&EPA
United States Industrial Environmental Research EPA-600/7-79-131
Environmental Protection Laboratory June 1979
Agency Research Triangle Park KlC 27711
Energy Conservation
Through Point Source
Recycle with High
Temperature
Hyperfiltration
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-79-131
June 1979
Energy Conservation Through Point
Source Recycle with High
Temperature Hyperfiltration
by
J.L. Gaddis, C.A. Brandon, and J.J. Porter
Clemson University
Department of Mechanical Engineering
Clemson, South Carolina 29631
Grant No. R803875
Program Element No. 1BB610
EPA Project Officer: Max Samfield
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
SUMMARY
Hyperfiltration and ultrafiltration are pressure driven membrane
processes which have been shown to have potential for recycle of water,
energy, and chemicals within textile wet finishing operations. This
project follows two prior projects which showed (1) reusability of
purified (permeate) and residual concentrate streams from hyperfiltra-
tion of spent dye water and (2) general reusability of permeate for a
cross section of textile waste streams. The project began with the
concept that reuse of water, energy, and chemicals would be most
readily adopted if the separations were applied to individual point-
source streams rather than total plant mixed effluents. It was
apparent that the energy value in many spent streams was possibly
sufficient to finance the cost and operation of the required equipment.
In consultation with plant personnel, five wet processes were
selected which were estimated to comprise a large fraction of the
energy and water use of current textile operations. These were
preparation in rope ranges, preparation in open width ranges, dyeing
in continuous thermosol ranges, dyeing in pressure becks, and dyeing
in atmospheric becks. It is estimated that the energy used in these
proceses comprises over half the total of all textile operations.
A team of researchers visited plant sites, received specifications
from plant personnel, measured flow rates and temperatures and using
the data with accepted engineering models calculated the energy use of
the processes. Fundamental calculations of the evaporative heat loss
were made for atmospheric becks and the open washtubs on a range. The
energy fraction drained in effluent water was determined from which
energy recoverable could be estimated. The results, on a basis of
energy consumed per mass of cloth, are given in the following table
together with the percentage estimated to be recoverable using membranes
for water recovery.
Rope preparation' 4337 kJ/kg (1860 Btu/lb) 61% recoverable
Open-width preparation 9300 kJ/kg (3990 Btu/lb) 62% recoverable
Continuous Dyeing 3.95 x 10^ kJ/kg 62% recoverable
(wet portions only) (1690 Btu/lb)
Atmospheric beck 4.47 x 104 kJ/kg 45% recoverable
dyeing (19170 Btu/lb)
High pressure beck 1.15 x lO1* kJ/kg 34% recoverable
dyeing (4930 Btu/lb)
Low pressure beck 4.79 x 103 kJ/kg 41% recoverable
dyeing (2054 Btu/lb)
ii
-------
Samples were taken of process effluent at each drain or of each
drained fluid in a sequential process. Analysis of samples was used
as a basis for decision on which process (UF or HF) to apply, and
for guidance as to the conditions of operation: A drum size sample of
each waste was shipped to the laboratory for screening tests to
ascertain rejection and cccnpatability of the membrane. These tests
provided steering for the subsequent field tests.
Two small equipment skids were fabricated to allow membrane operation
on fresh feed in the field. The hyperfiltration membrane modules were
applied to atmospheric dye waste, dye fluid, and soaper fluid from
continuous dyeing, and water desize, chemical desize, and caustic
drains of the open width range. The ultrafiltration modules were
applied to dye fluid and soaper fluid from continuous dyeing, to
water wash in open width preparation, and to chemical desize wash
in open width preparation.
The permeate water in each case was reported by plant personnel
to be reusable. In the atmospheric dye becks and the dye range the
residual fluid (membrane concentrate) showed some promise for reuse.
Reuse of residual on the preparation range was not evaluated through
water desize wash concentrate (PVA solutions) is being recycled
in several locations in the industry. Thus the applicability of
membranes for at least water recycle (with energy) has been demonstrated.
The long term flow properties as estimated from field data has been
used to design (estimate the size of) prototype units. These designs
form the basis for capital and operating cost estimates. Using a
common economic basis for evaluating each system and a six year
amortization of capital (selected arbitrarily), a simple total cost
estimate has been formulated. Two of the fluids show negative costs
(i.e., positive return) - the caustic washer of the preparation range
and the soaper washer of the dye range. Small costs (net) are
projected for chemical desize washer (preparation range) and larger
costs for the water desize washer (preparation range) and the dye
washer of the dye range. The atmospheric beck dye fluid could be
operated at a cost savings or loss depending on whether a technical
problem concerning an unusually large flux loss can be solved.
