balance
over the tower. Using the nomenclature in Figure 3-2, an overall
material balance for a tower without condenser is:
F + A = B + Vl
( 3.. 1 )
Here, A is the total amount of live steam or air introduced at
the base of the tower. It is not necessarily equal to Vl; the
vapor leaving the top of the tower as transfer of solute and
evaporation or condensation takes place along the length of the
tower. If reboil steam is used, A is zero as no additional
. material is introduced into the tower. The design calculations
are not otherwise different. The effect of using reboil steam is
that the bottoms will be more concentrated (for the same steam
rate and tower height), but more importantly, clean water (steam)
is not introduced into the tower and contaminated. The price for
this is the cost of the reboiler.
A material balance for a component with mole fraction %F'
%B' and Yl in the feed, bottomS and overhead vapor respectively,
becomes:
F xF = B xB + Vl Yl
(3.2)
We can define a fractional removal efficiency, f as the
fraction of the solute in the feed that is removed in the
overhead:
3-3
-------
OVERHEAD rAPOR
Vl,Vl
(NO CONDENSER)
At
- - OVERHEAD VAPOR
VI,VI
(PARTIAL CONDENSER)
I
- - - - ~ - - - OVERHEAD LIQUID
/" REFLUX D~ xD . .
Lo' xD . (TOTAL CONDENSER)
TOP STAGE
FEED
F, xF
y--
E
~
~
I Ln LJ
n-l
STAGE n
n+l
BOTTOMS
B, xB
STRIPPING MEDIUM
A
(LIVE STEAM OR AIR)
Figure 3-2.
Stripping Tower Nomencl.ature. ...
3-4
-------
f = Vl Yl/(F xF)
(3.3)
In qeneral, we specify the strippinq medium rate at t~e top of
the tower At (equal to Vl for the case of no condenser) and the
removal efficiency required, and calculate the resultinq overhead
composition:
Y1 = fFxF/At = fFxF/Vl
(3.4)
The vapor composition on any staqe n in the. tower is
obtained by doinq material balances .from that staqe to the top of
. .
the tower (See Figure 3-1):
Overall Material Balance
~ + Vn+l = Ln + Vl
(3.5)
Balance for Solute
F xF +Vn+lYn+l = Ln xn + Vl Yl
(3.6)
From these two equations, we obtain the vapor composition on
staqe n+l as:
Yn+l = (Vl Yl + Ln xn - FXF)/(Vl + Ln - ~)
( n> 0 )
(3.7)
The composition %n is the liquid composition on
and is in equilibrium with YD, the vapor composition
tray. Usinq Henry's Law we can write:
the nth tray
leavinq that
Xn = Yn/Kn
(3.8)
Eq. ( 3 . 7 )
now'becomes
Yn+l= (Vl Yl + Ln Yn/Kn - FXF)/(Vl + ~ - F)
(3.9)
. 3.-5
-------
For example, with n=l
Y2 - (Vl Yl + Ll Yl/Kl - FXF)/(Vl + Ln - F)
"Eq. (3.9) can be-applied with successive values of n until a
vapor composition is obtained that equals or is less than the
vapor leaving the bottom tray, YB~ We get the liquid composition
leaving. the tower, zB' from Eqs. (3.1) and (3.2) as:
Xa = F xF (1 - f)/{F + A - Vl)
:: xF (1 - f)
(3.10)
50
Ya :: xF(l - f)/KN
(3.11)
tfthe tower operates isothermally, and K is
the l.iquid and vapor rates are constant with L
V = Vo =At = Vo then Eq. (3.9) simplifies to
constant, and
= P = Ln and
Yn+l = Yl + (L/VK)Yn - (L/V)xF
(3.12)
The factor.. (V K / L) , called the stripping factor, S , is an
. .
important parameter in correlating the performance of strippers.
Thepreceeding equations apply to the case of no condenser.
. ,
Next we have to consider the use of condensers above the tower.
Either a partial condenser or a total condenser may'be installed,
and some or all of the condensate may be returned to the tower as
reflux. The amount of reflux returned to the top of the tower is
defined in terms of a reflux ratio R. The condenser may- be
followedb~a decanter in which the liquid is separated into
organic rich ~nd organic lean fracitions. The reflux .to the tower
may comefro~t~. condenser or decanter, 6r from another source.
Thefollawing specific conditions apply to the equations
given hate for use with condensers:
3-6
-------
If a partial condenser is used, it is assumed to be the top
equilibrium stage in the tower. All the liquid leaving the
partial condenser is assumed to be returned to the tower, so
D = O. If a total condenser is used, the reflux composition
returned to the tower is the same as the vapor composition
leaving the top of the tower, that is xD = Yl. This means
that a decanter is not included in the flow-sheet. A total
condenser is not an equilibrium stage and as the name
implies, no vapor leaves the system.
The design equations are summarized in Table 3-1. Note that
the generic equations for Yl' Yn+l and xB reduce to those given
above for the no condenser case. If reflux is not used, the feed
,must enter on the top stage. If reflux is used, the feed may
enter lower down the tower. Note that pmust be zero in Eqs.
(23.14) & (3.15) when determining Yn+l for stages above the feed
tray. The lower the feed tray from the top, the more the overhead
vapors are enriched in organics. Two trays above the feed is
typical for refluxed strippers. If decanters are included in the
, ,
flow sheet, do not use the equations in Table 3-1; the simplified
equations given later should be used.
Both the stripping medium rate, At, and t~e refl uxratio, R,
must be specified when determining the number of stages.' If these
are not known, a value of At equal to 15 percent of the feed for
no condenser, and up to 25 per~ent of the feed for a condenser
with reflux may be used initia'lly. If a decanter is used, the
entire aqueous part of the overhead is normally returned as
reflux. The reader is referred to Reference 12- for a discussion
on the minimum stripping medium rates and the effect of reflux on
tower performance.
, 3-7
-------
TABLE
3-1
SUMMARY OF STREAM RATES AND COMPOSITIONS.
No Partial Total
Variable Condenser Condenser Condenser
Stripping medium Steam or air Steam Steam
Specified parameters
Stripping medium rate, At VI V 1 + Lo Vl (=Lo+C)
Reflux ratio, R 0 . . LO/Vl L~/D
Overhead liquid Lo' C = 0 C = 0 Lo' C>O
Removal efficiency, f V lY1/ (FxF) V 1Y1/ (FxF) CXC/(FxF)
Calculated Parameters
Composition of
overhead vapor
Yl = fFXF(R + l)/At
(3.13)
Composition of
overhead liquid, Xc
Eqm. with Y1
XC.= Y1
Tower vapor
composi tion
n>O
AtYl/(R + 1) + Lnxn - FXF
Yn+1 =
(3.14)
At/(R + 1) + Ln - F
Ln xn - FXF (1 - f)
or
Yn+l =
(3.15)
At/(R + 1) + Ln - F
Bottoms
composi tion
Fx F (1 - f)
XB =
(3.16)
F - A - At/(R + 1)
.
For design of adiabatic stripping towers without decanters.
3-8
-------
Liquid and Vapor Rates. The liquid and vapor rates vary
along the tower length due to ~ransfer of the solute and
evaporation or concentration of water. To determine these flows,
we must combine enthalpy balances with the material balances
given previously. Before doing s6, it is instructive to first
exam;ne under what conditions these rates may be assumed
constant.
If the feed is relatively dilute, then both the heat and
mass transfer effects of the stripping solute may be assumed
negligible. Most organic contaminan1:;s are.sparingly soluble3, and
we will normally not have to consider concentrated feeds. If more
concentrated feeds are encountered, however, the equations can be
written on a "solute free" basis. Instead of Vn being the total
moles of vapor, it becomes the moles of pure stripping medium,
which is constant. Concentrations must then be expressed in terms
of mole ratios, (X, y), that is moles of solute per mole of
stripping medium, or per mole of water, instead of mole fractions
(x, y). Heat effects must also be accounted for with concentrated
feeds. A method for including an enthalpy balance in .the solution
is discussed under adiabatic air stripping below, and
incorporated in the BASIC design program in the Appendix.
In steam stripping, the feed can be assumed to be a
saturated liquid and each equi~ibrium stage. taken to be at the
saturated water vapor temperature. Because the- pressure .increases
from the top to the bottom of the tower, so does the saturation
temperature. Consequently, some. steam must condense on each stage
to supply the heat necessary to-raise the liquid temperature. For
low pressure drops, this condensation may be neglected and the
column assumed to operate isothermally with Ln ~nd Vn constant.
In air stripping, water will generally evaporate to saturate
the air, and the 1 iquid stream wi 11 coo 1 down towards the wet
bulb temperature. A constant water rate equal to the feed rate
less the water vapor transferred to the air may be assumed for
3-9
-------
rough estimates. If computers are not available and simplified
isothermal calculations must be made, then it is suggested that a
temperature which. is the. arithmetic mean of the feedwater
temperature and air wet bulb temperature be used3.
Because of the great difference in ~ower performance that
can result between summer and winter when stripping with air, it
has been suggested that the cooling effect be eliminated by
saturating the entering air with steam.3 If th~ air is
simultaneously heated to near the liquid feed temperature, then
the simplified design equations for constant flows and isothermal
conditions should satisfactorily approximate t~e iower
performance.
Simplified Equations for Steam Stripping
Making the assumption of constant liquid and vapor rates
along with isothermal conditions considerably simplifies the
calculational procedure. Because the Henry's Law constant is now
also invariant, an analytical solution can be obtained to the
design equations. The solution is summarized in Table 3-2.
These equations, based on a derivation by Kremser12, are
applicable to most steam stripping operations.
Adiabatic Stripping with Air
In general, the simplified equations will not be suitable
for. air stripping. The following procedure should be used
instead. The equations are derived for strippers with~ut
condensers treating dilute wastewaters. (the usual application £or
air stripping). The equations in Table 3-1 for no condenser
apply. The vapor rate is constant if we use pure .(water- and
solute-free) air for V. We must express vapor concentrations as
mole ratios Y, where
y = y/ (1 - y)
(3.19)
.. '"
3-10
-------
TABLE 3-2
*
DESIGN EQUATIONS
FOR TRAY TOWERS20
f = (SN+l - S)/[SN+l - S + k(S - 1)]
(3.17)
N = In[S + f k(S - 1)/(1 - f)]/ln S -.1
(3.18)
where
f = fractional removal ef~iciency of a solute
S = stripping factor (= KV/L) > 1)
N = number of theoretical stages
k = (1 + R' YD/YS)/(R' + 1)
R'= R {l + [Cp(tb'- tR)/L]}
R = reflux ratio
tR= reflux temperature
tb= boiling point of the aqueous reflux
L = latent heat of steam
Cpa specific heat of reflux
YD= activity coefficient of pollutant in the organic
phase (= 1.0 for pure organic)
YS= activity coefficient of pollutant in the aqueous
phase3,20
. .
. For isothermal operation, constant flow rates, with or without
reflux and decanter.
3-11
-------
The temperature and the amount of water evaporated for each
stage can be determined from an enthalpy balance. An enthalpy.
balance over the top portion of the tower is:
F hf + V in+l = V il + Ln hn
(3. 20 )
h is the water enthalpy in calories/mole,
i is the humid air enthalpy in calories/mole
V is the pure, ,dry air rate in moles.
The amount of water transferred on stage- n is given by
Here
dry air,
Wn = (Hn - Hn+l) V
moles' water
(3.21)
where a is the humidity, moles water vapor per mole dry air.
The above equations can be solved together with the
following correlations for enthalpy and humidity.3
Water enthalpy:
h = 18 t
cals/mole water
(3.22)
Enthalpy of humid air
i = 6.96 t + H(lO,750 + 8.1 t)
cals/mole dry air
(3.23)
Mole fraction water vapor in air:
y = (1/P)exp[21.2 - 592l/T
6.98xlO-3T]
(3 . 24 )
Humid i ty
H = y/ (1 - y)
moles water vapor/mole air
(3.25)
Where
t is the temperature in °C, T is in Kelvin, and
P is pressure in psi
3-12
-------
We must know the top temperature to initiate the
calculations. A temperature ~hich is the mean of the feed
temperature and wet bulb temperature of the entering air may be
used as a first guess. The calculations then proceed stagewise
down the tower. First the temperature and water rate from the top
stage (n = 1) are calcutated using the above equations. Next the
equiJ..ibrium concentrations are calculated using the Henry's Law
const~nt at the calculated temperature and the equations in
Table 3-1. The procedure is then repeated for the second stage
and so on until the bottom stage is reached. If the feed air
temperature does not match the calculated temperature ~or the air
entering the bottom stage, a new top temperature must be guessed.
A computer program for these calculations is given in the
Appendix.
Some results obtained using the computer program are shown
in Figure 3-3. These data illustrate the effect of the
temperature of the liquid and air feed, as well as the air
humidity, on stripping tower design.
3.1.2 Tower Diameter and Height
The tower diameter and consequently its cross section must
be sufficiently large to handle the gas and liquid at velocities
which will not cause flooding or excessive entrainment. The
diameter can be obtained from a determination of the maximum
allowable vapor velocity. The following equation applicable for
any specific location in the tower may be used21,22:
v m = Kv ( a /20 ) 0 . 2' [ ( P L - P G ) / P G] 0 . 5
(3.26)
where
vm = max. allowable vapor velocity based on cross-
sectional area of empty tower, ft/s
Kv = empirical constant, ft/s
PL' PG = density of liquid and gas respectively, and
a = liquid surface tension, dyne/ern .
:
3-13
-------
9
..J
<
>
o
x
~
cr::
a4
a'I
.
a'I
a'I
cr::
o
1.1.
Q
'-'J
=:
-
So 5
~
cr::
""
~
c.D
~
""
7
LIQUID FEED = 20°C
AI~ FEED = 10°C; RH = 5%
LIQUID FEED = 35°C
AIR FEED = 30°C;
RH = 70%
a
50
100 .150
AIR RATE, % OF FEED
200
30
LIQUID FEED = 35°C
AIR FEED = 30°C; RH = 70%
AIR RATE = 63%
u
°
~ 20
:::I
I-
<
cr::
~
c..
X
~
I-
~ 10
<
I-
"",
LIQUID FEED = 20°C
AIR FEED = 10°C; RH = 5%
AIR RATE = 80% OF LIQUID FEED
a
a
2
4
6
8
10
Figure 3-3.
STAGE NUMBER
Effect. of Temperature on Air Strii?per qesign
3-14
-------
The constant Xv is dependent mainly on the tray spacing but
also on the depth of liquid on the tray, the ratio of liquid to
gas flow rate, the density of the gas and liquid, and the
geometry of the tray components. Values for the empirical
constant can be obtained using the equations given below.12
For bubble~cap trays:
.KV= a log(l/c) +.b
(3.27)
For perforated trays:
KV= [S(Ah/Aa) + O.S][a log(l/c) + b]
(3.28)
where
Aa = Acti ve area of the tray which is the tota 1 cros s-
sectional area of tray minus the area of the
two downcomers (ft2).
Ah = Hole (perforation) area per tray.
c = (L/G)(PG/PL)0.5
The va 1 ues of a and b and the recommended va 1 ues for tray spacing
are given in Table 3-3. A tentative value of 0.13 may be used for
Ah/Aa in the absence of other data.
Although' the allowable velocity given by these methods is
cons ervati ve for many types of operations., one can employ
. ..
velocities in the range of 65 to 85 percent of Vm to ensure the
tower will be operable. The volumetric flow rate of the vapor is
a = G/(3600 PG)
ft3/s
and the cross-sectional area required to give a velocity v ft/s
is
A = 1.1 a/v
ft2
(3.29)
The factor 1.1 is to allow for area occupied by the downcomers.
3-15
-------
TABLE 3-3
RECOMMENDED VALUES FOR TRAY SPACING AND FLOODING CONSTANTS
FOR TRAY TOWERS 1.2
Tray spacing, t
Tower diameter, ft Tray spacing, inches
4 or less 18-20.
4-10 24
10-12 30
12-24 36
Flooding constants in terms of tray spacing, t
Bubble-cap trays
Value of c
0.01-0.03
0.03-0.2
0.2-1. 0
a
Us e val ues
b
0.0041t + 0.0135
0.0068t + 0.049
for 0.03
0.0047t +
0.0028t +
0.068
0.044
Perforated trays
0.01-0.1
0.1-1.0
Us~ values for 0.1
0.006t + 0.04 0.002St+0.OS
3-16
-------
If .the vapor rate changes over the length of the tower, the
theoretical diameter based on the available vapor velocity will
also change. Vapor rates at different heights of the tower
(usually bottom, middle and top) can be used to estimate
diameters. Occasionally, two different diameters are used for
different sections of one tower. When the variation in flow is
consiaerable, and especially when expensive alloys are used in
construction, use of two diameters may be considered. Usually a
single diameter based on the tow~r location where allowable
.velocity requires the largest diameter, is used.
Once the diameter has been determined, the recommended tray
spacing from Table 3-3 together with the number of trays as
determined in the next Section may be used to determine the
height of the tray section of the tower.
3.1.3
Stage Efficiency
The number of theoretical-stages determined must be divided
by the overall tower efficiency Eo to obtain the number of actual
stages required for the desired removal. There are several ways
to estimate the tower efficiency. Most of them are empirical and
developed for a particular application. The method presented here
resulted from a research program sponsoreQ by the American
Ins ti tute of Chemica 1 Engineers. Mos t of the~ork was done with
bubble-cap trays, but the few data together with subsequent
information indicate that the empirical expres~ions represent the
performance of sieve trays as. we 11. The method uti 1 izes the
concept of transfer units which are discussed more fully in
Section 3.2. The procedure is presented without discussion in the
~ppendix. The equations are given in a simplified form that are
valid for typical conditions: the reader is referred to Reference
23 for further details.
3-17
-------
3.1.4. Pressure Drop
As the vapor passes through the tray tower assembly, its
pressure decreases due to:
1. Pressure drop through the contactor assembly. This includes
Contraction loss as the vapor
Friction in the tray opening
bubblecap unit is used:
Friction due to change in direction of vapor flow for
bubblecap and valve units:
Friction as the vapor th~ough the ~lots fora bubblecap
unit.
enters the tray openings:
and in the annular space if a
2. Pressure drop due to the static liquid head above slots, sieve
openings, or valve openings.