The impact of the environmental problem by each of the above mem-
brane applications is (a) in concentrate reuse situations the individual
contribution of the pollutants in the total stream is eliminated from the
plant's discharge, (b) in all situations the waste hydraulic loading is
lessened, reducing costs or promoting efficiency, or (c) process modifica-
tions may be developed which reduce chemical use or promote the use of
superior products.
ill
-------
Several process modifications have been recognized which will
enhance the membrane use through allowing lower costs of installation
or will allow alternative processes or chemicals to be used thus
reducing operating costs.
All processes studied have applicable modifications which appear
to have some promise. Two such modifications have been demonstrated
in full scale. The addition of dye in atmospheric beck dyeing to an
already hot solution with the cloth is required to allow energy reuse.
Through careful application of dye in a uniform and deliberate manner,
a test dyeing was performed without the feared streaking effect. The
open width preparation range was changed from high flow to lower flow
by increasing the operating temperature. A net energy use decrease
resulted and the effect on the membrane system cost projection is
significant as well. The realization of the usefulness of superior
washing to yield a lower flow requirement prompted a brief investigation
of the washer equipment market and technology.
The incorporation of membranes for water recycle is viewed as
promising. The energy benefit alone will pay for a substantial
portion of the total cost. Membrane application to textile streams
appears to be broadly feasible leading to a reduction in effluent
volume of an estimated 38 percent based on the processes of this study
and 90 percent or more if fully implemented. Thus membrane application
allows a substantial step toward closed cycle operation. The interests
of the EPA are served best by such closed cycle operation and the
disposition of concentrate remains the principal hindrance to achieve-
ment of the zero discharge of liquid effluent.
This report was submitted in fulfillment of Grant R803875 by
Clemson University under the sponsorship of the U. S. Environmental
Protection Agency. This report covers a period from July 1975 to
December 1978 and work was completed as of March 1979.
4v
-------
CONTENTS
Summary
Figures vii
Tables ix
Units and Conversions x
Acknowledgments xi
Introduction j_
Conclusions and Recommendations 3
1.0 ENERGY AND WATER CONSUMPTION OF SELECTED PROCESSES 5
1.1 Atmospheric Beck Process 5
1.1.1 Process Description 5
1.1.2 Chemicals Used 7
1.1.3 Energy and Water Use 9
1.2 Continuous Preparation Range 9
1.2.1 Process Description 9
1.2.2 Chemicals Used n
1.2.3 Energy and Water Consumption 13
1.3 Continuous Dye Range 15
1.3.1 Process Description 15
1.3.2 Chemicals Used 15
1.3.3 Energy and Water Reusage 19
1.4 Pressure Becks 21
1.4.1 Process Description 21
1.4.2 Chemicals Used 23
1.4.3 Energy and Water Consumption 23
1.5 Continuous Rope Preparation Range 26
1.5.1 Process Description 26
1.5.2 Chemicals Used 26
1.5.3 Energy and Water Consumption 26
2.0 CHARACTERIZATION OF WASTEWATER DISCHARGES 31
2.1 Atmospheric Beck 31
2.2 Pressure Beck 31
2.3 Open-width Preparation Range 31
2.4 Continuous Dye Range 31
2.5 Rope Preparation Range 37
3.0 POSSIBLE PROCESS MODIFICATIONS 38
3.1 Atmospheric Dyeing 33
3.2 Continuous Dyeing 39
3-3 Continuous Rope Preparation Range 41
3.4 Pressure Beck Dyeing 41
3-5 Open-width Preparation Range 43
-------
4.0 ENERGY SAVINGS POTENTIAL THROUGH RECYCLE 44
4.1 Atmospheric Beck 44