A detailed me~hod of calculating pressure drops for
individual steps is given by Peters and Timmerhaus14. For a
bubblecap tower, the following pressure drops per tray are
considered typical, and may be used for sieve or valve trays if
other data are not available.
Total pressure
Pressure drop per tray
30 rmn Hg
1 atm
300 ps i
3 mm Hg or less
0.07-0.12 psi
0.15 psi
3-18
-------
3.2
PACKED TOWER DESIGN
The design of a packed tower consists of the following
s tepa :
1. Selection of the packing type and size.
2. Calculation of the total height of packing required.
3. Estimation of the tower diameter to avoid flooding.
4. Estimation of the pressure drop.
3.2.1 Selection of Packing
The types of packing available are described in Section 2.1,
and a summary of their important characteristics is given in
Table 3-4. There are many packing types on the market and it is
frequently advantageous to compare the characteristics of more
than one packing type and size before settling on a final design.
The following properties are representative of a satisfactory
packing:
1. Low pressure drop and high capacity. A large free cross-
sectional area should be available between the packing particles
to give low pressure drops and high fluid rates.
2. Large active surface ~ per unit ~l~. To give high
efficiencies, the packing must provide a large amount of contact
. ,
area between the two fluid phases.
3. Low weight ~ low liquid holdup. The total weight of the
tower and the resultant foundation load is low if the weight of
the packing and the liquid holdup in the tower are low. The
amount of liquid holdup, however, must be sufficiently great to
retain an effective driving force for mass transfer.
4. Misce 11 aneous. High durabil i ty, high corros ion res is tance,
low side thrust on tower walls and low cost must all be
considered in the selection of a packing.
3-19
-------
3.2.2 Tower Diameter
The flow rates through the packed bed must first of all be
high enough to ensure that the packing is .loaded., that is that
the packing surface is continuously wetted and good distribution
of the liquid and vapor flows occur. However, if the flow rates
are too high, the packing will .flood. with rapid increase in
pressure drop and loss of performance. The tower diameter is
generally selected to be as small as possible for economic
reasons, but large enough to ensure that. flooding will not occur
under anticipated operating conditions.
Flooding and pressure drop data are provided by packing
manufacturers. Figure 3~4 along with Table 3-4 may be used for a
variety of packings.
The method of calculating' the tower diameter using
Figure 3-4 can be summarized as follows:
1.
Calculate the parameter (L/G) (PG/PL)O.S. Note that Land G
are in mass units.
2.
At flooding, find from Figure 3-4 the corresponding value of
the parameter (with ~L in centipoise):
(G t ) 2C ( ~ ) 0.2/ (g P P )
- f L c G L
3.
Calculate Gt, the superficial mass velocity, (G/A).
4.
Reduce
ratio.
this flow by a flood safety factor, called the flood
A flood ratio of 0.7 to 0.8 is typically used but it
lower ,if dictated by pressure drop limitations.
may be
S. .
The tower cross-section is the entering flow rate of the
gas, G, in lb/hr divided by qt in Ib/(hr.ft2).
3-20
-------
N
.
o
-
0.40
0.20
0.10
0.06
0.04
-I -I
;1 a.
- J!' 0.02
u"" - u
N D'I
-
..
(.!)
-
w
I
N
.....
0.01
0.006
0.004
0.002
.
0.001
0.01
~p 1b/ft2 = 5.2 (in. If )
Z ft water t
1.50
1.00
0.50
0.25
0.10
0.05
L GAS PRESSURE DROP.
IN. WATER/ft °
0.02
0.2 0.4 0.6
L J PG
GO Pl
2
4
6 8 10
0.04
0.1
1.0
12
Figure 3-4. ' Flooding and Pressure Drop in Packed Towers
K fLc~ct.uf~.J
-------
TABLE 3-4
CHARACTERISTICS OF COMMERCIAL PACKINGSl2
Data for Wet-Dumped (Random) Packings *
Nominal size, inches
Packing 1/2 5/8 3/4 1 1-1/2 2
. Raschig rings
Ceramic
Cf 640 380 255 160 95 65
e: 0.63 0.68 0.73 .0.73 0.71 0.74
ap 111 100 80 .58 38 28
Metal
Cf 340 290 230 145 82 57
e: 0.73 0.78 0.85 0.90 0.92
a~ 118 71.8 56.7 41.2 31.4
Pa~~ r1Dgs
P 1as tic
Cf 97 52 32 25
e: 0.88 0.90 0.905 0.91
ap 110 63 39 31
Metal
Cf 71 48 28 20
e: 0.90 0.9", 0.95 0.96
ap 131. 2 66.3 48.1 36.6
Inta~oz sadd~es
Ceramic
Cf 265 130 98 52 40
e: 0.78 0.77 0.78 0.81 0.79
ap 190 102 78 59.5 36
Berl saddles
Ceramic
Cf 380 170 110 65 45
e: 0.63 0.66 . 0.69 0.75 0.72 .
ap 142 82 76 44 32
.Ct= packing factor, ft-1. e: = voidage:
,
ap= superficial area, ft 2 / ft 3
3-22
-------
3.2.3
Height of packing
The concept of a transfer unit is useful for interpreting
and correlating mass tans fer data. It is based on the idea ~f
dividing the packed section into a number of conta~ units called
tran&£er units. Although transfer units are, in fact, equilibrium
stage~, the derivation of the design equations differs from that
given for tray packings. In addition to the texts,10-12 the
reader is referred to Reference 24 for a good description of the
design of packed air strippers.
The depth
the height of
packed section
of packing required by one transfer unit is called
a transfer unit (HTU). The total height of the
is given by the relation:
Z = (HTU) (NTU)
(3.30)
where
Z = packed
HTU = Height
NTU = Number
height
of a transfer unit
of transfer units
The NTU characterizes the difficulty of stripping a cqmpound
to a desired level whereas the HTU characterizes the efficiency
of mass trans fer. The usual order of magnitude of HTU is 0.5 to 5
ft. Th~ actual value should be determined fr~m pilot studies for
'the system under consideration, but data sbpplied b~ packing
manufacturers are also useful. In the absence of data, mass
trans fer correlations or models from the literature can be used
for feasibility analyses.
In this section, the equations are presented without
:. derivation. Much of the discussion in Section 3.1 on tray towers
'is relevant, especially with reference to the constraints related
to temperature and flow rates. The design equations given here
are for isothermal towers and constant flow rates.
.
3-23
-------
Height of a Transfer Unit (HTU)
HTU can be estimated in several ways.24-26 Most of the
methods use some kind of mass transfer model. Good experimental
data are required to relate packing performance to mass transfer
models. These models must account for resistance to transfer in
both vapor and liquid phases. To accomplish this, the transfer
height is taken to be made up of the heights required for
transfer through the liquid and gas pha~es:
HTU = (HTU)v+ 5 (HTU)L
(3.31)
(BTU)V and (BTU)L can be calculated using the equations
given in Table 3-5, which were developed by Bolles and Fair.26
Note that the value for HTU is itself a function of the tower
. height, so the height determination is an iterative procedure.
Number of Transfer Units (NTU)
The NTU is defined in terms .of an integral of the feed (xp)
and bottoms (xB) concentrations. This integral equation can be
solved analytically for isothermal operation, constant flow rates
of dilu.te solutions, and solutes obeying Henry's Law. The
resulting equation to determine NTU for packed towers with and
without reflux is:20
NTU = 5/(5 - l)ln[(k(5 - l)/{5(1 - f}} + 1 + k(l - ~)/S]
(3.~4)
The nomenclature is defined in Table 3-2.
. '. .
3-24
-------
where
where
TABLE 3-5
DETERMINATION OF HTU FOR PACKED TOWERS26
(HTU)V=. 'I' dm(Z/10)1/3(Scv)0.5/(3600 G' f\J fp fa)n
(3.32)
(HTU)V = Height of a vapor phase transfer unit, ft
Z a Total packing height, ft
. . .
'I' = Vapor phase packing parameter. Determined from
Figure 3-5
G'= Vapor flux, lb/(hr.ft2)
d = The lesser of tower diameter or 2 ft
m = 1.24 (rings), 1.11 (saddles)
n = 0.6 (rings), 0.5 (saddles)
f\J = (\JL/JJW)0.16
fp - (PL/pw)-1.25
fa = (aL/aW)-O.S .
ScV = \JV/(PG Dv)
.
(HTU)L = ~CfL(Z/10)0.15(scL)0.5
(3.33)
(HTU)L=
~ =
Height of a liquid-phase transfer unit,. ft.
Liquid phase packing parameter. Determine from
Figure 3-5.
\JL/(PL DL)
Vapor-load coefficient. For flood ratios between
0.55 and 0.9, use:
sCL =
CfL=
CfL= 1.75 - 1.4(Flood ratio)
3-25
-------
0.5
-c.- ".-..-
I 1 1111 .-:::
'~I ~ ~~
1.0.;...
'Q,-'
. I.s... I II
I II
0.1
0.01
0.5
.-- "-'"-
I I II - ~~
I II ".s... ~
2.0.;...
I I" ---.,; '.0.;...-
" I
II -0.-.
II
0.1
~
.;
:j
i
~ 0.01
a. 0.5
1
1,
.
~~-l
II
II ......:::::
$ioo.
11' .- I
"'C II -...0.;...
, .' 'I
'1.ti"1
0.1
0.1
1.0.0...
0.01
102
101 10-
Maa wIocicy, I.. , IbIlhllftll
101
PARAMETER FOR (HTU\
2DD
,~
'-
1!D
".s...
100
!D
o
2DD
- "-""-
.,-~ \
-... ".-'
- .o.s... ".......
-'
--" - \,;0.;... - :::;.
-'
c:.-... -
I
,1.Ok ..I.J.ioo.
. '" -
_." -- "".~
'- I
1!D
~ 100
..
I 50
i
1 0
1200
...
~ 1150
100
!SO
o
200
-.... "-
1 !SO
"
-."
,1.-'
100
-'
aD
0.-"
so
o
2D
~
"--'t ftGGd
PARAMElER FOR (HTU)V
Figure
3-5.
. 26
Packing Parameters for Determination of HTU .
3-26
-------
'.
NTU is shown as a function of str.ipping factor for a range
of removal efficiencies in Figure 3-6.3 An important observation
made from this Figure is that for compounds with high Henry's Law
constant and consequently high stripping factor (>50), the number
of transfer units is nearly independent of the magnitude.of
Henry~s Law constant for any given removal efficiency. It seems,
therefore, that for compounds with high Henry's Law constant, NTU
is pr~portional to removal efficiency alone. Note .that the ~a~io
of vapor to liquid flow rate does.not have a significant impact.
on the stripping factor when Henry's Law constant is large.
3.2.4. Pressure Drop
packed towers are usually designed to operate with a gas
pressure drop well below flooding conditions. Most stripping
towers are designed for gas pressure drops of 0.25 to 0.5 inches
water per foot of packed height. In towers picked with rings or
saddles, flooding typically occurs at a pressure drop between 2
and 3 inches per foot of height.27
The total gas pressure drop in a packed tower can be due to
the following resistances:
1.
Pressure drop due to the irrigated pa~king height. The
generalized pressure-drop correlation shown in Figure 3-4
can be used to make estimates if manufacturer's data is not
ava ilable.
2.
Pressure drop for the dry packing height used above the
liquid inlet as an entrainment separator. The pressure drop
of a single fluid flowing through a bed of packed solids,
when it alone fills the voids in the bed, is corr"elated by
the Ergun equation. Approximately, for packed bed
applications, this can be written:
. 3-27
-------
IX
W
W u.
I ~
N 4:
- en IX
.--
8
15 >-
~
::> 0\
~. \
~ ...
~ 10 -b ---
\
\ 0'0-
\
5,- .~
REMOVAL
EFFICIENCY, /'1
8 99.99
8-8
8
o
o 99.9
o
u.
o
IX
W
co
~
z
.
.99
.
A
\
~A- A A A 90
-------
~P = 1.4xlO-9 Cf(G,)2Z/PG
lbf/ ft 2
(3.35)
where G' is in l'b/(hr ft2). Cf is obtained from Table 3-4.
3.
Press ure drop for the packing supports and 1 iquid
distributor, and for inlet expansion and outlet contraction
~osses for the gas. These may be negle~ted in comparison to
~he pressure drop across the irrigated packing.
3.3
ANCILLARY EOUIPMENT
3.3.1. ~ Exchangers
The design parameter of interest is the total heat transfer
surface area which is the primary determinant of cost. In
gen~ral, the surface area required, A, increases with increasing
quantity of heat to be transferred and decreasing temperature
difference; ~T. The basic design equation is:
A = O/(u ~Tm)
(3.36)
This equation is based on the tube outside
it provides a relationship between the three
determipe the heat transfer area required:
surface area and
parameters that
l.
The total amount of heat to be transferred Q is fixed by
process demand and is calculated from s~nsible abd latent
heat requirements.
2.
u is
the overall heat transfer coefficient.
Table 3-6
provides values of U for a variety of applications
encountered in stripping, and may be used for preliminary
estimating purposes.28 These values represent typical values
found in the chemical industry and therefore a range is
given. Values in the lower end of the range should be used
for conservative estimates.
3-29
-------
3.
The temperature difference ~T between the two streams. This
normally varies throughout the length of exchanger, so the
log mean temperature difference is used:
~Tm::l (6T2- ~Tl)/ln(~T2/(~Tl)
( 3 . 37) .
The temperature differences
are those at the inlet and
the log mean temperature
used in calculating the log mean
outlet of the exchanger. Because
difference is based. on ideal
cocurrent or countercurrent flow, i~ is necessary to apply a
correction factor, to allow for flow paths in actual
exchangers. The type of exchanger us~d is normally selected
so that the correction factor is greater than about 0.8. The
reader is referred to Reference 28 for details on exchanger
geometries and. associated correction factors.
3.3.2. Storage Vessels
Storage vessels can be designed from the following factors:
1.
The volumetric flow rate determined
energy balances.
from material
and
2.
The residence time of the fluid in the vessel. This can be
the minimum time for which the fluid has to be
example feed wastewater storage in the case o,f
unit shutdown for emergency re~airs.
stored, for
a stripping
3.
A fixed over capacity factor, usually 20 percent.
The volume of a storage vessel can be calculated as follows:
Vol ume = 1. 2 V t
(gallons ).
(3.38)
where
V= volumetric flow rate, gallons/hr
t = residence time, hours.
3-30
-------
TABLE 3-6
OF THE
TYPICAL MEAN VALUES
OVERALL HEAT-TRANSFER COEFFICIENT,28 U
(Btu)/(OF.ft2.hr)
Shell side
Tube side
Des ign U
--
Liquid-liquid Exchangers
Water
Water
200-250
Demineralized
water
Water
.
300-500
Condensing vapor-liquid
Steam
Feed water
400-1000
Gas-liquid Exchangers
Air, compressed
Water or brine
40-80
Air, atmospheric
Wa'ter or br ine
10-50
Water or brine
Air, compressed
. 20-40
Water or brine
Air, atmospheric
5-20
3-31
-------
SECTION 4
COST ANALYSIS
Presented in this .chapter ar~data and methods. to allow
estimation of capital costs and annualized costs for the various
components of air and steam stripping systems. These costs can be
used .to determine the economi~ viability of stripping, evaluate
alternative waste treatment processes, and for budget planning.
The purpose of the estimate determines the accuracy
required, and in turn how much time and mo~ey is spent on it.
Estimates have been classified and thei.r accuracy standardized,
as listed in Table 4-1, by the American Association of Cost
Engineers. The potential limitations of these cost estim~tes must
be recognized. Capital cost estimates of equipment that may be
derived from the methods in this report are considered to be of
study grade or better.
TABLE 4-1. AACE CLASSIFICATION OF CAPITAL COST ESTIMATES
!YE! of Estimate
Order of.magnitude
Study
preliminary or budget
Definitive or project
Detailed or firm
Accuracy, +%
control
40
25
12
6
3
4-1
-------
Information on process equipment costs has been published in
. .
various engineering books, journals' and several EPA reports. The
cost data and methods presented in ~is report have been derived
from these sources, and not from vendor quotes or case histories.
4.1
CAPITAL COSTS
Capital costs are the costs of the equipment used. They are
discussed in terms of purchased cost, delivered cost and
installed cost. Pikulik and Dias29 sUg'gest that the delivered
cost of equipment may be approximated by increasing the purchased
cost of equipment, f.o.1:). manufacturer's .sho.p, by 3 percent. An
installation factor, usually different for each type of
equipment, can be utilized to determine the installed capital
eost. Installation factors are usually based on the purchased
equipment cos t.
The most accurate method of determining equipment costs is
to obtain firm bids from manufacturers or s upp 1 iers. 0 ften
manufacturers can supply quick estimates whichwil.l fall close to
the bid price. Second bes t in re 1 iabi 1 i ty are cos ts from the fi 1 e
of past purchase orders. When used for costing new equipment,
purchas e-order cos ts mus t be corrected to the current cos t
indexes.
It is often necessary to estimate the cost of a piece of
equipment when no cost data are available for the particular size
required. The. logarithmic relationship known as the six-tenths-
factor rule can be used if the new piece of equipment is similar
to one of another capacity for which cost data are available.l~
Costl = Costo (Sizet/Sizeo)0.6
(4.1)
However, the application of this rule of thumb fo~ most purchased
equipment is an oversimplification since actual values of the
cost capacity factor vary from less than 0.2 to..great~r..than 1.0.
4-2
-------
"
Typical exponents are presented in this report for major
stripping components. In general, the cost-capacity concept
should not be used beyond a tenfold range of capacity. Care must
be taken to make certain the two pieces of equipment are similar
with regard to type of construction, material of construction,
and temperature and pressure operating range.
Purchased equipment costs can often be estimated on the
bas is of weight. The fact that.a wide variety of types of
equipment have about the same cost per unit weight is quite
useful, particularly when other cost data are not available. For
, . '
fabricated vessels, such as stripping towers, fairly accurate,
costs can be derived on the' basis of their weights. The
procedures for translating the process design data from the
previous Section into vessel weight equivalents are presented
here.
The capital cost of air/steam stripping systems
grouped into costs for the following major components,
, '
which is discussed separately:
can be
each 0 f
1.
Mass transfer equipment (tray and packed towers)
2.
Heat transfer equipment (heat exchangers, condensers and
reboilers)
3.
Fluid transfer and handling equipment (pumps, compressors
and tanks)
4.
Installation materials including foundation, structural,
instrumentation and controls, paint, insul,ation, electrical
and piping, as well as labor - estimated by means of factors
related to the purchased equipment costs.
All the capital cost data presented here are based on May
1983 cost indexes unless otherwise stated'--
4-3
-------
4.1.1.
Mass Transfer Equipment
The purchased cost for tray and packed towers can be
divided into the following components:
1.
Shell cost, including heads, skirts, manholes and nozzles,
2.
Cost for internals, including
packing, supports, and plates, and
trays
and
accessories,
3.
Cost for auxiliaries,
handrails, and insulation.
such
as
platforms,
ladders,
The published methods for determining these costs are
largely graphical. However, with the advent of computer aided
design and cost analysis, the use of correlations is becoming
more convenient.
Hall et al30 present purchased costs of tray type and packed
towers in a series of graphs, as a function of diameter, height
and specific operating parameters. The probable accuracies of
these costs are in the range of !lOto 15 percent.
A method for determining installed costs of plate type
distillation towers, that includes items such as bulk material
and field labor, overhead, engineering and contingencies, is.
presented by Miller and Kapella3l. The base. cost of 'the colum~ is
presented graphically as a function of the product of diameter,
he igh t and th ic kness. These cos ts represent Un i on Ca rb ide's
actua 1 purchas ing exper ience and can prov ide a reasonab 1 y
accurate estimate (within 25 percent) for a completely installed
tower.
The costs for essential components of a stripping tower
provided in this report are in terms of correlations that were
developed by Corripio ~ .!!32 through the use .of computer-aided
4-4
-------
statistical analysis of cost data for 200 distillation towers and
200 adsorp~ion towers. In these correlations, tower cost
represents the sum of the costs of the shell, platforms and
ladders, and internals, either trays or packing. The costs of the
shell are calculated from a knowledge of the shell weight.
The shell cost data on which the correlations are based
include the cost of the skirt and a standard number of nozzles
and manho 1 es. Thes e are functions_of tower diameter, 1 ength and
pressure rating. The shell weight is calculated (assuming 2:1
elliptic heads and ignoring the nozzles, manholes and skirt) from
tower diameter, tangent-to-tangent l~ngth and design pressure
(external or internal) by the procedure us ed for des igning
pressure vessels.33 This procedure, as outlined below, provid~s a
set of analytical equations and takes into account wind load
effects and allows for different shell thickness at the bottom
and top of the tower.
Shell weight
Shell weight depends on material density, diameter, tangent-
to-tangent length and wall thickness. Thickness is a function of
des ign press ure, diameter, 1 ength, and either the tens i 1 e
strength or the modulus of elasticity of the material of
cons truction (depending on whether the des ign press ure is
pos i ti ve or negative). For towers, the thickness requirement to
withstand wind resistance must also be taken into consideration.
The equation to calculate shell. weight is:
w =
s
Di(L + O.SlDi) TsPs
(4.2)
where
Ws = Shell weight, lb
Di = Inside tower diameter, ft
L = Tangent-to-tangent length, ft
Ts = Tower wall thickness, ft, and
P = Density of shell material, Ib/ft3' ..
s
4-5
-------
!.!1.! thickness for internal press ure. I f the press ure in
the tower is positive, the stress on the longitudinal seam
governs. The maximum allowable stress S, which is a property of
the specified material of construction, is then used in the
following equation to calculate the wall thickness:
Tp - Pg(Di/2)/(SE - O.~ Pg)
(4.3)
where
T -
P
Pg
S
E
Thickness to withstand design pressure, ft
= Gage pressure, psig
- Maximum allowable stress, psi, and
- Joint efficiency
Criteria for the selection of materials of construction are
discussed later in this section. The wall thickness is one
criteria in selecting the grade or quality of material. For
example, in the case of carbon steel, the lower grade SA-285 C
can be used for wall thickness up to 2 inches, and the higher
grade SA-5l5-65 can be used for wall thickness greater than 2
inches.
The welding joint efficiency, E, is determined from tests in
, ,
accordance with ASME codes. For carbon steel up to 1.25 inches in
thickness, the weld need only be 10 percent spot-checked by X-
rays, and a joint efficiency of 0.85 must be used in the formula.
For any other thickness and for other materials of ',construction,
the weld must be 100 percent X-ray tested and the joint
efficiency is 1.0.
Wall thickness for external pressure. An iterative graphical
procedure to calculate thickness under vacuum conditions is
outlined in several sources. A method suitable for computer
calculation is given here. The thickness required to withstand
external pressure, Tp' is given by:
Tp = Te + (Te)c
'., (4.4)
-,\
4-6
-------
Te must be high enough for the collapsin~ pressure, Pc' given by
Eq. (4.5) to be five times the difference between the external
(atmospheric) pressure and the design (vacuum) pressure in the
tower.
Pc ~ (2.6(Te/Do)2.5EMJ/(L/Do - 0.4S(Te/Do)0.5J
(4.5)
where'
Do a Outside tower diameter
EM = Modulus of elasticit¥
The correction factor (T e) c is obtained from the. fo 11 owing
equation, with Te' Land Di all in. inches:
(Te)c = L(O.lS Di - 2.l7)xlO-5- 0.19
(4.6)
Thickness ~ wind load. The required wall thickness of a
tower normally varies from top to bottom. The thickness at the
top is that ~equired to withstand either the in~ernal or .external
pressure. The thickness at the bottom is additionally required to
withstand the wind load, which is assumed to be sufficient also
for any earthquake load. Assuming that the wind acts with a
uniform intensity. over the entire length of the column and that
the drag coefficient for the wind resistance is 1.0 (drag past a
cylinder in turbulent flow), the thickness necessary to withstand
the wind load is calculated from:
T = P V2(D + Z)L2/(S~D2)
w a 0 0
(4.7)
After substituting common values. for some variables such as:
Pa. ~ Air density, 0.075
V = Wind velocity, 140
Z = Allowance for cage
lb/ft3 at,70oF and
miles/hour, and
ladders, 1.5 ft.
1 atm~
we get the following simplified equation with Tw' Do' and Lin
inches, and S in ps i:
4-7
-------
Tw = 0.22(Do + 18)L2/(SD~)
(4.8)
The thickness required to withstand the internal pressure
when the girth seam governs is calculated by:
Tg = Pg(Di/2)/(2SE + 0.4 Pg)
(4.9)
The total thickness at the bottom of a tower is then given by:
Tb =. Tw + Tg
(4.10)
Finally, the total shell thickn~ss
greater of the calculated to~ and bottom
corrosion allowance, Tc, which is specified
is taken to
thicknesses,
by the user.
be the
plus a
Ts = (greater of Tb' Tp) + Tc
(4.11)
This final calculated thickness should be rounded up to the
nearest standard plate thickness.
Shell cost
The shell cost or the base cost of the tower, which covers
fabrication and prime painting in the shop, is f.o.b.
manufacturer's plant. Correlations for base cost, in carbon
steel, Gf the shell are given for both tray towers and packed
towers in Table 4-2. To calculate shell cost for :.a material of
construction other than carbon steel, use thematerial-of-
con s t r u c t ion f act 0 r s , F M ' a 1s 0 9 i v e n in. Tab 1 e 4 '- 2 ,
as a mul tip I ier to the base cost in c.arbon steel.
To fabricate a shell of thickness varying .from top to
bottom, the. additional labor cost required is significant only
for towers having large length-to-diameter ratios, as these must
be thicker at the bottom to withstand wind loading. This is
evident in Eq. (4.13) for towers taller than 40 ft~ The
4-8
-------
TABLE 4-2
CORRELATIONS FOR COST* OF STRIPPING TOWERS32
Carbon steel shell, base cost £BL !
Ce= 1.3exp[6.3 + 0.18 lnws + 0.02 In(ws~2]
for
Lt ~ 40 ft, and
4250 lb ~ ws~ .980,000 lb
Ce= 1.3exp[6.8 + 0.14 1nWs + 0.02 (lnWs) 2 +
+ 0.016(L/Di)ln(Tb/Tp)]
for
Lt > 40 ft .
9020 lb ~Ws ~. 2,4'70,000
Final shell cost£S~ !
Cs= FM Cg
where
Material
Carbon steel
Stainless steel, 304
Stainless steel, 316
Carpenter 20CB-3
Nickel - 200
Monel": 400
Inconel 600
Incoloy - 825
Titanium
Cost factor,' FM
1',0
L7
2.1
3.2
5.4
3.6
3.9
3.7
7.7
(4.12)
(4.13)
(4.14)
* .
Costs were escalated to May 1983 using the Che~ical Engineering
Fabricated Equipment Index of 327.1
, 4-9
-------
requirement is smaller, however, ,for tall towers of higher design
pressure because of the greater thickness at the top of such
towers. The ratios of tower length to diameter and bottom to top
thickness in the cost correlations accounts for this additional
labor cost requirement. Eqs. (4.12) and (4.13) are each based on
about 200 data points on the cost of distillation and adsorption
towers covering a wide range of design variables and materials of
construction. The standard deviation is 9.9 percent for Eq.
(4.12) and 10.6 percent for Eq. (4.13).
platforms and Ladders
Correlations for the base cost, CL' for carbon steel
platforms and ladders are given in Table 4-3. The standard
deviation for Eq. (4.15) is 8.9 pe~cent and that for Eq. (4.16)
is 3.4 percent. The significant. difference between the two
standard deviations reflect the,'effect of correlating the
discretely varying number (and cost) of platforms with a
continuous variable, tower length.Tbe effect of this error is
much larger for the shorter towers than for the taller towers.
Cost of Internals
~ of trays, CT. Correlations for the ba~e cost of trays
were developed from cost data for Glitsch 8Truss type8 one-pass
removable ball~st trays. The base cost of,valve t~ays in ca~bon
steel are given as functions of tower diameter in Table 4-4. For
the correlation of 14 trays of different diameter, the standard
deviation is 1.3 percent.
Material-of-construction cost factors, FMT' and tray-type
factors, FT' that must be applied to the base cost are given in
. ,
Table 4-4'. If a design calls for: fe~erthan 20 trays, a number-
of-trays factor, FH' recommended by, £'nyedy34 must be appl ied: '
FN = 2.25/(1.0414}N
. .. ..'
(4.l9)
4-10
-------
TABLE 4-3
CORRELATIONS FOR COST OF PLATFORMS AND LADDERS32
CL- 237 D~.74 LO.71
for
27 ft < L < 40 ft
3 ft < Di ~ 21 ft
C = 198 DO.6 LO.8
L 1.
for 57 ft < L < 170 ft
3 ft < D. < 24 ft
1.. ,-
(4.15)
(4.16)
TABLE 4-4. CORRELATIONS FOR COST OF TOWER TRAYS32
Carbon steel valve tray base ~ £ST~ !
CST= 362 exp(0.174
Do>
for
2 ft < Do ~ 16 ft
Final cost
CT = CST FMT FT
Material
Carbon steel
Cost Factor,: FMT
1.0 ..
1.19 + 0.06Di
1040 + 0.07.Di
1. 53 + 0.08 Di
2.31 + 0.11 Di
Stainless steel, 304
Stainless steel, 316
Carpenter 20 CS-3
Monel
Tray ~
Valve
Grid
Subble cap
Sieve. (with
Cost Factor, FT
1.0
0.8
1.59
0.85
d~wncomer>
4-11
(4.17)
(4.18)
-------
~ ~ tower packings, Cpo Cost of tower packings are based
on d~ta from pikulik and Diaz29. The costS are estimated from the
required volume of packing and its cost per unit volume as listed
in Table 4-5. The cost of a distributor plate in a packed tower
can be assumed to be the same as tha~ for one bubble cap tray.
TABLE 4-5. COST OF TOWER PACKING PER UNIT VOLUME29
packing ~
Ceramic Raschig rings or
Intalox saddles, 1 in.
£P~ $/fd
18.9
Ceramic Raschig rings or
Intalox saddles, 2 in.
13.2
Metal Raschig or Pall rings, 1 in.
31.1
Metal Raschig or Pall rings, 2 in.
22".1
Total Purchased Cost
Total cost of t.cay towers is the sum of the costs for
individual components as discussed in previous sections. If N is
the number of trays required, then the total cost can be
estimated using the following equation: .,
C = Cs + CL + N CT
(4.20)
Similarly, the total cost of a tower packed to height Z is:
C = Cs + CL + (~D~/4)Z Cp
(4.21)
Correlations for the cost of towers having two diameters
have been published by Enyedy.34
4-12
-------
4.1.2
Heat transfer equipment
The basic capital cost for shell-and-tube heat transfer
equipment - which incl udes heat exchanc;ers, condens ers and
reboilers - is a function of many desic;n and cost parameters. The
most jmportant parameters are exchanc;er type (fixed tubesheet, U-
tube, floatinc; head and kettle-type), heat transfer surface area,
diameter, 1 enc;ths and c;auc;e of the tubes, des'ic;n press ure,
material of construction, and the -type and inlet temperature of
the fluids. The fixed tubesheet desic;n is low in cost and
cleanable on the tubes ide. It should not be used in applications
involvinc; severe thermal expansion stresses and severe foulinc;.
The U-tube desic;n is moderate in cost and eliminates the problem
of thermal expansion. However, they impose difficulty in tubes ide
cleaninc;. The floatinc; head desic;n is the hic;hest in cost, but
both the tubes and the sheli are easily cleaned. If low
temperatures are required for condensation, refric;eration can add
sic;nificantly to the total capital and operatinc; costs of a heat
trans fer s ys tem. 40
Costs of ~hell-and-tube heat exchanc;ers are correlated
ac;ainst heat transfer area and are presented by Corripio et.al.4l
The base cost is for floatinc;-head.shell-and-tube exchanc;ers made
of carbon steel and desic;ned for 100 psic; pressure. Cost. factors
. ..
for desic;n type, desic;n pressure and material-of-construction are
provided separately and must be used to adjust the base cost.
Table 4-6 lists correlations, factors and limits for their use.
The cost estimation procedure that takes into account shell
diameter, nutnber, lenc;th and c;auc;e of tubes, types of heads and
other construction details can be more accurate and is discussed
in several References .14,40 The accuracy of the correlation of
cost vs. area provided here is adequate for study cost estimates.
, 4-13
-------
TABLE 4-6
CORRELATIONS FOR COSTS OF HEAT TRANSFER EQUIPMENT4l
carbon steel, 100 psig design
Sase cost for floating-head,
*
pressure:
Cs ~ 1.3exp[8.55 - 0.3l(lnA) + 0.07(ln A)2]
Final cost:
CE = Cs FT Fp FM
!Y2!
Fixed tube-sheet
U-tube
Kettle reboiler
Factor FT
exp[-1.12 + 0.091
exp[-0~98 + 0.083
1.35
ln A]
In A]
Pressure
100 to 300 psig
300 to 600 psig
600 to 900 psig
Factor Fp
0.78 + 0.05 ln ~
1.03 + 0.07 ln A
1.14 + 0.12 ln A
Material
Carbon steel
Stainless Steel, 316
Stainless Steel, 304
Stainless Steel, 347
Nickel-200
Monel-400
Inconel-600
Incoloy-825
Titanium
Hastelloy
Factor FM
1.0
0.86 + 0.23
0.82 + 0.16
0.61 + 0.22
1.51 + O. 61
1 . 3'0 + O. 43
1. 20 + 0.51
1.19 + 0.50
1.54 + 0.43
0.15 + 1.52
* A in ft2; 150 < A < 12,000
4-14
In A'
ln A'
ln A
In A -,
ln A
ln A
ln A
In A
ln A
.. .. ..'
(4.22 )
(4.23)
-------
4.1.3
Centrifugal Pumps
pumps used for transferring aqueous dilute solutions are
commonly centrifugal-type and the easiest way to estimate thei r
cost is by using a cost-data manual which all pump manufacturers
provlde to their customers. These manuals are detailed and give
relia~le estimates. For designs not covered in the manual, the
usual procedure is to send detailed specificati~ns to several
pump manufacturers for their bids~
Correlating the cost of centr.ifugalpumps against size or
capacity is difficult bec~use a pump of a given capacity can
serve in a variety of combinations of flowrate (capacity) and
developed head. This di ff icul ty is circumvented by corre 1 a ting
pump cost against the maximum value of the size parameter 5,
defined below, that could be handled by a pump of a particular
cost.
5 - Q HO.5
Here
Q is design capacity in gpm, and
H is the required head in ft-lb/lb.
The base cost is for a one-stage, 355Q rpm, vertically-
spl it-case (V5C) pump of cast iron. Correlations for 'pump base
cost and design-type factors, and cost factors for material-of-
con'struction are given in T'able 4-7.42 The cost of a pump
calculated from these correlations includes the base plate and
driver coupling but not the driver. The capacity, head and
horsepower limits for each type of pump are noted in Table 4-8.
.4-15
-------
TABLE 4-7. CORRELATIONS FOR COST* OF CENTRIFUGAL PUMPS42
,"
Base Cost CB for single-stage, 3550 rpm, cast iron, V5C:
Cs = 1.53exp[8.4 - 0.6(ln S) + 0.05(ln S)2]
Total cost:
(4.24)
Cp .. CB FT FM
( 4 . 25 )
!lE!
I-stage 3550 rpm, VSC
I-stage 1750 rpm, VSC
I-stage 3550 rpm, SSC
I-stage 1750 rpm, SSC
2-stage 3550 rpm, HSC
Multistage 3550 rpm, SSC
Factor FT
1.0
exp[5.10 - 1.22 In 5 + 0.077(ln S)2]
exp[0.06 + 0.27 1n S - 0.025(ln 5)2]
exp[2.03 - 0.24 In S + 0.010(ln S)2]
exp[13.73 - 2.831n S + 0.154(ln 5)2]
'exp[9.88 - 1.62 1n 5 + Q.083(ln S)2]
Factor FM26
1.0
1.35
1.15
2.00
2.00
3.50
3.30
4.95
4.60
9.70
2.95
1.15
1.90
Material
Carbon steel
Cast steel
304 or 316 fittings
Stainless ~teel, 304 or 316
Cast Gould's Alloy No. 20
Nickel
Monel
ISO B
I50 C
Titanium
Haste110y C
Ductile iron
Bronze
*Costs were escalated to May 1983 using Chemical Engineers's
Pumps and Compressors Index of 413.4. . ,...