4.1.1 Energy Recovery by Plant-scale Heat Exchangers 44
4.1.2 Energy Recovery with Filtration Devices 47
4.1.2.1 Plant-scale Filtration Application 47
4.1.2.2 Filtration Applied to Specific Processes 47
4.2 Pressure Beck 52
4.3 Dye Range and Preparation Range 53
4.4 Summary of Energy Recovery Estimates 53
5.0 SCREENING TESTS 55
5.1 Facility 55
5.2 Procedure 55
6.0 FIELD TESTING 66
6.1 Operation of Atmospheric Beck Dye 71
6.2 Operation on Dye Range 83
6.3 Operation on Preparation Range Fluids 93
6.4 Ultraf iltration on Dye Range Fluids" 103
7.0 ECONOMIC ESTIMATES HI
7.1 Individual Designs HI
7.1.1 Water Washer on Preparation Range 111
7.1.2 Chemical Desize Washer 112
7.1.3 Caustic Washer 113
7.1.4 Dye Washer #1 113
7.1.5 Soaper Washer on Dye Range 113
7.1.6 Atmospheric Dye Becks 114
7.2 Costs of Systems 114
7.3 Results of Economic Survey 115
7.4 Economic Impact on Industry 119
8.0 FULL SCALE PROCESS MODIFICATIONS 122
8.1 High Temperature Dye Addition Experiment 122
8.2 Open-width Preparation Range 123
APPENDICES
A. Fundamental Evaluation of Evaporative and Convective
Loss in Full-Scale Textile Equipment 125
B. Calculation of External Radiative and Convective
Heat Losses in Textile Equipment 133
C. Energy Use in Atmospheric Beck Processes 135
D. Energy Use Analysis of a Continuous Dye Range 150
E. Analysis of Mass Removal from Cloth in Preparation 161
F. Optimization Computer Program 162
G. Textile Washing Study 168
References 171
VI
-------
List of Figures
Figure 1 Schematic Diagram of an Atmospheric Beck
Figure 2 Typical Dye Cycle with Scour for Atmospheric
Beck Dyeing
Figure 3 Schematic Diagram of Open Width Preparation Range
Figure 4 Schematic Diagram of Continuous Dye Range Process
Figure 5 Energy Disposition on Continuous Dye Range
Figure 6 Cycle Procedures for Low Pressure Beck
Figure 7 Cycle Procedures for High Pressure Beck
Fgiure 8 Energy Consumption for Pressure Becks
Figure 9 Schematic Diagram of Rope Preparation Range
Figure 10 Distribution of Energy on Rope Preparation Range
Figure 11 Implementation of Process Modifications for
Application to Dye Range
Figure 12 Caustic Washer Section Process Modification
Figure 13 Energy Recovery by Heat Exchangers - Atmospheric
Becks
Figure 14 Comparison of Energy Recovery with Drop-fill
and Overflow Washing
Figure 15 Energy Recovery Related to Water Recovery
Figure 16 Schematic Loops in Mobile Hyperfiltration
Laboratory
Figure 17 Screening Tests - Ultrafiltration on Open Width
Preparation Range
Figure 18 Screening Tests - Hyperfiltration on Open Width
Preparation Range
Figure 19 Screening Tests - Hyperfiltration on Atmospheric
Dye Beck
Figure 20 Screening Tests - Rope Preparation Range
Figure 21 Screening Tests - Hyperfiltration on Dye Range
Figure 22 Flow Schematic of Skid Mounted Unit
Figure 23 Ceramic Tube Module Arrangement
Figure 24 Stainless Steel Tube Module
Figure 25 Ultrafiltration Skid Schematic Flow Diagram
Figure 26 Operating Conditions History - Atmospheric Dye Beck
Figure 27 Permeate Flow Results - Atmospheric Dye Beck
Figure 28 Results of Flux Decline Study with Atmospheric Dye
Beck Fluid
Figure 29 Copper and Chromium Effect on Flux Behavior
Figure 30 Field Results - Hyperfiltration on Dye Washer
Figure 31 Copper and Chromium Comparison to Flux
Figure 32 Membrane Exposure - Soaper Feed on Dye Range
Figure 33 Flux and Temperature - Soaper Feed on Dye Range
8
12
18
22
24
25
27
28
30
40
42
46
49
51
56
58
59
60
61
62
67
69
70
72
74
75
81
82
86
87
89
90
Vii
-------
Figure 34 Flux Results - Soaper Feed on Dye Range 91
Figure 35 Field Results - Hyperfiltration on Preparation
Range Water Washer - High Concentration 94
Figure 36 Field Results - Hyperfiltration of Preparation