4-16
-------
TABLE 4-8
CAPACITY, HEAD AND HORSEPOWER LIMITS FOR CENTRIFUGAL PUMPS 42
Flow Head Mo to r
- !.l:E! s.E!!! ft-lbf/lb hE (max)
1-stage 3550 rpm, VS 50 to 900 50 to 400 75
I-stage 1750 rpm, VSC 50 to 3; 500 50 to 200 200
I-stage 3550 rpm, HSC 100 to 1,500 100 to 450 150
1-stage 1750 rpm, HSC 250 to 5,0.00 50 to 500 250
1-stage 3550 rpm, HSC 50 to 1,100 300 to 1,100 250
Mul tistage 3550 rpm, HSC 100 to 1,500 650 to 3,200 1,450
4.1.4
Electric motors
To determine the cost of a driver for a.pump, the required
brake horsepower is determined using the following formula:
PB =
o HI (33 ,000 Ep) .
where,
p = fluid density, lb/ga1
o = liquid flow, gpm
H = developed head, ft-lbf/lb
Pump efficiency Ep may be calculated from:
Ep = -0.316 + 0.24(ln Q) - 0.012(ln 0)2
19 < 0, gpm ~ 5000
The cost correlations and coefficients for motors are given in
Table 4-9 and are for three motor types and three speeds. Because
electric motors come in discrete sizes, the motor size to use
with Table 4-9 is the available motor horsepower size that is
. -.. of ,-
just equal to or bigger than the required brake horsepower.
4-17
-------
TABLE 4-9
CORRELATIONS FOR COST* OF ELECTRIC MOTORS42
Cm a 1.38 exp[al + a2(ln P) + a3(ln p)2]
(4.26)
P a nominal power of motor, hp (~ brake horsepower Pa)
60 Hz, standard voltage motor & insulation
.!l
Open, drip-proof
3,600 rpm
4.83
4.15
4.24
1,800 rpm
4.71
4.52
7.40
1,200 rpm
4.93
5.10
4.62
Totally enclosed,
3,600 rpm
fan-cooled
5.11
3.85
5.32
1,800 rpm
4.97
4.53
1 , 200 rpm
5.15
5.39
Explosion-proof
3,600 rpm
5.39
4.44
1,800 rpm
5.29
4.82
1,200 rpm
5.42
5.57
.!2
0.097
0.535
1. 033
-0.015
0.472.
-0.065
0.301
0.359
0.885
0.033
0.833
1. 085
-0.009
0.571
0.289
0.310
-0.003
0.608
o
0.511
0.312
0.313
.!3
0.11
0.053
~ -0 . 036
0.229
0.048
0.054
0.126
0.061
~0.022
. 0.154
0.024
~O.057
0.226
0.046
0.144
0.074
6.155
0.052
0.200
O.~ 053
.0.106 .
...0..072
!!E limit
1-7.5
7..5-250
250-700
1-7.5
7.5-250
250-600
1-7.5
7.5-250
250~500
1-7.5
7.5-250
250-400
1-7.5
7.5-250
1-7.5
7.5-350 .
. ,
1-7.5
7.5-200-
1-7.5
7.5-250
1-7.5
7.5-200
"
Costs were escalated to May 1983 '~sln~ Ch~mical
Electrical Equipment Index of 242.4
4-18
Engineer's
~ " h
-------
4.1.5
Air compressors
Air stripping requires a compressed air supply. Compressors
having high capacity and low discharge pressures are required.
Cent~ifugal air compressors will adequately do the job. Their
cost, Cc' can be determined as a function of capacity using the
correiation fitted to data from Peters and Timmerhaus.13
Cc - 1.53xexp[0.31(ln 0) + 8.52]
( 4 . 27 )
o :8 capaci ty, ft3/min; 50 ~ 0 ~ 1000'
Typical speeds and horsepower for various capacities are:
,q
143
275
325
rpm
230
275
300
HP
50
75
100
The costs were escalated to May 1983 using Chemical Engineering's
Pumps and Compressors Index of 413.4.
4.1.6
Storage vessels
Preliminary-grade estimates of storage vessels can be made
using the correlations and factors for material-of-construction
given in Table 4-10. Total vessel or tank volume is calculated
from ,the knowledge of tank residence time, a fixed overcapacity
factor of 20 percent, and volumetric flow ~ate. The last is
determined from material and energy balance calculations. Omitted
, in the cost estimating procedure are the number and sizes of
manholes and nozzles and other design details.
4-19
-------
TABLE.4-10
CORRELATION FOR COSTS OF STORAGE VESSELS41
Base cost for shop-fabricated carbon steel tanks:
CB = 1.3 exp[2.33 + l.37(ln V) - 0.063(ln V)2j
1,300 < V gallons < 21,000
- , -
Base cost for field-erected carbon steel tanks:
CB = 1.3 exp[ll.4 - 0.6l0(ln V) + 0.045(ln V)2]
21,000 ~ V gallons ~ 11,000,000
Final Cost:
Cv = CB FM
Material
Carbon steel
Stainless steel 316
Stainless steel 304
Stainless steel 347
Nickel
Monel
Factor FM
. 1.0
2.7
2.4
3.0
3.5
3'~ 3
3.8
11.0
ll~O
Inconel .
Zirconium
Titanium
Brick-and-rubber or
brick-and polyester lined steel
Rubber- or lead-lined steel
Polyester, fiberglass-reinforced
.Aluminum
2.75
1.90
0.32
2.70
2.30
0.5$ '..
Copper
Concrete
4-20
(4. 28 )
(4.29)
(4.30)
-------
Judgment on optimum allowance for corrosion and on the use
of the most suitable construction materials needs experience or
vendor assistance. Guidance can be obtained by reference to a
materials handbook.47
~t is customary to have spare pumps and motors available. In
larger plants, standby pumps may be included in the process flow
sheet on by-pass lines. For a stripper facility, consider having
back-ups for all pumps and motors ~s well as the air compressor.
These must be included in the capital cost estimate.
4.1.7. Installation Costs
The installation of equipment involves costs for labor,
foundations, supports, platforms, construction-expenses, and
other factors directly related to the erection of purchased
equipment. The following list of additional items/materials must
be included in estimating the total cost of installation
materials:
Equipment insulation
Instrumentation and controls-
Piping and insulation
Contractors fees
" "
Contingencies
Service facilities
Engineering, Supervision and
Electrical installations
Buildings
yard improvement
Land
Start-up
Construction
Capital costs of equipment;, system or plant, that include
costs for all of the above items are often referred to as the
total invest~ent cost. From experience one can _generate factors,
as a percent of purchased equipment cost (or some other
conven.ient base cost), for each of the above categories to
es"timate the investment cost or .turnkey. cost for a system. We
have not found factors for air and steam stripping systems in
- the 1 i tera ture.
4-21
-------
There are- several methods available to estimate the total
investment cost of a system when installation costs are not
available. These include the methods of Lang,35 Hand,36 Guthrie37
and viola.38 In Lang's method the total investment cost ~f a
stripping system is estimated by mul tiplyingthe total del ivered
cost of equipment by a factor of 4.74. As discussed above, the
delivered cost of equipment may be approximated by increasing the
purchased cost, f.o.b. manufacturer's shop, by 3 percent.
In Hand's method,36 different installation factors are used
for each type of equipment. Some typical factors39 based on the,
purchased equipment cost are:
Equipment
Towers
Heat exchangers
Compressors
Factor
4
3.5
2.5
The purchased equipment cost must be multiplied by these
factors to arrive at total installed costs, and the sum of these
products represents the estimated inside-battery-limit cost of
the installed plant. While these meth~ds are convenient to use
they provide, at best, study estimates.
4.2
Operating Costs
. ,
De t e r m i n a t ion' 0 f c a pi tal co s t s i son 1 yon e par t 0 f a
complete cost estimate. Another equally importan~ part is "the
estimation of costs for operating the system and maintaining - it
to run efficiently. Operating costs include both fixed and
variable components.
Total operating costs are commonly calculated on one of
three bases: daily basis, unit-of-product basis, or annual basis.
The annual cost basis is probably the best choice because:
4-22
-------
1.
2.
The' effect of seasonal variation is smoothed o~t, and
The pl an t stream facto r is inc 1 uded.
Fixed operating costs include labor, supervision, overhead,
laboratory labor, maintenance, services, insurance and taxes,
serv~ce water, and the cost of capital. Variable components
include utilities - efectricity, steam, cooling water, and
compressed air.
4.2.1
Fixed Costs
The bases and. factors for fixed operating costs are given in
Table 4-11 obtained from the EPA cost model. Cost factors in the
original are for July 1977; values adjusted with appropriate cost
indexes to May, 1983 are also shown in the Table.
4. 2 ~,2
Variable Costs
Power Requirements
These include power for the feed,
overhead 1 iquid pumps, and for the air
horsepower is estimated using:
bottoms, reflux, and
compressor. The pump
HP = 1.21xlO-6 x Q x H
(4. 28 )"
where
Q is the flow rate, Ib/hr, and
H is the head, ft.
A 20 percent factor for flow variation and a 50 percent pump
efficiency is assumed. The following heads may be used:
Feed-pump head
Bottoms-pump head
Overhead and reflux pumps
L is the total tower height
50 + L
50 + L/2
20 ft
in feet.
ft
ft
of 0.""
4-23
, .
-------
Element
Labor*+
.. *
Supervlslon
*
Overhead
**
Laboratory
Maintenance
Services
Insurance
and Taxes
Service Water
Amortization
TABLE 4-11
FIXED OPERATING AND MAINTENANCE COST BASIS
AND UNIT COST FACTORS FOR STRIPPING8
Cost Basis
Equivalent Unit Quantity
. - July 1977
Base Unit Cost
May .1983
0.25Weeks (6hr/day)
10' Labor (0.60 hr/day)
75' Labor Cost
$ .9.80/hr
$11.76/hr
NA
$10.70/hr
NA
NA
0.20 Shifts (1.6
4.13' Capital
0.40' Capital
hr/day)
2.50' Capital
103 gal/day
16' Capital
NA
$0.50/103
(estimated) NA
gal
$14.70
$23.25
NA
$16.00
NA
NA
NA
$1.00/103
NA
gal
NA - not applicable.
* .
Labor may vary from 0.7 to 1.2 times the standard amount
indicated depending on the overall scale6f the plant.
. .
.. "
Labor, Supervision, and Overhead may also be adjusted for
the scale of the plant.
+ One week = 7 days = 168 hours = 4.2 shifts
**
One shift = 40 hours
Note: The May, 1983 data were estimated using Chemical
Engineering's indexes for construction labor.
4-.24
.. .. ..'
-------
The horsepower for' the compressor can be calculated from the
equation for an ideal gas undergoing isothermal compression:14
HP - 3.03xlO-S Pl In(P2/Pl)
(4.29)
where,
Q . air flow rate in ft3/m
Pl' P2 .- intake and delivery pressure, lbf/ft2
Equations for isentropic compression should be' used for high
compression ratios.
The total annual power cost is given by:
Cp a 6535 SF P ($/kWhr)
(4.31)
where SF is the stream factor and P is the sum of the horsepower
. requirements for all the pumps and compressors.
Utilities
The cooling water requirement is approximately equivalent to
the latent heat of the steam and organics condensed in the
condenser. Assuming a 200Ftemperature rise and a 20 percent
factor for flow variations, the cooling water requirement is:
CW = 63 SF (condenser duty, Btu/hr)9als/year
(4.32)
The steam rate is the design stripping medium rate plus any
requirements for heating the feed and compensating for heat
. losses in the tower.
Costs for steam and cooling water are normally obtained from
plant records. If data are not available, steam costs can be
based on an energy cost of $3/106 Btu, and a cooling water cost
based on 3 to 5 cents/1000 gallons water circulated.
4-25
-------
SECTION 5
SUMMARY OF DESIGN PROCEDURES
,ep by step approach suitable for handling most design
s i tUi t'1S is gi ven here to ass is t the reader through the
numerous correlations and situati~~s:covered in the previous
sections. An important part of obtaining a reliable design is
having accurate physical and chemical property data available.
Henry's Law constants and activity coefficients for the organics
on the EPA's toxic pollutant list are included in the Appendix.
These should be checked and updated with recent data as it
becomes available. Lists of densities, viscosities, specific
. ,
heats, latent heats and surface tensions for a range .of species
can be found in many handbooks.28,43,44 Diffusivitie~ ~re less
easily found, but may be estimated using correlations and data
given in Reference 17.
5.1
PROCESS DESIGN
1.
Basic Design Data
The basic design data req~ired includes the flow rate and
temperature and pressure of the stream to be stripped.
Temperature and pressure constraints due to chemicals in the
. - .
stream or imposed by the process should be noted.- 1".detailed
stream composition is useful, particularly with r~~pect to
suspended matter and reactive or corrosive substances... Treatments
to be provided upstream of the stripper should be identified.
5-1
-------
Henry's Law constants for all the components should be
obtained, preferably as a function of temperature as given in the
Appendix. In the case of multicomponent streams, select a key
component on which to base the design. Molecular weights will
be required to convert the concentrations to mole fractions.
Other physical property data that will be required are listed
later.
2.
Preliminary Tower Height Calculations
The design options available can be relatively quickly
ascertdined by doing a few equilibrium s~age. calculations Using
the s impl ified equations in Table 3-2. Determine the number of
stages when stripping with steam at .from 10 to 30 mole percent of
the feed, and with air at from one-half to twice or more the feed
rate (molar basis). Examine the effect of vari6us amounts of
reflux in the case of steam.
3.
Selection of Stripping Medium
The preliminary designs will give an indic~tion of the
effectiveness of stripping and should also sugge~t whether air
stripping is satisfactory or whether steam should be used. Steam
will usually be less expensive if it means an a-fold decrease in
the amount of stripping medium used.3 Steam may also be dictated
for process reasons related to recovery of th~ organics and/or
disposal of the overhead vapors. If the choice is not obvious,
complete cost estimates for both steam and air, and select the
less expensive system.
Li ve and reboi 1 s team. In the cas e of s team, a chqice mus t be
made between live steam or the use of a reboiler.. Areboiler adds
to the capital costs, but the steam is recovered as a clean
condensate and is recycled. An equivalent am6u~t~f vapor is
driven off the feed in the overhead vapors, so the bottoms rate
is less than the feed rate. Live steam does not significintly
. -0. ...
5-2
-------
alter the feed stream rate. The live steam must be continually
made up from the boiler feedwater treatment plant. In some
situations, it may be possible to use a low purity steam raised
from was tewa ter.
conde.ns er and ref 1 ux. The incorporation of a condens er in the
system may also influence the decision between reboil and live
steam" A condenser will be used wherever recovery o~ the organics
is required. A partial condenser m~y be used to concentrate the
overhead in systems that incinerate stripped vapors. The amount
of reflux is determined by the quality.o~ overhead required. When
a decanter is used, all the aqueous phase may be returned to the
tower as refl ux.
4.
Selection of Tower ~
Select between a packed or tray tower on the basis of the
information given in Tables 2-1 and 2-2.' This information
together with the data in Table 3-4 and the discussion in Section
.3.2.1 will assist in the selection of the type of packing or
tray. Availability and costs should be ascertained by consulting
the manufacturers.
5.
Detailed Process Design
Sufficient background information is now available to carry
out the process design. Tower capital costs decrease with
increasing stripping medium, but operating costs increase. An
optimum design having minimum overall costs can usually be found.
It is therefore recommended that the designs be repeated for a
range of stripping medium rates.
Tower height. For steam, use the equations in Table 3-2 for tray
towers, and Table 3-5 together with Eqs. (3.31), (3.32) & (3.34)
for packed towers. In the case of (adiabatic) air stripping, the
'. . '. ...
number of equilibrium stages can be obtained using the BASIC
5-3
-------
program given in the Appendix.
provided for packed towers, the
but with some loss of accuracy.
Because a similar program is not
isothermal equations may be used,"
The equation for the determination of HTU incorporates
factors containing the surface tension, viscosity and density of
the liquid relative to water. Density data is usually readily
available, but surface tension and viscosity data may be less
easily found. Although these latter variables can vary
" .
significantly with even small concentrat~ons of organics, the HTU
is not strongly dependent on them (they are rais ed to low
powers). Neglecting these factors normally will not introduce
large errors, especially when designing systems in which the
organic is stripped down to trace amounts. The diffusion
. .
coefficient of the organic through the liquid is always required,
and may be estimated using published correlations.17
Note that the determination of HTU and tower height for.
packed towers is an iterative procedure. The use of a
programmable calculator or microcomputer is recommended.
Tower diameter. Liquid and .vapor dens i ties are required
addition, the liquid surface tension for tray towers,
liquid viscosity for packed towers, shoul~ be known.
and, in
or the
For tray columns, use Eqs. (3.26) to (3..28) and the
parameters in Table 3-3. Manufacturer supplied d~~a should~be
used if available. The tower diameter also sets the tray spacing
(See Table 3-3), which, when multiplied by the numb~r of stages,
is the height required for the tr~y section of the tower.
Additional height may be required for a reboiler, and an extra
static head or 'skirt' height (typically about 15 ft) may be
added to preve~t cavitation in the bottoms pump.
For packed towers, the tower diameter is determined from
Figure 3-4 as described in Section 3.2.2. If packing parameters
.. . ~ ..'
5-4
-------
are not available from the manufacturer~,
use the data in Table
3-4.
Stage Efficiency. For tray towers use the equations summarized in
Table A-l, tog~ther with Figure A-l in the Appendix. Extensive
physi~al property data are again required. The actual number of
trays required is given by the number of theoretical stages
divided by Eo' the overall tower efficiency.
-
Pressure drop. The pressure drop is required for costing the
pumps and air fan. Refer to Section 3.1.4 and 3.2.4 for tray and
packed towers, respectively.
6.
Ancillary Equipment
Obtain the surface area for all condensers, heat exchangers,
reboilers, and storage vessels as outlined in Section 3.3. Refer
to Figure 3-1 for an indicatiori of the equipment required for a
typical stripping plant.
.'
5.2
COST ES:I'IMATION
The correlations given in Section 4 are provided as a guide
to cost estimating. Wherever possible, costs from previou~
projects or vendor quotes should be used. The, correlat,ions are
readily programmabl~ for use on calculators or .small computers.
Finally, the capital cost, operating cost and overall
annualized cost should be plotted against the stripping medium
..
rate, and the optimum design selected.
5-5
-------
SECTION 6
REFERENCES
1.
Berger, B. B., editor, Contr~l
Water. and Wastewater, Office of
u.S. Environmental Protection
EPA-600/8-83-011,April, 1983.
~f organic ~ub~tances in
Research and Development,
Ag ency, Wash ing to n D.C.,
2.
Hicks, R. E., et. .!l.:. .Wastewater Trea.tment in Coal
Conversion., u.S. Environmental Protection Agency, Research
Triangle park, N.C., Contract 68-03-2207; April, 1979.
3.
Goldstein, D. J., .Air and Steam Stripping of Toxic
pollutants., Volumes 1 & 2, Industrial Environmental
Research Laboratory, U.S. Environmental Protection Agency, .
Cincinnati, -_9~it.-~~_-03-0_Q~,_May, 1982... -., '..
~. --.~.---- '-, " - ""''''''
/. - ~....-......
G.osset, J. M., .packed Tower Air Stri.rping of TCE 't'r~
~ilute Aqueous Solution", Engineering and Services,
.,~atory, Air Force Engineering and Services center,'
Tyndal-1..........~ir ~~_~ce B,~se, Flor..:..d_:.,__~:~~.,. -~---_._-- ,-- .
--.--..--
4.
Stover, E. L., .Removal of Volatile Organics from
Contaminated Ground Water., Proceedings of the Second
Rational Symposium on Aquifer Restoration and Ground Water
Monitoring, Columbus, OH, May, 1982.
6-1
-------
6.
Kincannon, D. F. and Stover,
. Groundwater Treatability - A Case
75(6),292, 1983.
.Contaminated
E.
L.,
Study., AWWA Journal,
7.
u.s. Environmental protection Agency, organic Data Base;
Plant Nos. 1290-010, 2930-035, and 3390-005, Effluent
Guidelines Division, washington D.C.
8.
U.S~ Environmental Protection Agency, .Treatability.Manua1-,
Volume 1, Treatability Data., Report No. 600/8-80-042a,
JUlY: 1980.---"-~ - --- --..-----.-------
Mfcarty, P. L. "Trace Organic RelOova1bY-AdvanCed wastew~?eJ
T~' J1. WPCF, 52(7) ,1907-1922, 1980. ----.- . ... .
King, C. J.-;-sep;r;~-s'es,.--M-cG'r-awHiil Boo.k Co., New
York, N. Y., 1980.
9.
10.
11.
Smith, B. D., Design of Equilibrium Stage processes, McGraw
Hill Book Co., New York, N. Y., 1963.
12.
Treybal, R. E., Mass-Transfer operations, Third Edition,
McGraw Hill Book Co., New York,N. Y.; 1980.
13.
Davies, J. A., -Bubble Trays - Design and Layout-, Part II,
Petrol Refiner, 29 (9), 121-130, 1950.
14.
Peters, M.S., and Timmerhaus, K. D., Plant Design and
Economics for Chemical Eng ineers, McGraw Hi 11 -Book Co.; New
York, N. Y., 1980.
15.
Lewis, A. N. and Randall, M., Thermodynamics, McGraw Hill
Book Co., New York, N. Y., 1950. ..,: ).
16.
Gilliland, E. R., Elements of Fractional Distillation,
McGraw Hill Book Co., New York, N. Y., 1950~.. . ".
6-2
-------
17.
18.
19.
20.
21.
22.
23.
24.
Reid, R. C., prausnitz, J.M. and Sherwood, T.K., The
properties of Gases and Liquids, McGraw Hill Book Co., New
York, N. Y., 1977.
Holmes, M. J. and Van Winkle, M., .prediction of Ternary
Vapor-Liquid Equi 1 ibria for Binary Data.' Ind. , Bng. Chell.,
-62(1),21-31, Jan., 1970.
Ambrose, D., .vapor-pressure Equations., Nat. Phys. Lab.
Rep. Chem., 19, Nov., 1972.
Hwang, S. T. and Fahrenthold'; p'., .Treatabi1ity of the
Organic priority pollutants by Steam Stripping., Water-1979,
AIChE Symposium Series No. 197, 76, AIChE, New York, 1980.
Fair, J. R., Petrol Chem. Bng. 33(10),45,1961.
Souders M. and Brown,'G. G., Des!gh. of Fractionating
Columns. I Entrainment and Layout., Ind. Bng. Chem., 26, 78-
103, 1934.
.Tray Efficiencies in Distillation Columns., Final Report
from the University of Michigan, AIChE, New York, 1960.
Kavanaugh, M. C., and Trussell, R. R., .,Design of Aeratio,n
Towers to Strip Volatile Contaminants fr~m Drinki~g Water.,
Journal ANWA, 72 (12), 684-692, 1980.
25. Onda, K., Takeuchi, H., ~nd Okumoto Y., .Mass Transfer
Coefficients Between Gas and Liquid Phases in Packed
Columns., Journal Chem. Bng. Japan, 1 (1), 56-62, 1968.
/_;~---,._------ ---,.,_. . -.. -_._-_._,.~----......
26., . Bolles, W. L. and Fair, J. R., .Improved Mass-Transfer Mo'd..~)
! Enhances Packed-Co 1 umn Des ign., Chemical Bngineering, ,:
" . pp 109-116~u.4'-i2, 1~
-------. ---........~.-- -..-
6-3
-------
27.
28.
29.
30.
31.
32.
330
34.
35.
36.
McCabe, W. L. and Sm i th, J. C., Uni t Operations of Chemica 1
EngineeriD~, McGraw ~i1l Book Co., New York, N. Y., 1976.
perry, R. H. and Chilton, C. H., Chemical
Handbook, McGraw Hill, New York, N. Y., 1973.
Engineer's
piku1ik, A. and Diaz, H. E., .Cost Estimating for Major
Process EquipmentW, Cheaical Bngineering, 84(21), 106-122,
OC t. 10, 1977.
Hall, R. So, Mat1ey, J. and McNaughton, K. J., .Current
Costs of Process EquipmentW, Chemic~l Engineerinq, pp 80-
'116, April 5, 1982.
,,""
e' ,____,e ,,_r--'.'-- " '.-----------
r:'~i..ler, ~,./S: and Kapella, W. A., wInsrffrl-e'd-Co,s.t--of a
~ U 10 tion Co 1 umn". Cheaica 1 Eng ineer ing. 84 (8). 12 9~3",. .
Ap ~-~ 19Z1~___,---_._----,. ---" '
------.
-,'
Corripio, .A:' B., M!'u1et, A. and Evans, L. B'., WEstimate Costs
, ,
of Distl'i lation & Absorption Towers vla~'.)cor.re1ationsW,
Chemicall Engineering, 88 (26), 77-82, Dec. 28.., 1981.
~~_h__" """.., d'_~_" ' "", ,--..----.---
Corripio, A. B., Mulet, A. and Evans, L. B., WEstimate Costs
of Pressure Vessels via Correlations., Chemical Engineering,
88 (20), 145-150, Oct. 5, 1981.
Enyedy, G., wA Computer Based Cost Estimation Service-, PDQ$
Inc., Gates Mills, OH, 1979.
Lang, H. J., wSimplified Approach to Preliminary Cost
EstimatesW, Chemical Eng ineer ing, 55, 112-113, June, 1948.
Hand, W. E., .From Flow Sheet to Cost EstimatesW, Petroleum
Ref., pp 331, Sept., 1938.
6-4
-------
37. Guthr i.~, ..K.M., '..pat'a"'ahc:r-'Tecfirilques, ..~~r preliminary Capital
Cost Estimates. , Cheaical Engineering~\ 76 (6), 114-142, March
)
,/
38. Viola Jr., L. J., .Estimate capi~a~/~osts via a New Shortcut
....-.-
~ Method., Cheaica1,E.n.gineerln9' 88(7), 80-86, April 6, 1981.
~...~-..- -.' . . ..~.., -...'
39.
40.
41.
42.
43.
44.
45.
/24,1969.
./:
/
Desai, M. B.', .pre 1 im i na ry Cos tEst ima t i ng 0 f Process
Plants., Cheaica1 Engineering, 88(15),65-70; July 27, 1981.
Shukla, H. M., Newmann Jr., C.
.Condensation Organic Vapor Control
the 75th APCA Annual Meeting, June,
R . and. Len, t , j . W. ,
Methods.,proccedings of
1982.
. Corri~i6, A. B., Chrien, K. S. and Evans, L. B., .Estimate
Costs. of Heat Exchangers and Storage Tanks via
Correlations., Che.i~a1 Engineering, 89(2), 125-127, Jan.
25, 1982.,
Corripioi A. B., Chrien, K. S. and Evans, L. B., .Estimate
Costs of Centrifugal Pumps and Electric Motors., Che.ica1
Engineering, 89(4), 115-118, Feb. 22, 1982. .
Weast, R. C. Ed., Handbook of Che.istry and Physics, 64th
edition, CRC Press, Boca Raton, FL 33431, 1983.
Wagman, D. D.
Ther.odynamic
Wash i ng ton D.C.,
..
et.a.!..:., The HBS Tables of Chemical
Pro per tie s , Am e r i can C hem i c a 1 So c i e t y ,
1983.
Arbuckl.e,W. B., .Estimating Activity Coefficients for use
inCal~ulating Environmental Parameters., Environ. Sci. «
Tech.,' ;17..<9):, 537-542, 1983.
6-5
-------
46.
47.
f f9 5-L-.1J /J 3 3
Weber, .~~ ~h, .Vapor pressur~.Di-str~l'b~ion of.
Sel ect(' Organic Chemical s., . u.s. Env i ronmenta 1 protection
Agency, Cincinnati, OH, EPA-600/2-81-021, Feb., 1981~)
\ ------:------- - . --_.-/
'.-..... ----.------- ----.".
Brady, G. S:~---Ed., Materials Handbook, 11th edition~ McGraw
H i 11 Boo k Company , New Yo r k, N. Y., 1974.
j..
.i>
6-6
-------
APPENDIX A
DETERMINATION OF TRAY EFFICIENCY
TABLE A-l
SUMMARY OF AIChE PROCEDURE FOR PREDICTION OF TRAY EFFICIENCY
1. Predict a value for (NTU)V' the number of phase transfer
. units from:
(NTU)V = 1.Sl/(Sc)0.S
where
Sc = dimensionless gas-phase Schmidt number
= ~ vi ( P G Dv) .
~V = gas-phase viscosity
PG = gas-phase density
DV = gas~phase diffusivity
of the pollutant into
water vapor
2. Compute the liquid residence time on a tray, tL' seconds
us ing :
tL = 0.66 D
where
D = tray diameter in ft
A-l
-------
3. predict a value for (NTU)L' the number of liquid-phase
transfe~ units, from:
(NTU)L = 89.2 D (DL)0.5
where
DL = liquid-phase diffusivity of pollutan~ into water at
nearly infinite dilution in water, ft2/hr
4. Compute the overall gas-phase transfer units, NTU, from
l/NTU = l/(NTU)v + S/(~TU)~
where
S = stripping factor, KV/L
5. Predict the point efficiency:
Ep = 1 - exp(-NTU)
6. .Compute a value for eff.ective diffusivity in the direction of
liquid flow:
DE = (10.034 + 1.026 v)2
ft2/hr
where
. ,
v = vapor velocity through the active trai area, ft/~
7. Compute the peclet number, pe, from:
pe = 3600 Zf/(DE tL)
where
Z1 = distance traveled on the tray by the liquid, ft;
may be taken as the distance between inlet and outlet
weirs, = 0.71 x diameter.
. '. . .
A-2
-------
8. Obtain the ratio EM/Ep
EM/Ep = (l - exp[-{M + pe)]}/{{M + Pe) [1 + {M + Pe)/M]}
+ (exp M - l)/{M[l + M/{M + Pe)]}
where
M = (Pe/2) [(1 + 4 S Ep/pe)O.S - 1]
9. Obtain the fractional entrainment (mass fraction of liquid
that is entrained), " from Figure A-l.
10. Correct the resulting tray efficiency for the effect of entrainment
using Colburn's equation:
EA = EM/[l + EM '/(1 - ,)]
11. The overall column efficiency can be rel~ted to the tray
efficiency by the following relatioship:
Eo = In[l + EA(S - l)]/ln S
. " ..'
A-3
-------
~
-' 0.2
~
z
3:
co
Q
en
en 0.1
co
a:::
-c,:,
~
-'
i 0.05
.......
~
-'
co
:E
..
~
~
z 0.02
~
:E
Z
-
<
a:::
~
z 0.01
~
-'
<
z
co
-
~ 0.005
<.,)
<
a:::
~
0.002
0.001
0".005
1.0
0.5
SIEVE PLATES
- - - BUBBLE-CAP PLATES
0.01
0.02
0.05
0.1
FLOOD,
PERCENT
90
80
70
60
50
30 354045
0.2 -- 0.5
1:.0
( )0.5
L PG
G p
L
Figure- A~l.
Entrainment Correlation 10
.. ..,'
A-4
-------
APPENDIX B
DESIGN EXAMPLE
A dilute aqueous solution Qf an organic is to be steam
stripped from an initial mole fraction of 27xlO-6 down to 0~6XIO-9
Desi3n the tower to handle a feed of 2100 moles/hr,' and' use a
. .
steam rate of 260 moles/hr (about 12.4 percent of the feed).
Assume the column operates at a pressure of one atmosphere and
100oe, and that at these conditions the vapor-liquid equilibrium
constant, K, has a value of 15.
1.
TRAY TOWER DESIGN (Perforated trays)
Theoretical Stages
Stripping factor
Removal efficiency
From Eq. (3.18)
S = KV/L = 15x260/2l00 = 1.86
f = 1 - 0.6xlO-9/27xlO-6 = 0.99998
N = 16.0 for R = 0, and
N = 8.73 for R = 99
In the reflux case, it was assumed that ~~ percent af the
overhead would be recovered in a decanter and returned to the
tower as aqueous reflux. Other values used wer~:
tB = 1000e tR = 950e
YD'= 1.0 (nearly pure organic phase ~ssumed)
YS= 1131 (values for specific compounds are tabulated
in the Appendix and in Reference 20)
ep = 1.0 cal/g.K}
L = 542 cal/g
which gives R' = 99.9 and k = 0.0108
B-1
-------
Tower Diameter
For water at 1000C:
Liqu.iddensity= 60 Ib/ft3; Vapordensity=0.037
Surface tension = 59.2 dyne/em
Liquid mass rate = 2100x18 = 37,800 Ib/hr
Gas mass rate = 260x18 = 4680 Ib/hr
1 b/ ft 3
To calculate the vapor velocity through ~he active tray area we
use Eq. (3.28) with:
Ah/Aa = 0.13
Tray spacing,
a = 0.16
b = 0.10
c = 0.20
Kv = 0.24
(assumed)
t. = 20 inches
From Eq. (3.26), Maximum allowable velocity, vm = °12.2 ft/s
The volumetric flow rate of the gas is 4680/(3600xO.037) or
35.1 ft3/s
so
A = 1.lx35.1/9.7 = 4.0 ft2
diameter = 2.26 ft
From Eq. (3.29)
Overall efficiency
The procedure in Table A-l is uS'ed with the following
values:
Diffusivity in the vapor phase,
Gas density, PG = 0.037 Ib/ft3
Gas viscosity, ~V = 294.3x10-4
DV = 1.73 ft2/hr
Ib/(ft.hr)
.. '. .,'
B-2
-------
Eq. (11-3.2) in Reference 17 may be used to estimate the
diffusivity. This reference ~nd other~28,43 contain data and
estimation methods for a number of physical properties.
The above values yield:
Sc = 0.46, and
(NTU)v :8 2.23
The liquid residence time on a tray.of diameter 2.26 ft is:
tL = 0.66x2.26 = 1.5 seconds
Diffusivities in the liquid phase may be estimated from
Eq. (11-10.1) in Reference 17.. Using a value of DL= 1.67x10-4
ft2/hr gives the number of liquid phase transfer units to be:
(NTU)L = 2.61
. .
Consequently the number of transfer units is:
NTU :8 (1/2.23 + 1.86/2.61)-1
:8 0.86
The point efficiency is:
DE = (10.03 + 1.03x9.7)2 = 401 ft2/hr
I f wet a k e the d i s tan c e t r a v e 1 e don a p 1 ate be twe e n we i r s to be
0.71 D, or 1.6 ft, then the peclet number becomes
Pe .= 3600xl. 62/ (401x1.5) = 15.3.
B-3
-------
In obtaining the Murphree tray efficiency we get:
M = 1.012, and
EM/Ep = 1.633, so EM = 0.947
Correcting this efficiency for entrainment using Figure A-l, we
get 'I' = 0.01 and
EA = 0.938
The overall towe~ efficiency therefore be~omes
Eo = 0.953
The actual number of trays is, therefore,
N = 16/0.95 = 17 without reflux
N = 8.7/0.95 = 9 with reflux
If a reboiler is used, one stage can be subtracted from the
actual tray count
2.
PACKED TOWER (for 2 inch ceramic packing)
Tower Diameter
From Figure A-l, for abscissa = 0.2, get ordinate = 0.09 at
f 1 0 od i n g. For a pac kin g f act 0 r C F = 6 5 and a 1 i qui d vis c 0 sit y .0 f
0.29 cP, we get for the vapor flux,
or
G1 = 1282 Ib/(hr.ft2)
G1 = 1026' 1b/(hr.ft2)
(at flood)
(at 80% of flood)
Therefore:
Area = 4680/1026 = 4.56 ft2, and
D = 2.5 ft
8-4
-------
Number of Transfer Units
From Eq. (3.34), using the same values as above for tray
towers,
NTU ::I 21.5
NTU ::I 11. 7
without reflux
wi th reflux
Height of ~ Transfer Unit and.Tower Height
Using the equations in Table 3-5 with viscosity, density and'
surface ten$ion factors of unity, we get:
(HTU)V::l 0.66 (Z)1/3, and
(HTU)L ::I 0.32 (Z)0.15
so,
HTU = Z/NTU ::I (HTU)v + S(HTU)L
and
Z/2l.5 = 0.66 zl/3 + 0.60 ZO.15
for no reflux
Solving for Z gives a packing height of 90 feet.