Range Water Washer - Moderate Concentration 95
Figure 37 Field Results - Hyperfiltration of Chemical
Desize Washer Fluid 97
Figure 38 Chemical Desize Fluid - Flux versus Solids 98
Figure 39 Chemical Desize Fluid - Flux versus Pressure and
Velocity 99
Figure 40 Caustic Washer Fluid - Parametric Flux Dependence 101
Figure 41 Field Results - Ultrafiltration on Dye Washer Fluid 104
Figure 42 Relative Light Transmission - Membranes on Dye Washer 105
Figure 43 Field Results - Ultrafiltration on Soaper Washer
Dye Range Fluid 106
Figure 44 Field Results - Ultrafiltration of Water Desize
Fluid 108
Figure 45 Field Results - Ultrafiltration of Chemical Desize
Fluid 109
viii
-------
List of Tables
Table 1
Table 2
Table 3
Table 4
Table 5
Table 6
Table 7
Table 8
Table 9
Table 10
Table 11
Table 12
Table 13
Table 14
Table 15
Table 16
Table 17
Table 18
Table 19
Table 20
Table 21
Table 22
Table 23
Table 24
Table 25
Table 26
Table 27
Table 28
Table 29
Table 30
Table 31
Analysis of Samples
Analysis of Samples
Analysis of Samples
Analysis of Samples
Typical Dye Cycle Details - Atmospheric Beck 10
Evaporative Heat Loss Rates for Open Width
Preparation Range 13
Energy Losses by Radiation and Convection 14
Energy Loss to Drain 15
Energy Summary - Open Width Preparation Range 15
Chemical Utilization and Prospects for Reclamation 16
Evaporative Loss Estimate 20
Energy Consumption for Dye Range 21
Analysis of Samples - Atmospheric Beck 32
Atmospheric Beck 33
Pressure Becks 34
Open Width Preparation Range 35
Continuous Dye Range 36
Analysis of Drains in Rope Preparation Range 37
Energy Recovery Estimates for Pressure Beck Dyeing 52
Potential for Energy Savings by Water Reuse - Summary 54
Summary of Tests Undertaken 63
Membrane Operation on Dye Waste from Atmospheric Beck 73
Analysis of Samples from Hyperfiltration of Dye Beck
Fluid 77
Dye Additions for Reuse Tests 83
Summary of Testing of Hyperfiltration on Dye Range 84
Description of Dye Laboratory Reuse Testing 92
Industrial Plant Analysis of Spot Samples
(Caustic Washer) 102
Total Solids Rejection on Dye Range 103
Preparation Range Water Washer: Comparison of
Ultrafiltration Permeates 107
Results of Economic Analysis for Preparation Range 116
Results of Economic Analysis for Dye Range 117
Results of Economic Analysis for Atmospheric Dye Becks 118
Energy and Number of Selected Processes 120
Impact on Industry of Full Membrane Implementation 121
Flow and Temperature Modifications 124
ix
-------
Units and Conversions
Multiply English Unit
Btu
ft
ft2
gal
fl ounce
pound (mass)
psi
gal/min
gal/ft2-day (GFD)
ohm'1
F minus 32
by
1.0587 E+3
3.048 E-l
9.290 E-2
3.785 E-3
2.957 E-5
4.536 E-l
6.895 E+3
6.308 E-5
4.716 E-7
1.0
5/9
To Obtain Metric Quantity
J (joule)
m (meter)
m
kg (kilogram)
N/m2 (Pascal)
m3/s
m/s
S/( Siemens)
C (-K - 237.16)
Metric prefixes used
G
M
k
d
c
m
P
(giga)
(mega)
(kilo)
(deci)
(centi)
(milli)
(micro)
~
=
=
=
—
=
=
X
X
X
X
X
X
X
103
106
103
10~ l
ID'2
ID'2
10"6
Thus the liter (= m3/1000) is properly written as dm3
The microm (10~6m) is properly ym (micrometer)
-------
ACKNOWLEDGMENTS
Four industrial plants contributed to the effort reported herein
by supplying information, opening their plants for surveys/ performing
chemical and reuse analysis, allowing the operation of equipment at
their expense, and other varied contributions. Their contributions
are gratefully acknowledged. The plants are:
Graniteville Company, Gregg Division
M. Lowenstein & Sons, Lyman Printing and Finishing Company
Riegel Textile Corporation, La France Industries
Springs Mills, Grace Bleachery
Contributing significantly to portions of the effort were Dr.