8-5
-------
APPENDIX C
ADIABATIC AIR STRIPPER: BASIC PROGRAM
This program solves the material balance equations in
Table 3-2 (no-condenser case) and the. enthalpy' balance,
. .'. .
Eq. (3.20), to determine the number of theoretical stages
required for an adiabatic air stripper.
The procedure involves an iteration within an iteration. The
temperature of the,air leaving the stripper is estimated, and
then the enthalpy balance applied to each stage until the bottom
stage is reached. If the air temperature entering this stage is
not equal to the specified air temperature, a new top temperature
must be assumed. The other iteration is made arou~d each stage.
The temperature on stage D+l must be estimated 'to calculate the
. entha lpy of the air entering stage n. Eq. (3.20) is then tes ted,
and a new temperature for n+l estimated until the equation
balances.
The program typically ,runs for
refined convergence procedures'
significantly. .
several minutes. The use of
should reduce this ~ime
C-l
-------
ADIABATIC AIR STRIPPER: PROGRAM LISTING
l~ GOTO 2~~ .
15 REM *** BASIC PROGRAM FOR ADIABATIC AIR STRIPPING
2~ REM ENTER THE VAPOR-LIQUID CONSTANT K HERE
25 REM WHERE K=y/x - concentrations are in mole fractions
3~ REM K is dimensionless
35 REM (K.IS THE HENRY LAW CONSTANT DIVIDED BY PRESSURE)
4~ REM THE FOLLOWING EQUATIONS MAY BE USEFUL:
45 REM LOG (K*P) = LOG (VAP.PRSSS) + LOG (GAMMA)
5~ REM LOG (VAP. PRESS} = A - B/(T + C}
55 REM LOG (GAMMA) = D + E/(T + 273}
6~ REM . THE FOLLOWING EQUATION IS FOR TCE
65 K = 1~A(7.~3 - 1315/(T + 23~} + 1134/(T + 273}}/(P * 76~}
7~ RETURN
75 REM *
***
EQUATION FOR ENTHALPY OF HUMID AIR, AIRHT, cal/mole
dry air
8~ AIRHT = 6.96*T + HUMID * rl~75~ + 8.1*~}.:RETURN
85 REM * EQUATION FOR MOLE FRACTION WATER VAPOR IN AIR, YW
9~ YW = EXP(21.158 - 592~.8/(T+273} - .~~6977 * (T+273})/(14.7 * P)
95 REM * EQUATION FOR HUMIDITY OF AIR,. HUMID, moles water/mole
. dry air
l~~ HUMID = YW/(l - YW}:RETURN
1~5 REM * EQUATION FOR ENTHALPY OF WATE, WATHT, cal/mole
11~ WATHT = 18 * T:RETURN
2~~ DIM L(2~},V(2~},HUM(2~},HEATW(2~},HEATA(2~},TEMP(2~}
2~5 DIM Y(2~}, X(2~}
21~ REM **** TABLE OF VARIABLES
215 REM * CONCENTRATION IN LIQUID, X,
22~ REM * CONCENTRATION IN VAPOR, y,
225 REM
23~ REM
235 REM
24~ REM
245 REM
25~ REM
255 REM
26~ REM
265 REM
27~ REM
275 REM
28~ REM
285 REM
29~ REM
* FLOWS ~RE IN MOLES/TIME FOR
L IS LIQUID RATE, MOLES/TIME.
AIR IS WET AIR RATE, VAIR IS THE DRY AIR RATE
N IS STAGE NUMBER, NB IS BOTTOM STAGE
TMAX,TMIN, TTMAX, TTMIN, TWMAX,TWMIN, TNMAX, TNMIN
ARE THE MAX. & MIN. TEMPS FOR THE
. TOWER, TOP, WET BULB, AND PLATE N, RESPECTIVELY
YW IS MOLE FRACTION WATER VAPOR IN SATURATED! AIR
YWIN IS MOLE FRACTION WATER VAPOR IN ENTERING AIR
YWIN IS MOLE FRACTION WATER VAPOR IN ENTERING AIR.
TEMP(N} IS TEMPERATURE ON PLATE N
EVAP IS THE WATER EVAPORATED ON A STAGE
TERROR, HERROR, YERROR, ARE THE
ERRORS ON THE TEMPERATURE, HUMIDITY, AND
IS MOLE FRACTION
IS MOLE RATIO (PER MOLE.
DRY AIR)
A FEED OF 1~~ MOLES/TIME
295 REM
3~~ REM
305 REM
310 REM
315 REM
4~0 REM
-
MOLE FRACTION
WATER VAPOR
HUM(N) IS THE AIR HUMIDITY, MOLES WATER/MOLE DRY AIR
HEATW IS THE ENTHALPY OF WATER, CAL/MOLE WATER
HEATA IS THE HUMID AIR ENTHALPY, CALS/MOLE DRY AIR
ALL TEMPERATURES ARE DEGREES CELSIUS
SEE LINES 42~ TO 450 FOR INPUT VARIABLES
**** SPECIFY INPUT VARIABLES ****
.. -...
C-2
-------
435 CLS:PRINT"ADIABATIC AIR STRIPPER"
413 PRINT"BEFORE STARTING, YOU SHOULD ENTER THE K-FACTOR".
415 PRINT" K = y/x IN LINES 20 TO 65"
423 PRINT"ENTER MOLE FRACTION ORGANIC IN LIQUID FEED "i:INPUTiXF
425 PRINT"ENTER DESIRED CONCENTRATION IN TREATED WATER"i:INPUTiXB
430 PRINT"ENTER FEED TEMPERATURE IN CENTIGRADE "i:INPUTiTF
435 PRINT"ENTER AIR RATE AS MOLE % OF LIQUID FEED "i:INPUTiAIR
443 PRINT"ENTER TEMPERATURE OF ENTERING AIR lIi~INPUTiTDRY
445 PRINT"ENTER RELATIVE HUMIDITY OF ENTERING AIR, % ni:INPUTiRH
450 PRINTnENTER MEAN PRESSURE IN TOWER, ATMOSPHERES "i :INPUTiP
455 REM ***** DETERMINE INITIAL VARIABLES *******
463 L13) = 130:TEMP(0) = TF:X(0) = XF
465 TMAX = TF:IF TDRY > TF THEN TMAX = TDRY:REM MAX. POSSIBLE
TEMP IN TOWER
470 TMAX = TMAX + .2
475 TTMAX=TMAX:REM LIMIT FOR TOP TE~PERATURE
483 M= 3:REM COUNT ON TOWER ITERATIONS
485 LN = 3:REM LAST STAGE CALCS
493 REM ** CALCULATE DRY AIR RATE FROM SAT. HUMIDITY AND
REL. HUMIDITY
533 T = TDRY:GOSUB 90:HUMIDITY = HUMID * RH/133
535 YWIN = HUMIDITY/(HUMIDITY + l):VAIR = AIR * (1 - YWIN):
. PRINTnDRY AIR RATE = lIiVAIR
513 REM **' CALCULATE THE FEED ENTHALPY (ASSUMING PURE WATER)
515 T = TF:GOSUB 113:FEEDHT = WATHT .
523 REM ** CALCULATE THE WET BULB TEMPERATURE OF THE ENTERING
AIR
525 IF RH = 130 THEN TWET = TDRY:GOTO ~85
533 TWMAX = TDRY:TWMIN = -10 .
535 TWET = TDRY - 10:REM INITIAL GUESS, ITERATION FOLLOWS:
549 YTEST = YWIN:IF YWIN <9.999999E-96 THEN YTEST = 9.999999E-06
545 T = TWET:GOSUB 99
550 YW2 = YW - (1. - YW)*(TDRY -TWET)/(1555 - .7 * TWET):
REM CARRIER EQN.
555 PRINT"TWET"iTWET,"CALC"iYW2,YWIN
569 YERROR = (YTEST - YW2)/YTEST
565 IF ABS(YERROR *199) < 1 THEN 585
579 IF YERROR > 0 THEN TWMIN = TWET:GOTO 580
575 IF YERROR < 3 THEN TWMAX = TWET
580 TWET = (TWMIN + TWMAX)/2:GOTO 545
585 TMIN = TWET - .2: IF TF < TWET THEN TMIN ='rF - .2: .
REM MINIMUM TOWER TEMP.
599 TTMIN=TWET + (TF - TWET)/4:REM LIMIT ON TOP TEMPERATURE
595 IF TWET > TF THEN TTMIN = TF
639 REM **** ESTIMATE TOP TEMPERATURE FOR-FIRST TIME
695 TEMP(l) = TF.- (TF - TWET)/5.
613 REM START OF TOWER/TOP TEMP ITERATION
615 REM ** CALCULATE THE HUMIDITY OF THE LEAVING AIR
620 LN = 0:T.= TEMP(l):GOSUB 99:HUM(1) = HUMID:
PRINTnTOP HUMIDIT~ = niHUM(l)
625 REM ** CALCULATE THE ENTHALPY OF THE LEAVING AIR
639 GOSUB 89:HEATA(1) = AIRHT .
635 REM ** CALCULATE BOTTOMS. WATER RATE
640 REM WATER ADDED TO SATURATE AIR IS
VAIR(INITIAL - FINAL HUMIDITY)
C-3
-------
645 WATER = VAIR * (HUM(l) - HUMIDITY)
650 BOTTOMS = 100 - WATER
655 REM ** CALCULATE THE ORGANIC IN THE- AIR LEAVING THE TOWER
660 Y(l) = (100 * XF -BOTTOMS * XB)/VAIR
665 PRINT"BOTTOMS =";BOTTOMS:PRINT"TOP VAPOR = ";Y(l)
670 PRINT"TOP TEMP =";TEMP(l)
700 REM *** START OF STAGE CALCULATIONS
705 N=O.
710 L(0) = 100
715 N=N+1:IF N = 20 THEN PRINT "Increase air rate":STOP
720 PRINT:PRINT:PRINT N
725 REM ** ENTHALPY BALANCE TO DETERMINE TEMPERATURE ON NEXT PLATE
730 TEMP(N+1) = TEMP(N) - (TEMP(N) - TDRY)/3:REM INITIAL GUESS
735 TNMAX = TMAX:TNMIN = TMIN
740 REM ** CALCULATE ENTHALPIES AND HUMIDITIES FOR
745 REM F.FEEDHT + VAIR.AIRHT(N+1) = .
VAIR.AIRHT(l) + L(N).WATHT(N)
750 T = TEMP(N):GOSUB 110:HEATW(N) = WAT~T .
755 T = TEMP(N+1):GOSUB. 90:HUM(N+1) = HUMID
760 IF N = NB THEN HUM(N+1) = HUMID * RH/100:PRINT"BOTTOM"
765 HUMID = HUM(N+1):GOSUB 80:HEATA(N+1) =.AIRHT
770 REM** CALCULATE THE WATER EVAPORATED ON STAGE N
775 EVAP = (HUM(N) - HUM(N+1» * VAIR:L(N) = L(N-1) - EVAP
780 REM ** CALCULATE AIR ENTHALPY ON (N+1) BY.HEAT BALANCE
785 AIRHT = (VAIR * HEATA(l) + L(N) * HEATW(N) - 100 * FEEDHT)
/VAIR
790 HERROR = (HEATA(N+1) - AIRHT)/HEATA(N+1) * 100
795 PRINT TEMP (N+1) ,"ERROR= "; HERROR;" 'ON STAGE"; N
800 IF ABS(HERROR) < .01 THEN 860
805 IF HERROR < 0 THEN TNMIN = TEMP(N+1):GOTO 815
810 IF HERROR > 0 THEN TNMAX. = TEMP (N+1)
815 TEMP(N+1) = (TNMAX + TNMIN)/2
820 IF (TEMP (N+1) - TMIN) < .01 THEN 835
825 IF (TMAX ~ TEMP(N+1» < .01 THEN 840
830 GO TO 750
835 TTMIN = TEMP(l) - (TEMP(l) ~ TTMIN)/3:GOTO 845
840 TTMAX = TEMP(l) + (TEMP(l) - TTMIN)/3
845 TEMP(l) = (TTMAX + TTMIN)/2 .
850 PRINT:PRINT "TRYING A NEW TOP TEMPERATURE"
855 GOTO 620 .
860 REM **** CALCULATE THE COMPOSITION ON THE NEXT STAGE
865 REM Y(N+1) = (VAIR * y(l) + L(N) * x(N) - F * xF)/VAIR
870 REM ** CALCULATE EQM. LIQUID CONCENTRATION ( x = y/K
875 T = TEMP(N):GOSUB 55
880 X(N) = Y(N)/«Y(N)+
885 IF X(N) > XF AND M >
1) * K)
o THEN PRINT.
"Increase air rate,x(n)=";~(N):STOP
890 IF NB > 0.THEN 905
895 REM * CHECK IF LIQUID CONCENTRATION IS LESS THAN BOTTOMS
900 IF X(N) <= XB THEN NB = N: GOTO 730
905 IF N = NB THEN .935
910 REM ** MATERIAL BALANCE FOR Y(N+1)
915 Y(N+1) = (VAIR * Y(N) + L(N) * X(N) - 100 * XF)/VAIR
920 IF Y (N+1) < 0 THEN Y (N+1) = 0
925 PRINT Y(N+1),Y(N),X(N),K
.. ...."
C-4
-------
930 GO TO 715
935 IF X(N) <= XB THEN 1000
940 NB = NB + 1
945 TTMAX = TTMAX + (TMAX - TTMAX)/5
950 GOTO 620
1000 REM ** CHECK INITIAL TEMPERATURE GUESS
1005 TERROR = (TORY - TEMP(N+1»/TORY
1010 IF ABS(TERROR) < 5.000001E-03 THEN 1200
1015 M = M + 1
1020 PRINT N,TEMP(N+1),TORY,TERROR
1025 PRINT:PRINT:PRINT
1030 IF TERROR < 0 THEN TTMAX = TEMP(l):GOTO 1040
1035 IF TERROR> 0 THEN TTMIN = TEMP(l)
1040 TEMP(l) = (TTMAX + TTMIN)/2:GOTO 1045
1045 IF M > 20 THEN 1200
1050 GOTO 620
1200 REM ** PRINT RESULTS TO SCREEN ** Modify format as required.
1205 CLS
1210 PRINT"INPUT DATA"
1215 PRINT"FEED CONC=";XF;"
1220 PRINT"AIR RATE=";AIR;"
FEED T=";TF;" BOTTOMS=";XB;" K=";K
AIR TEMP= ";TORY;" PRESSURE=";P;
"RH = "; RH
1225 PRINT:PRINT"CALCULATEO VALUES"
1230 PRINT"TWET = ";TWET;" BOTTOMS = ";BOTTOMS
1235 PRINT"CALC AIR HUM= ";HUM(N+l);" GIVEN = ";HUMIDITY
1240 PRINT:PRINT"STAGE . TEMP LIQUID X*1000 Y*1000"
1245 PRINT"FEED ";: PRINT USING "titi. i#i ";TEMP (0),L (0),X (0)
. *1000
1250 FOR I= 1 TO N:PRINT I;" ";:PRINT USING "fH##.### ";
TEMP(I),L(I),X(I)*1000,Y(I)*1000:NEXT I
"; TEMP (N+1 ) ;." (GIVEN AS "; TORY;
. . n ERROR % =";-TERROR * 100;")"
1260 IF X(NB-1) < XB THEN PRINT:PRINT
" Number of stages falls between ";NB;" and ";NB-1
1255 PRINT "AIR
1270 STOP
TYPICAL OUTPUT
INPUT DATA
FEED CONe= .1
AIR RATE= 75
FEED T= 40 BOTTOMS= .0000001 K= 578.588
AIR TEMP: 2~ PRESSURE= 1 RH= 50
aw:tJLATED VALUES
TWET = 13.90625 BOTTOMS = 97.2744
CALC AIR HUM= 1.159221E-liJ2 GIVEN = 1.165858E-liJ2
STAGE
FEED
1
2
3
AIR
TEMP
40.000
31.916
25.935
20.368
19.91004
LIQUID
100.000
98.912
98.179
97.269
(GIVEN AS
X*1000
100.000
0.169 134.888
0.1iJ00 0.225..
0.1iJ00 0.000
20 ERROR % =-.449791 )
Y*1000
. C~5
-------
APPENDIX D
ANTOINE COEFFICIENTS: BASIC PROGRAM
The program calculates the Antoine coefficients A, B, C in
Eq. (2~15) from three pairs of temperature-vapor pressure data.
The data may be in mixed units. The coefficients calculated are
for determining vapor pressure in - atmospheres using temperature
in Kelvin.
PROGRAM LISTING
.10 REM
20 REM
30 REM
40 REM
50 REM
60 REM
70 REM * Requires 3 vapor pressure - temperature data points *
80 REM Conversion factors for pressure and temperature
90 DIM PVERT(5),TVERT(3,3) .
'100 RE;M Pressure. and temperature data points
110 REM 'Subscript' IN for input value, CALC for calculated value,
120 REM CON for input value converted to selected units
130 REM and ERR is for error between input and calculated value.
. . 140 DIM PIN(3),PCALC(3),PCON(3),PERR(3),TIN(3),TCON(3)
, , 150 DIM ALP(3):REM log oi vapor pressure
169 REM Alpha variables for units
179 DIM PUNIT$(5),TUNIT$(3)
189 REM Read units
199 DATA "inm Hg ", "inch Hg ", "psi ", "bar
290 DATA "F ", "C ", "K "
210 FOR I=l TO 5:READ Z$:PUNIT$(I)=Z$:NEXT I'
229 FOR I=l TO 3:READ Z$:TUNIT$(I)=Z$:NEXT I
239 REM Read conversion factors
240 REM Pressure: from mm Hg, inch Hg, psi, bar, to atm
250 DATA 760, 29.92,14.696,9.9872,1.09
'260 REM Temperature: from F, C, to Kelvin
279 DATA 9.5556,255.37,1.9,273.15,1.9,0
289 FOR I=l TO 5
290 READ Z:PVERT(I)=l/Z
. 399 NEXT I
.319 FOR I=l TO 3:FOR J=l TO 2
:3~9 READ Z:TVERT(I,J)=Z .