A. C. Elrod and Dr. P. J. Bishop, both Clemson University faculty.
This effort began through the interest and activity of T. N.
Sargeant who was initially Project Officer, but who was transferred by
EPA during the effort. The present Project Officer, Dr. Max Samfield,
is acknowledged for his special effort required to carry through work
in progress. His patience and continued support have been appreciated.
The extramural support of EPA-IERL in Cincinnati is acknowledged
in the persons of Dr. Harry Bostian and Bob Mournighan. Each has
played an active role in the execution of this grant.
xi
-------
INTRODUCTION
Hyperfiltration and ultrafiltration are membrane separation
processes which produce a large stream dilute in solute and a small
stream enriched in solute. Both types of membranes separate large
molecular solutes and colloids; while only hyperfiltration is
effective in separating small molecules. In this case, solutes
include spent dyes, warp size, waxes and oils, various auxiliary
chemicals and a host of contaminants. The membrane processes do not
eliminate chemical discharge; rather, they provide a dilute (permeate)
stream and a concentrate (residual) stream. The permeate was shown in
two previous studies1'2 to satisfy industrial water requirements. Further,
in the first of these studies the concentrate was reused in commercial
quality dyeing of various shades and a net decrease in dye use was
projected.
In the previous studies mentioned, the membrane separation was
applied to mixed effluent from the plant. It was apparent3 that
processing of the high temperature segregated streams would allow
energy recovery and encourage the reuse of concentrate. In the many
near-boiling temperature streams found in the textile industry, the
energy value of the water is approximately ten times the value of the
water.
Thus, the present program was conceived to identify candidate
processes having high temperature streams and to study the applicability
of membrane processes. The objective was to study the energy consumption
of the selected processes and characterize the effluent from each
appropriate source. Based on experience and screening tests of the
fluids, a processing strategy was devised to test the best membrane/fluid
combinations in the field for a period sufficient to evaluate long term
effects on fresh fluid. Using the field data, economic studies were made
1"Hyperfiltration for Renovation of Textile Finishing Plant
Wastewater," by Craig A. Brandon and John J. Porter, EPA-600-2-76-060,
March, 1976.
2"Hyperfiltration for Renovation of Composite Wastewater at Eight
Textile Finishing Plants," by Craig A. Brandon, John J. Porter, and
Donald K. Todd, Final Report, Grant S802973.
3"Application of High Temperature Hyperfiltration to Unit Textile
Process for Direct Recycle," by Craig A. Brandon and Max Samfield,
"Membranes: Desalination and Wastewater Treatment" conference, Jeru-
salem, 1978.
-------
to estimate the net cost of application of membranes for industrial
use.
Supporting the entire effort was an emphasis on process modifica-
tions. Some modifications result directly from recycle, i.e.,
recycled water is hot compared to water added cold and then heated.
Other modifications, to washing processes particularly, include trade-
offs between flow and operating temperature to decrease membrane costs
or permit the use of alternative chemical formulations.
-------
CONCLUSIONS AND RECOMMENDATIONS
1. A survey of energy requirements and chemical discharge of the
individual fluid streams from five common textile processes has been
conducted. The aggregate of all such streams is estimated to comprise
over half the energy utilization of the textile finishing operation.
2. It is estimated that approximately one half of the energy
used in the processes studied can be effectively recovered by hyper-
filtration or ultrafiltration if permeate recycle is adopted.
3. Screening tests were run on all but one fluid stream, followed
by field tests on six fluid streams. It is considered that the long
term effects were observed sufficiently to estimate costs on a reasonable
basis.
4. All permeate water was determined reusable by the textile plant
laboratory personnel. Simulation or estimation of build up effects was
not attempted for the evaluation. Some of the concentrate streams were
determined valuable for reuse.
5. A number of process modifications a) to permit membrane
recovery processing, b) to reduce membrane equipment costs, and c) to
allow chemical cost reductions were proposed.
6. TWO full scale process modifications were demonstrated and are
documented. Both were successful. In one, dyes were added at high
temperatures without streaking. In the other, water use was diminished
by operating at increased temperature - with a net energy decrease.
7. Based on membrane permeability observed in field testing,
hypothetical designs were made and the respective costs estimated.