330 NEXT J:NEXT I
.599 CLS
510 PRINT"Calcu1ate coefficients in the Antoine Equati6n"
511 PRINT" 10g19 p(atm) = A - B/[C + T(Kelvih)J"
512 PRINT"Any of the 'following units may be used in mixed mode"
Program to calculate the Antoine coefficients
Antoine equation: 10g(10)p = A - B/(C + T)
p is vapor pressure, T is temperature.
Units are program selectable.
", "atm
D-l
-------
513 PRINT"The data will be converted to atm. and Kelvin"
530 PRINT"Pressure units:";
540 FOR I-I TO 5:PRINT I; "-";PUNIT$(I);:NEXT I
550 PRINT"Temperature :";
560 FOR I-I TO 3:PRINT I; "-";TUNIT$(I)J:NEXT I
565 PRINT:PRINT"Data entry":
570 FOR N-l TO 3 .
580 PRINT "PRESSURE, unit '";:INPUT; I:IF I> 5 THEN GOTO 580
590 PRINT" Value";:INPUT;PIN(N):PRINT PUNIT$(I),
600 PCON(N)=PIN(N) * PVERT(I)
610 IF I < 5 THEN PRINT" or ";PCON(N); PUNIT$(5);
620 PRINT:PRINT "TEMPERATURE, unit ,u;:INPUT; I:IF I>3 THEN GOTO 620
630 PRINT" Value";:INPUT;TIN(N):PRINT TUNIT$(I),
640 TCON(N)-TIN(N) * TVERT(I,I)+TVERT(I,2)
650 IF I < 3 THEN PRINT" or ";TCON(N); TUNIT$(3);
66~ PRINT:PRINT:NEXT N
710 FOR N=1 TO 3
720 ALP(N)-LOG(PCON(N»/2.302585
730 NEXT N
740 Dl-(TCON(I)-TCON(2»/(ALP(2)-ALP(1»
750 02=(TCON(2) - TCON(3»/(ALP(3) - ALP(2»
760 03=(ALP(3) * TCON(3) - ALP(2) * TCON(2»/(ALP(3) - ALP(2»
770 04=(ALP(2) * TCON(2) - ALP(1) * TCON(1»/(ALP(2) - ALP(l»
780 A-(D3 - 04)/(01 - 02)
790 C=-l * (A * 01) - 04
800 a=(A-ALP(1» * (C + TCON(l»
805 PRINT:PRINT"Input T, K Input P, atm
810 FOR N=1 TO 3
820 PCALC(N) = 10A(A - a/(C+TCON(N»)
830 PERR(N)=(PCON(N) - PCALC(N» * 1~0/PCON(N)
840 IF PERR(N) < .001 THEN PERR(N) =0
860 PRINT" "; .
870 PRINT USING ".tt...t ";TCON{N);PCON(N);PCALC(N);PERR(N)
880 NEXT N
890 PRINT:PRINT"A =";A;" a =";a;" C =";C
Calc. P,
% error:"
0-2
-------
APPENDIX E
TOXIC ORGANICS LIST
Data from Reference 3 is summarized here for convenience.
For a method of predicting activity coefficients and Henry's Law
con s tan t for com po un d s not 1 i s t e d her e, see Ref ere n c e 4 5. Va po r
pressure data are list~d in Referen~e 4~.
Compounds directory
Ethers
phthalates
Nitrogen Compounds
Phenols
Aromatics
polynuclear Aromatic Hydrocarbons
PCB's and Related Compounds
Halogenated Hydrocarbons
Pesticides
Oxygenated Compounds
Miscellaneous
Stripping Category, ~
5-1
6-1
7-1
8-1-
9-1
10-1
11-1
-12-1
13-1
14-1
15~1
to 5-7
to 6-6
to 7-14
to 8-14
to 9-20
to 10-17
to 11-7
to 12-31
to 13-46
to 14-18
to 15-7
1
2
3
4
5
6
D
P
Very easily ~tripped
Easily stripped
Intermediate
Difficult
Very difficult
Cannot be stripped
Decomposes in water
" of"..
Poor data, ~ncertain classification
E-1
-------
I '
\ \
EASE OF STRIPPING, HENRY'S LAW CONSTANT AND ACTIVIT~ COEFFICIENTS
5-1
5-2
5-3
5-4
5-5
5-6
5-7
6-1
Bis(ch10romethy1) ether
109 H(mm H9) = decomposes
109 (Activity coefficient)
S = D
= decomposes
Bis(2-ch10roethy1}ether S = 3
.109H(mm B9) = 7.659 - 1969/(tOC + 234) + 862/(t~C+ 273)
109 (Activity coefficient) = 862/(~oC + 273)
Bis(2-ch10roisopropy1) ether .
109H(mm H9) = 7.94 - 123l/(tOC + 273)
109 (Activity coefficient) = ll17/(tOC + 273)
S = 2
2-Ch10roethyl vinyl ether
109H(mm B9) = 7.75 - 1079/(tOC + 273)
109 (Activity coefficient) = 774/(tOC + 273) .
S = 1
4-Ch10ropheny1 phenyl ether
H(100oC) = 3.9 atm
Activity coefficient(1000C) = 2.948x103
S = 4-P
4-Bromopheny1 phenyl ether
H (100°C) = 9.9 atm. . .
Activity coefficient(1000C)
S = 3-P
= 4418
Bis(2-ch1oroethoxy)methane ,
109 H(mmHq) = 8.151 - 2416/(tOC + 243)
109 (Activity coefficient) = 618/(tOC +
S = 6
+ 618/(tOC + 273)
273)
Dimethyl phthalate S = 6
10q H(mm Hq) = 8.293 - 2858/(tOC + 244) + 1036/(tOC + 273)
109 (Activity coefficient) = 1036/(tOC + 273)
E-2
-------
6-2
6-3
6-4
6-5
6-6
7-1
7-2
7-3
7-4
Diethy1 phthalate S = 5
109 8(mm 89) = 8.29 - 2883/(tOC + 239) + 1248/(tOC + 273)
109 (Activity coefficient) = 1248/(tOC + 273)
Di-n-buty1 phthalate
_8(250C) = 0.15 atm
Activity coefficient(250C)
s = 4-P
= 1. 2x106
Di-n-octy1phtha1ate
8(1000C» = 9,947 atm .
Activity coefficient(1000C)
S = 3-P
=.5.4X107
Bis(2-ethy1hexyl) phthalate
109 8(mm 89) = data uncertain
109 (Activity coefficient) = data
S = 4-P
uncertain
Butyl benzyl p~thalate
109 8(mm 89) = data uncertain
109 (Activity coefficient) = data
S = 3-P
uncertain
N-nitrosodimethy1amine
109 8(mm 89) = no data
109 (Activity coefficient)
S = 4-P
= no data
N-nitrosodipheny1amine
109 8(mm 89) = no data
109 (Activity coefficient)
S = 4-P'
= no data
. .
N-nitrosodi-n-propy1amine"
109 8(mm 89) = data uncertain
109 (Activity coefficient) = data
S = 3-P
uncerta.in
Benzidine
109 8(mm 89) = data uncertain
109 (Activity coefficient) = data uQ~erta'n..
S = 3-P
E-3
-------
7-5
7-6
7-7
7-8
7-9
7-10
7-11
7-12
7-13
3,3'-Dichlorobenzidine
H(lOOoC» = 0.013 atm
Activity coefficient(1000C) = 65
1,2-Diphenylhydrazine
8(1000C» = 0.006 atm
Activity coefficient(lOOoC) = 19
s - 4-P
S . 4-P
Acrylonitrile . S = 2
109 8(mm 89) = 6.655 - 1208/(tOC T 222) + 545.6/(tOC + 273)
109 (Activity coefficient) = (545.6/T) - 0.2606
n-Butylamine
109 8 (mm 89)
109 (Activity
= no data
coefficient) = no data.
S = P
Diethylamine S = 3-P
109 8(mm 89) = 16.52 - l127/(tOC.+ 220) - 2804/(tOC ~ 273)
109 (Activity coefficient) = 9.55 - 2804i(tOC + 273)
Ethylene diamine
109 8(mm 89) = no. data.
109 . (Activi ty coeff ic ient)
= no data'
Monoethylamine
109 .8 (mm 89) = no data
109 (Activity coefficient)
= no data
Monomethylamine
109 H(mm 89) = no data
109 (Activity coefficient)
= no data
Triethylamine
H(200C) = 26 atm
Activity coefficient(200C)
= 375
E-4
S = 6
s =P
S = P
S = 1
-------
7-14
8-1
8-2
8-3
8-4
.8-5
8-6
8-7
8-8
Trimethylamine
109 H(mm H9) = no data
109 (Activity coefficient)
s = P
= no data
Phenol
s = 5
-109 H(mm 89) = 9.071 - 0.00352 tOC - 1517/(tOC + 174)
.109 (Activitycoefficient) = 1.941-0.00352 tOC(55 to 240°C)
2-Ch10ropheno1
109 H(mm H9) = 7.24 - 1668/(tOC + 210) + 715/(tOC
10g(Activitycoefficient)= 715/(tOC + 273)
S = 4
+ 273)
2,4-Dich10ropheno1
109 H(mm H9) = 8.205 - 2380/(tOC + 237)
109(Activitycoefficient)= 992/(tOC +
S = 4
+ 992/(tOC + 273)
273)
2,4,6-Trich10rophen01 . S
109 H(mm H9) = 8.096 - 2484/(tOC + 230) + 1230/(tOC
10g(Activitycoefficient)= 1230/(tOC.+ 273)
= 4
+ 273)
Pentachlorophenol
log H(mm H9) = 12.047 - 7751/(tOC + 509)
109(Activitycoefficient)= 1800/(tOC +
S = 3
+ 1800!(tOC + 273)
273)
2-Nitrophen01, O-nitropheno1 5 = 3
10g H(mm H9) = 8.014 - 2173/(tOC + 231) + 971/(tOC' + 273)
10g(Activitycoefficient)= 0.251 + 971/(tOC + 273)
4-Nitropheno1, p-nitropheno1
109 H(mm H9) = 7.483 - 2604/(tOC + 273)
10g(Activitycoefficient)= -3.412 + 1816/(tOC
S = 6
+ 273).
2,4-Dinitropheno1
log H(mm Hg) = no data
10g (Activity coefficient)
S = 6
= -0.767 :t 1173/ c.~oc + 273)
, E-5
-------
8-9
8-10
8-11
8-12
8-13
8-14
./ 9-1
9-2
9-3
Resorcinol
log B(mm Bg) = 5.31 - 3049/(tOC +
10g(Activitycoefficient)= -3.50
S = 6
238) + 1191/(tOC + 273)
+ 1191/(tOC + 273)
2,4-Dimethyl phenol, 2,4 Xylenol S = 5
log B(mm, Bg) = 7.06 - 1587/(tOC + 170) + 923/(tOC
10g(Activitycoefficient)= 923/(tOC + 273)
+ 273)
Total phenols
See ind~vidual phenols
p-chloro-m-cresol
B'(lOOoC) = 8.2 atm
Activity coefficient(lOOoC)
S = 3-P
= 376
4,6-Dinitro-o-cresol
10g B(mm Bg) = uncertain data
log (Activity coefficient) = no data
S = 5
Cresols S = 5
log B(mm Bg) = 7.51 - 1856/(tOC + 199) + 760/(tOC +273)
10g(Activitycoefficient)= 760/(tOC + 273)
Benzene
log H(mm Hg) = 7.455 - 1211/(tOC +
10g(Activitycoefficient)= 0.5455
S = 1
221) + 848.2/(tOC + 273)
+ 848.2/(tOC + 273)
Chlorobenzene S = 1
log H(mm'Bg) = 6.845 - 1431/(tOC + 218) + 1274/(tOC + 273)
10g(Activitycoefficient)= -0.135 + 1274/(tOC + 273)
1,2-Dich10robenzene, o-Dichlorobenzene
log H(mm Hg) = 7.07 - 1650/(tOC + 213) +
10g(Activitycoefficient)= 1421/(tOC +
S = 1
1421/(tOC + 273)
273)
. o. ..'
E-6
-------
9-4
9-5
9-6
9-7
,9-8
9-9
9-10
9-11
9-12
1,3-D~ch10robenzene, m-Dich10robenzene S = 1
10g'H(mm Hg) = 7.30 - 1782/(tOC + 230) + 1407/(tOC + 273)
10g(Activitycoefficient)= 1407/(tOC + 273)
1,4-Dich10robenzene, p-pich10robenzene S = 1
-log H(mm Hg) = 7.00 - 1575/(tOC + 209) + 1425/(tOC + 273)
.10g(Activitycoefficient)= 1425/(tOC + 273)
1,2, 4-Tr ich10robenzene S = 1
log H(mm Hg) = 7.601 - 2175/(tOC + 248) + 1533/(tOC + 273)
10g(Activitycoefficient)= 1~33/(tOC + 273) ,
Hexach10robenzene S = 1
log H(mm Hg) = 9.836 - 4630/(tOC + 356) + 2138/(tOC + 273)
10g(A~~ivityco~fficient)= 2138/(tOC + 273)
Ethyl Benzene
log H(mm Hg) = 6.286 - 1424/(tOC +
10g(Activitycoefficient)= -0.664
S = 1
213)+ 1529/(tOC + 273)
+ 1529/(tOC + 273)
Nitrobenzene S = 3
10g H(mm Hg) = 6.856 - 1740/(tOC + 200) + 1141/(tOC + 273)
10g(Activitycoefficient)= -0.262 + 1141/(tOC + 273)
Toluene, 'S.= 1
10g H(mm Hg) = 6.95 -'1345/(tOC + 219) + 1192/(tOC + 273)
10g(Activitycoefficient)= -0.0004 + 1i92/(tOC + 273)
2,4-Dinitroto1uene
log H(mm Hg) = 7.450 - 1238/(tOC + 273)
10g(Activitycoefficient)= 1380/(tOC + 273)
S = 3
2,6-Dinitroto1uene
Use data for 9-11
S = 3
.. ..."
E-7
-------
9-13
9-14
9-15
9-16
9-17
9-18
9-19
9-20
10-1
Aniline
See Referenc'e 3
Benzoic acid
S = 6
S = 6
109 H(mm H9) = 9.922 - 2816/(tOC + 273)
109(Activitycoefficient)= 0.8894 + 516.6/(tOC + 273)
Benzyl chloride S = 1
109 H(mm H9) = 7.546 - 1923/(tOC + 233) + 1265/(tOC + 273)
109(Activitycoefficient)= 1265/1tOC + 273)
Styrene, S =1
109 H(mm 89) = 9.31 - 1446/(tOC + 209) + 570/(tOC + 273)
1Q9(Activitycoefficient)= 2.34 + 570/(tOC + 273)
Quinoline
109 H(mm H9) = 7.89 - 2398/(tOC + 244) + 619/(tOC
109(Activitycoefficient)~ 619/(tOC+ 273),
S = 5
+ 273)
xy1enes S = 1
109 H(mm H9) = 7.00 - 1460/(tOC + 215) + 1339/(tOC + 273)
109(Activitycoefficient)= 1339/(tOC + 273)
Nitroto1uene
ortho: 109
meta: 109
para: 109
109 (Activity
H(mm 89) =7.97 - 1263/(tOC +273)
H(mm 89) = 8.07 - 1368/(tOC + 2.'73)
H(mm 89) = 7.98 - 1359/(tOC + 273)
coefficient) = 1250/(tOC + 273)
Napthenic acid
No data
2-Ch10ronaphtha1ene
109 8(mm H9) = 9.93 - 1644/(tOC + 273)
109 (Activity coefficient) = 1785/(tOC
+ 273)
E-8
S = 2
S = 1
.. 'w..'
-------
10-2
10-3
10-4
10-5
10-6
10-7
10-8
10-9
10-10
Benzo(a) anthracene
Uncertain data
Benzo (b) f1uoranthene
H(1000C) = 208 atm
- 109 (Activity coefficient)
= 2600/(tOC + 273)
Benzo(k)f1uoranthene
109 H(mm H9) = 7.445 - 837/(~oC + 273)
109 (Activity coefficient) = 2600/(tOC
+ 273)
Benzo(a)pyrene
109 H(mm H9) = 5.57 + 47/(tOC + 273)
109 (Activity coefficient) = 2716/(tOC
+ 273)
Indeno(1,2,3-cd)pyrene
Uncertain data
Dibenzo(a,h)anthracene
Uncertain data
Benzo(9hi)pory1ene
H(1000C) = 0.104 atm
109 (Activity coefficient) = 2946/(tOC + 273)
. ,
Acenaphthene
109 H(mm H9) = 8.033 - 1200/(tOC + 273)
109 (Activity coefficient) = 1635/(tOC + 273)
Acenaphthylene
H(250C) = 6.3 atm
Activity coefficient(250)
= 20
10-11 Anthracene
109 H(mm H9) = 8.91 - 1739/(tOC + 273)
109 (Activity coefficient) = 2022/(tOC + 2~3)
E-9
5 = 3-P
5 = 1-P
5 = I-P
5 = I-P
5 = 3-P
5 = 3-P
5 = 6
.5 = 1
5 = 2
5 = 3
-------
10-12 ehrysene
Data uncertain
S == 1-P
10-13 F1uoranthene
Data uncertain
S == 1-P
10-14 Fluorene
109 H(mm H9) = 8.06 - 1301/(tOe + 273)
109 (Activity coefficient) = 1632/(tOe + 273)
S = 2
10-15
Naphthalene S = 1
109 H(mm H9) = 7.01 - 1734/(tOe + 2Q~) :'" 1473/(tOe + 273)
109 (Activitycoefficient)= 1473/(tOe + 273)
10-16 Phenanthrene S = 2
109 H(mm H9) = 7.26 ~ 2379/(tOe + 204) + 2022/(tOe + 273)
109 (Activitycoefficient)= 2022/(tOe + 273)
10-17
pyrene
H = Data uncertain
109 (Activity coefficient) =
S = 1
2365/ (tOe + 273)
10-18
Aroc10r 1016
109 H(atm) = 6.22 - 1555/(tOe + 273)
Activity coefficient (100oe) = 1.0X105.
S = 1
11-2
Aroclor 1221
109 H(atm) = 8.05 - 2026/(tOe + 273)
Activity coefficient (lOOoe) = 1.1x105
S = l'
11-3-
Aroclor 1232
109 H(atm) = 8.24 - 1954/(tOe + 273)
Activity coefficient (1000e) = 3.6x105
S = 1
.. " ,."
E-10
-------
11-4
. ,11-5
11-6
11-7
12:"'1
12-2
/12-3
J12-4
Aroclor 1242 .
log H(atm) . 10.01 - 2577/(tOC + 273)
Activity coefficient (100°C) = 1.3xl06
Aroclor 1248
- log H(atm) . 9.80 - 2313/(tOC + 273)
Activity coefficient (100°C) . 5.9xl06
Aroclor 1254
log H(atm) = 9.46 - 2024/(tOC + 273)
Activity coefficient {100°C) = 2.1x107
Aroclor 1260
10g.H(atm) = 9.28 - 1928/(tOC + 273)
Activity coefficient (100poC) = 7xl07
Methyl chloride, chloromethane
log H(mm Hg) = 5.233 + 276/(tOC + 273')
log (Activity coefficient) = -2.247 + 1424/(tOC
Methylene chloride, dichloromethane
log H(mm Hg) = 9.58 - 1139/(tOC + 231)
Insufficient knowledge on the variation of
coefficient with temperature; a value of 315
of temperature) can be used.
"
Chloroform
log H(mm Hg) = 10.67 - 1171/(tOC +
log (Activitycoefficient):= 3.73 -
S = 1
S = 1
S = 1
S = 1
S = 1
+'273)
S = 1
the activity
( independent
S = 1
227) - 260/(tOC + 273)
260/(tOC + 273)
Carbon tetrachloride
S = 1
. .
log H(mm Hg) = 6.796 - 1220/(tOC + 227) + 1237/(tOC + 273)
log (Activitycoefficient)= -0.0979 + 1237/(tOC + 273)
E-ll
-------
12-5
~2-6
J12 -7
J 12-8
J 12-9
J 12-10
ehloroethane, ethyl chloride S ~ 1
log 8(mm 8g)" = 6.324 + 993/(tOC + 273) - 1013/(tOe + 237)
10g (Activit~oefficient)= -0.616 + 993/(tOe + 273)
l,l-Dichloroethane
10g 8(mm 8g) = 6.99 - 1171/(tOe + 228) + 895/(tOe
log (Activitycoefficient)= 895/(tOe + 273)
S ~ 1
+ 273)
1,2-Dichloroethane S = 1
109 8(mm 8g) = 7.03 - 1271/(tOe + 223) .:.. 824/(tOe .+ 273)
log (Activitycoefficient)= 824/(tOe + 273)
l,l,l-Trichloroethane S = 1
log H(mm Hg) = 7.03 - 1276/(tOe + 234) + 947/(tOe + 273)
log (Activitycoefficient)= 947/(tOe + 273)
1,1,2-Trichloroethane
log H(mm Hg) = 7.17 - l351/(tOe +
log (Activit~oef£:icient)= 0.199
S = 1
217) + 884/(tOe + 273)
+ 884/(tOe + 273)
1,1,2,2,-Tetrachloroethane
log H(mm Hg) = 7.54 - 1683/(tOC +
log (Activitycoefficient)= 0.227
. S = 1
234) + 960/(tOe + 273)
+ 960/(tOe + 273)
12-11 Hexachloroethane
log H(mm 8g) = 8.649 - 903/(tOC + 273)
109 (Activity coefficient) = 1745/(tOe + 273).'
\ 12-12 vinyl chloride
H (lOOoe) = 4.6x103 atm
Activi~y coefficient (lOOoe) = 145 .
12-13
S = 1
S = 1.
1,2-Dichloropropane S = 1
10g H(mm 8g) = 6.966 - l296/(tOe + 221) + 10.10/(tOe + 273)
log (Activitycoefficient)= 1010/(tOe + 273)
.. ....'
E-12
-------
12-141,3-Dich10ropropene
109 H(atm) = 4.543 - 670/(tOC + 273)
5 = 1
12-15 Hexach10robutadiene .5 = 1
109 H(mm H9) = 5.819 - 1029/(tOC + 130) + 1926/(tOC + 273)
109 (A c t i v i t yc ° e f" f i c i en t) = 1 9 28 / ( to C + 273)
12-16 "Hexach10rocyc10pentadiene
109 H(mm H9) = 8.00 - 644(tOC +273)
109 (Activity coefficient) =-2067/(tOC
5 = 1
+ 273)
12-17
Methyl Bromide 5 = 1
109 H(mmH9) = 6.96 - 987/(tOC + 238) + 1123/(tOC + 273)
109 (Activitycoefficient)= -0.00200 + 1123/(tOC + 273)
12-18 Dich10robromomethane
109 H(atm) = 3.795 - 513/(tOC + 273)
12-19 Ch10rodibromomethane
109 H (atm) = 5.262 ~ 1078/(tOC + 273)
12-20 Bromoform, Tribromomethane
109 H (a tm) = 5.464 - 1188/(tOC + 273)
5 = 1
5 = 1
5 = 1
12-21
Dich10rodif1uoromethane 5 = 1
109 H(mm H9) = 5.954 - 382.6/(tOC + 145)(+ 1305/(t?C + 273)
109 (Activitycoefficient)= 1305/(tOC + .273)
12-22 Trichlorofluoromethane
. -
109 H(mm H9) = 6.884 - 1043/(tOC + 237)
109 (Activitycoefficient)= 1040/(tOC +
5 = 1
+ 1040/(tOC + 273)
273)
112-23
Trichloroethylene 5 = 1
109 H(mm H9) = 7.03 - 1315/(tOC + 230) + 1134/(tOC + 273)
109 (Activitycoefficient)= 1134/(tOC + 273)
E-13
-------
)12-24
j 12-25
j 12-26
12-27
1,1-Dich10roethylene
109 8(mm 89) = 7.606 - 1471/(tOC + 280)
109 (Activitycoefficient):I 903/(tOC +
S :I 1
+ 903/(tOC + 273)
273)
1,2-Trans-Dichloroethylene . S:I 1
109 8(mm 89) = 7.037 - 1l45/(tOC + 228) + 846/(tOC + 273)
109 (Activitycoefficient)= 846/(tOC + '273)
Tetrachloroethylene S :I 1
109 8(mm 89) = 6.96 - 14l5/(~oC + ~2l) + 1449/(tOC'+ 273)
109 (Activitycoefficient)= -0.061 + .l449/(tOC + 273) .
Allyl Chloride
8 (~250C) = 20,000 atm
Activity coefficient (25°C)
= 42,500
12-28 2,2-Dich10ropropionic acid
No data
12-29 Phosgene
Hydrolyzes in water
12-30
12-31
13-1
13-2
S = 1
S :I P
S = D
Ethylene dibromide, 1,2-Dibromoethane S = 1
109 H(mm 8g) = 9.044 - 2989/(tOC + 353) + 1024/(tOC + 273)
log (Activitycoefficient)= 1,024/(tOC + 273)
Epichlorohydrin
109 H(mm H9) = 7.487 - 1609/(tOC + 231)
109 (Activitycoefficient)= 586/(tOC +
a-Endosulfan
Data uncertain
Endosu1fan sulfate
No data
E-14
S = 3.
+ 586/(tOC + 273)
273)
S = P
S = 6
~ " .."
-------
13-3
13-4
13-S
13-6
13-7
13-8
13-9
13-10
13-11
13-12
6-Endosulfan
No data
s = P
a-SHC
Data uncertain
s = p
6 -SHC
'Data uncertain
s = p
o -SHC
Data uncertain
s = p
y -SHC
Data uncertain
s = p
Aldrin
H (20° to 2SoC) = lS atm
Activity coefficient (20°C) = 1.84xld9
s = 2
DieldI:in
Data uncertain
S = 2
4,4.'-DDE
H (20° to 2SoC) = 1.3 atm
Activity coefficient (2SoC)
s = 3-P
= 1. Sxl08
4,4'-DDT
H (approx. 2SoC) = 0.8 to 0.03 atm
Activity coefficient (approx. 2SoC)
S = 6
= 1:lxl08 to 3.3xl09
4,4'-DDD
H (2So. to 30°C) = 1.2 atm
Activity coefficient (2SoC)
S = 3-P
= 8.9xl08
E-1S
-------
13-13
Endrin
H (2SoC)
Activity
= 0.039 to 0.021 atm
coefficient (2SoC) = 1.1xl08 to 8.1xl07
13-14 Kelthane
No data
13-1S Naled
No data
13-16 Dichlone
No da ta
13-17 Kepone
No da ta
13-18 Diuron
H (temp. not known) = 0.0013 atm
13-19 Endrin Aldehyde
No ,da ta
13-20 Heptachlor
H (250C) = 82 atm
13-21 Heptachlor Epoxide
H (25°C) = 1.8 atm
13-22
Carbofuran
H (25°C) = 4.6xl0-4 atm
Activity coefficient (25°C) = 1.76xl04
13-23 Mercaptodimethur
No data
E-16
5 = 6
5 = P
5 = P
5 = P
5 = 4-P
5 = 6
5 = P
5 = 1
5 = 3
5 = 6 .
5 = P
. '. ..>
-------
13-24 Chlordane
Data uncertain
S = 2
13-25 Toxaphene
Data uncertain
S = 1
13-26- Captan
"Data uncertain
S = 2-P
13-27 Carbaryl
No data
S = 3-P
13-28 Coumaphos
H (200 to 250C) = 0.0018 atm
13-29 Diazinon
H (20°C) = 0.078 atm
13-30 Dicamba
H (100°C) = 0.027 atm
13-31 Dich10benil
H (200C) = 0.28 atm
13-32 Malathion.
H = 0.0067. atm
S = 6
S = 6
S = 6
S = 6
S = 6
13-33 Methyl parathion
H (20° to 25°C) = 0.0034 to 0.0031 atm
S = 6
13-34 Parathion
H (20°. to 25°C) = 0.034 atm
S =.6
13-35 Guthion
H (20° to 25°C) < 0.21 atm
S = 6
E-17
-------
13-36 Ethion
No data
13-37 Isoprene
No data
13-38 Ch1orpyridos
8 (25° to 35°C) = 0.23 atm
13-39 Cich10rvos
B (20°C) = 0.019 atm
13--40 Ciquat
No data
13-41 Cisa1foton
8 (20° to 23°C) = 0.14 atm
13-42 Mevinphos
No data
13-43 Mexacarbate
8 (139°C) < 0.63 atm
Activity coefficient (139~C) = 4.8x103
13-44 Tri~hlorfon
8 (25°C) = 9.5x10-7 atm
13-45 propargite
No data
13-46 Carbon disulfide
log 8(mm 8g) = 6.942 - ~1169/0:oC + 242)
log (Activitycoefficient)~ 949/(tOC +
E-18
5 = P
5 = 1-P
5 = 6
'5 = 6
. 5 = 6-P
5 = 6
5 = 6-P
5 = 6
5 = 6
5 = P.
5 = 1
+ 949/(tOC + 273)
273)
-------
14-6 Butyric Acid
8 (1000e) = 1 to 2.2 atm
14-7 Formaldehyde
Data uncertain
14-8 Formic Acid
No data
14-9 Fumaric Acid
No data
14-10 Maleic Acid
Data uncertain
14-1
14-2
14-3
14-4
14-5
Aceta.ldehyde
109 8(mm 89) = 7.293 + 0.00751tOe -
109 (Activity coefficient) = 0.4829
S = 2
1071/(tOe + 236)
+ 00751tOe
Acetic Acid
109 8(mm 89) = 7.30 - 1479/(tOe + 217) + 133/(tOe
- 109 (Activitycoefficient)= 133/(tOe + 273)
S = 6
+ 273)
Allyl Alcohol S = 5
109 8(mm 89) = 9.97 - 1272/(tOe + 188) - 600/(tOe + 273)
109 (Activit~oefficient)= 2.63 - 600/(tOe + 273)
iso-Ainyl Acetate
109 8(atm) = 8.702 -
Activity cosfficient
S = 1
2l87/(tOe + 273)
(1000e) = 2730
n-Buty1 Acetate
. 109 8(mm 89) = 7.03 - 1368/(tOe + 204) + 965/(tOe
109 (Activitycoefficient)= 965/(tOe + 273)
S = 1
+ 273)
S = 6
S = 6
S = 6
S = 4-P
S = 4-P
E-19
-------
14-11 Methyl methacrylate
No data
14-12 propionic Acid
See Reference 3
14-13
14-14
14-15
vinyl Acetate
log 8(mm Hg) = 6.992 - 1192/(tOC + 217)
log (Activitycoefficient)= 697/(tOC +
S = P
S = 6
S = 1
+ 697/(tOC + 273)
273)
Adipic Acid S = 6
log 8(mm 8g) = 8.639 - 3074/(tOC + 196) + 787/(tOC + 273)
log (Activitycoefficient)= 787/(to'C +" 273)
Crotononaldehyde
log H(atm) = 5.24 - 1571/(tOC
Activity coefficient (20°C) =
+ 273)
22 to 25
14-16 Acrolein
log H(mm 89) = 8.17 - 1132/(tOC + 228)
. .
log (Activity coefficient) = 1.2634
14-17 Furfural
8 (100°C) = 1.3 to 25 atm
14-18
15-1
15-2
propylene Oxide
109 8(mm 89) = 8.244
Activity coefficient
- 1114/(tOC + 232)
(35° t01QOoC) = 15.3
Methyl Mercaptan
109 8(mm 89) ="7.032 - 1016/(tOc + 239)
log (A c t i v i t yc ° e f f i c i en t) = 604/ ( tOe +
Dodecyl Benzenesulfonic Acid
Ionizes in water, not strippable
E-20
S = 4
S = 2-P
S = 3-P
S = 2
S = 1
+ 604/(tOC + 273)
273)
S = 6
-------
15-3
15-4
15-5
15-6
15-7
Cyc10hexane S = 1
109 8(mm 89) = 6.84 - 1202/(tOC + 223) + 1442/(tOC + 273)
109 (Activitycoefficient)= 1442/(tOC + 273)
Isophorone
109 H(mm H9) = 7.99 - 1672/(tOC + 273)
- 109 (Activity coefficient) = 836(tOC + 273)
Strychnine
No data
2,3,7,8-Tetrachlordibenzo-p-dioxin
No data
Zinc Phenol Sulfonate
Non-volatile, not strippab1e
E-21
S = 4
S = 6
S = 3-P
S = 6
-------
APPENDIX F
SYMBOLS
A partial list of symbols, their common units, and an
indication of where the symbol is 'first used or defined is given
below. Symbols used once only, or"in'a limited context, are
defined where used and not listed here. Cost symbols, defined in
Section 4, are specifically not listed here. In general, any
consistent system of units may be used. Some correlations are for
specific units, however, and these are specified in context.
ap
Superficial surface area, ft2/ft3
A
Stripping medium rate, moles/hr
A,B,C
Antoine coefficients
B
Bottoms rate, moles/L
Cf
packing factor, ft-l
, ,
d,
Tower diameter
D
.,
Diffusivity, ft2/hr or m2/s
D
Overhead liquid rate, moles/hr
EA
Murphree efficiency
corrected for entrainment
F-l
Table "3-4
Eq. (3.1)
Eq. (2.15)
Eq. (3.1)
T~ble 3-4
Eq. (3.29)
Fig. 3-2
Table A-I
-------
EM
Murphree tray efficiency
Eo
Overall tower efficiency
Ep
point efficiency
f
Fractional removal efficiency
f
Fugacity, atm
F
Feed rate, moles/hr
gc
Gravitational conversion constant,
4.18x108 lbm.ft/(lbf.hr2)
G
Gas or vapor rate, lb/hr
G'
Gas or vapor flux, lb/(hr.ft2)
h
Liquid enthalpy, ca1s/mole
H
Henry's Law constant, atm
H
Humidity, moles water/mole dry air
HTU
Height of a transfer unit, ft
i
Air enthalpy, cals/mole dry air
K
vapor-liquid equilibrium constant
L
Total height of tower, f~
L
Liquid rate, moles/hr and mass/hr
n
Tray or stage number
F-2
Sec. 2.4.1
S ec. 2 . 4 . 1
S ec. 2 . 4 . 1
Eq. (3.3)
Eq. (2.1)
Eq.(3.1)
Sec. 3.2.3
Sec. 3.2.3
Eq. (3.22)
Eq. (2.11)
Eq. (3.25)
Eq. (3.30)
. ,
Eq. (3.23)
Eq. (2.11)
Fig. 3-2
-------
N
Total number o'f moles present
N
Total number of stages
NTU
Number of transfer units
p
Partial pressure, atm
*
p
Vapor pressure of pure substance, atm
P
Total pressure, atm
R
Reflux ratio
S
Stripping factor KV/L
t
Temperature, of, °c
T
Temperature, Kelvin
U
Overall heat transfer coefficient
V
Vapor rate, moles/hr
v
Vapor velocity in tower, ft/s
. ,
x
Mole fraction in liquid phase
y
Mole fraction in vapor' phase
y
Mole ratio in vapor
z
Height of packing
£
packing voidage
F-3
Eq. (2.2)
Eq. (3.30)
Eq. (2.3)
Eq. (2.8)
Eq. (2.3)
Table 3-1
'. Eq. (3.12)
Table 3
Eq. (3.1)
Eq. ( 3 . 26')
Eq. (2.6)
Eq. (2.2)
Eq. (3.19)
Eq. (3.30)
. Table 2-1
~
-------
~
Viscosity, Ibm/(hr.ft)
p
Density, Ib/ft3
y
Activity coefficient
Eq. (2.7)
a
Surface tension, dyne/cm
Subscripts ..
B Bottoms
F Feed
G Gas or vapor phase
i, 1, 2 Species
L Liquid phase
m Max !mum
n, 1, 2 Stage number
t Top of tower
V Vapor phase
F-4
------- |