Net operating costs and benefits were calculated. On the basis of
six year amortization and without credit for chemical recovery, three
of the six streams could offer payback potential on energy saved. The
other three would be operated at a deficit. The average of the six
streams is near a balance.
8. There is a general reluctance in management to adopt the
risk associated with capital investment in the new membrane technology
for the modest payback it affords.
9. There is a greater reluctance to consider the more difficult
reuse of concentrated spent chemicals. Concentrate reuse will either
-------
be adopted for its own value, as in the case of size recovery, or will
be adopted after a permeate recovery unit is installed and the plant's
laboratory further studies the concentrate for reuse. This is to say
that management will probably not decide to install equipment if the
payback depends on both permeate and concentrate energy recovery.
10. A membrane system may be considered on its merit as an energy
recovery device. If the installation is to significantly reduce
pollution (other than reducing total flow) a means of concentrate
disposal must be devised. Since there is reluctance to vigorously
pursue concentrate reuse, the technology of concentrate disposal
should be studied by EPA. If such technology is adopted, the plants
will, in time, learn how to "mine" their concentrated waste streams
for raw materials and in the meantime pollution will be reduced by
the action of the concentrate disposal.
11. Membrane processing may be particularly attractive for
toxic laden streams. The concentrate from such streams will require
special attention.
12. Depending on economic factors and EPA philosophy, a number
of demonstration projects should be considered. By absorbing some
of the risk, EPA can accelerate the application of the technology
and stimulate the growth and interest of membrane manufacture and
application.
-------
1.0 ENERGY AND WATER CONSUMPTION OF SELECTED PROCESSES
In consultation with members of the South Carolina Textile
Manufacturers Association and the Environmental Protection Agency
five manufacturing processes were selected to include (1) as many
as possible of the effluent guideline categories, (2) as many
as possible of the wet manufacturing processes with "hot" wastewater,
and (3) at least one case each of batch and continuous processing.
The following five processes were selected for this study: (1)
atmospheric beck process, (2) continuous preparation range, (3)
pressure beck process, (4) continuous dye range, and (5) continuous
(rope) preparation range. Four South Carolina industrial partners
were involved: La France Industries, Lyman Printing and Finishing,
Gregg Division of Graniteville, and Grace Bleachery of Springs Mills.
1.1 Atmospheric Beck Process
1.1.1 Process Description
Atmospheric beck dyeing is accomplished by exposing the fabric
to a hot (82 to 96 C) dye solution for a sufficient length of time
to achieve the desired shade of color. The procedure for dyeing a
typical lot of cloth is as follows:
(1) Cold water is admitted to the dye beck (see Figure 1)
and is heated to about 32 C.
(2) The cloth is loaded into the beck.
(3) Scouring (cleaning) chemicals are added to the water and,
while the cloth is circulated through this solution, the
solution temperature is increased to, and maintained at,
about 71 C for about one-half hour.
(4) The scouring chemicals are flushed out of the beck and
rinsed from the cloth. (This requires a large amount of
cold water.)
(5) Fresh water again fills the beck and is heated to 32 C.
(6) The dye and auxiliary chemicals are added to the water.
(7) The solution is heated and maintained at about 87 C for
sufficient time to color the cloth to the desired shade.
(8) The dye solution is flushed out of the beck and rinsed
from the cloth. (This, again, requires much cold water.)
(9) Cold water is admitted to the beck again, and is heated
to 32 C.
(10) Chemicals to make the dye "color fast" are added to the
water.
-------
DOOR
COLD
WATER
SUPPLY
CLOTH
BECK VENT
THROUGH ROOF
DAMPER
ROTATING
FRAME
DOOR
VERFLOW
WATER LEVEL
TO DRAIN
STEAM
DRAIN PORT
FIGURE 1 SCHEMATIC DIAGRAM OF AN ATMOSPHERIC BECK
6
-------
(11) This solution is heated to about 50 C and maintained at
that temperature for about fifteen minutes.
(12) These dye "fixing" chemicals are flushed from the beck
and carefully rinsed from the cloth. (Again, much cold
water is used.)
(13) The cloth is unloaded from the dye beck and sent to a dryer.
Because different materials and dyes may be involved, certain
procedures will vary from this thirteen-step generalization. However,
all procedures are similar. Figure 2 plots temperature versus time
for a typical cycle with scour.
The observed plant had 32 dye becks installed, ranging in width
from 0.91 to 3.66m. These becks were operated randomly around the
clock five days of each week. Normal operation had these becks dyeing
36 lots of cloth daily, with (typically) three becks undergoing
maintenance at any one time. Then, on the average, each beck dyes
36/29 (or 1.24) lots of cloth each day. The length of time that a
beck is kept at high temperature varies markedly from one lot of cloth
to the next, but 6.6 hours is average. When one adds the time required
for loading the beck with water and with cloth, time for controlled
gradual heating up and cooling down, time for rinsing the cloth, time
for sampling to determine whether the cloth has been dyed to the proper
shade of color, and time for unloading the material from the beck, it
is understandable that each beck averages dyeing only 1.24 lots of
cloth daily.
1.1.2
Chemicals Used
Many various chemicals are used in the dyeing and scouring
processes of the dyehouse. A sample of these chemicals is given
below for three arbitrarily selected processes:
Dye and Scour Chemicals
Process 2 - Formula 8545
Dye
Salt
Fix
Process 11 - Formula 1191
Dye Auxiliary
Dye
Salt
Fix
Process 12 - Formula 1197
Scour Chemicals
Scour Chemicals
kg mass of chemical used
1000 kg of cloth processed
2.160 Dye Mixture
150.000 Salt Liquid
5.000 Acetic Acid
20.000 Quadye NT
27.400 Dye Mixture
150.000 Salt Liquid
20.000 Fix GD
120.000
10.000
Peroxide
Caustic Soda
-------
DYE
CYCLE EVENTS
3 4
DYE TIME (HR)
FIGURE 2 TYPICAL DYE CYCLE WITH SCOUR FOR ATMOSPHERIC BECK DYEING
-------
Dye and Scour Chemicals
Process 12 (continued)
Scour Chemicals
Dye Auxiliary
Dye Auxiliary
Dye
Salt
Fix
kg mass of chemical used
1000 kg""of" "cloth processed
10.000 Sodium Bisulfite
20.000 Quadye NT
2.500 Leveler 37-5
0.826 Dye Mixture
50.000 Glauber Salt
20.000 Fix GD
1.1.3
Energy and Water Use
Energy and water use in the dyehouse was studied to determine
the potential for reclamation. During the course of this study two
different methods were employed to calculate the energy and water
consumed per kilogram of cloth processed. In the first method, an
energy and water mass balance is made on the dyehouse without
reference to the individual dyeing cycles. The dyehouse is taken
as an entity and the total water and energy usage is calculated
from the inlet and discharge water flow rates and temperatures and
the steam consumption in the dyehouse. To better understand the
distribution of energy, a second method is based on the analysis
of selected hypothetical dye cycles to determine the energy and water
consumed per kilogram of cloth processed during each cycle. Three
cycles, two representing the average cycle (with or without scour)
and the other a typical cycle (with scour), were studied through
this method. Table 1 shows the summary of calculations for a single,
typical dye cycle for illustrative purposes.
The results obtained by the two methods are given below. The
details of the averaging of hypothetical cycles are given in Appendix
C.
Energy consumption per kg Water usage per kg of
of cloth processed (kJ) cloth processes (m3)
Method I (overall
balance)
Method II (average
hypothetical cycle)
4.47 x 1(T
(19,170 Btu/lb)
4.57 x 104
(19,600 Btu/lb)
0.447 (53.7 gal/lb)
0.288 (34.6 gal/lb)
The calculations by each method agree well for energy but the
hypothetical cycles fail to fully account for the water usage
presumed to occur during sustained rinses.
1.2.1
Continuous Preparation Range
Process Description
-------
TABLE 1. TYPICAL DYE CYCLE DETAILS - ATMOSPHERIC BECK
Event
Duration
Temperature
Energy consumption
to heat from 18 C
to operating temp
Energy consumption
to maintain at
operating temp
Total energy
consumption
Water used
min
°C
GJ
GJ
GJ
m3
1 & 2
91.5
46
0.629
0.203
0.832
5.65
(hot)
3
37.5
74
0.658
0.303
0.961
0.27
(hot)
4 5
30.0 99.0
52
0.755
0.292
1.047
13.32 5.7
(cold) (hot)
6 &7
522.0
98
1.103
12.28
13.38
0.47
(hot)
8 & 9 10 & 11
45.0 21.0
52
0.755
.061
.816
26.1 5.7
(cold) (hot)
12
42.0
_
_
_
_
23.5
(cold)
Cumulative energy = 17.0 GJ (16 x 106 Btu)
Cumulative water = 80.7 m3 (21 kgal)
Beck volume (cold) = 5.40 m3 (1428 gal)
297 kg of cloth processed
Energy = 57200 kJ/kg cloth (24,500 Btu/lb)
Water = 0.271 m3/kg cloth (32.6 gal/lb)
-------
The continuous preparation range prepares the cloth for subse-
quent finishing, i.e., dyeing or printing, through a series of
chemical baths followed by high-temperature washing to "clean" the
cloth. On the range observed conditioned cloth (greige cloth) from
the heat-set oven initially enters a pair of Gaston County alterna-
tors that wash the cloth at 60 C with a flow rate of 5.14 dm3/s
(35 gal/min) to the drain on each washer. Then the cloth proceeds
to a 60 C chemical bath, the desize saturator. After exposure
to a steamer environment the cloth then enters the desize washers for
a hot rinse. The three desize washers were operating at 82 c, and each
washer has approximately 1.05 dm3/s (17 gal/min) of fresh water run
in at the nip sprays. The washers are counterflowed back toward the
desize saturator before going to the drain. The cloth then proceeds
to the second chemical bath (two caustic saturators, running at 77 C).
This is followed by a steamer and J-box. Subsequently, the cloth
enters the caustic washers for a second high-temperature bath. The
caustic washers were operated at 82 C and counterflow back toward the
caustic steamer. Each washer has nip sprays feeding the washers at
approximately 1.05 dm3/s (17 gal/min) on each washer. There are
four washers in this section gravity counterflowing to a single drain.
Then the cloth proceeds to the third and last chemical bath (the
peroxide saturator) running at 46 c. After a steamer and J-box
the cloth then enters the peroxide washers that were operating at 76 C
for the last high-temperature bath. Upon exit it is ready for subsequent
finishing. The four peroxide washers are counterflowed back toward the
peroxide J-box, before going to the drain. The washers were fed by
nip sprays at 0.79 dm3/s (12.5 gal/min) on each washer. A schematic
diagram of the flow arrangement on the continuous preparation range
is given in Figure 3.
1.2.2
Chemicals Used
The formulations of the chemical baths used in the continuous
preparation range are given below:
Desize Saturator (1)
Caustic Saturator (2)
Peroxide Saturator (1)
0.5% Albone D.S.
2.0% NaOH
1.2% Superterge LRC-37
1.8% Xylol
3.6% NaOH
1.2% Superterge LRC-37
2.5% Hydrogen Peroxide
1.0% Sodium Silicate
0.5% NaOH
0.1% Superterge LRC-37
0.1% Amquest 120-CF (chelate)
(stabilized peroxide)
(detergent)
(50%)
11
-------
DES.ZE
SATURATOR \ SATURATOR
WATER WASHERS \ CHEMICAL DESIZE
WASHERS
CLOTH
TEAMER
4.41 dn?/$
60 C
1
t
I
f
CONDENSATE
J
J
J*BOX
3.15 dirt3*
82C
CONOENSATE
CONTINUED
BELOW
FROM
ABOVE
PEROXIDE
SATURATOR
CAUSTIC WASHERS
4.41
82 C
/
y
PEROXIDE WASHERS
1
it
t
*
f
•
f
^
I
3.15
76 C
FIGURE 3 SCHEMATIC DIAGRAM OF OPEN WIDTH PREPARATION RANGE
-------
1.2.3
Energy and Water Consumption
An energy and water use study was made on the continuous
preparation range to determine the energy and water consumption.
The temperature data used in these calculations were obtained on
January 30, 1976, (but the flows were determined over a year later
after installation of instruments). Figure 3 presents the data used
for analysis. Incoming water temperature was assumed to be 18 C
which is considered to be the annual average.
The latent heat loss rates, qT, were read from the curve of
Figure A-5 corresponding to the appropriate operating temperatures
for the wash tubs and saturators. This information along with
estimated surface areas involved is given in Table 2. The net
latent heat loss per unit area, for 9.11 m2 of estimated cloth and
water surface, is scaled to the loss rate for the 12.5 m2 of the test
washer. The latent heat loss, qiaf was then calculated for the
observed temperatures by
•Hat- Ifig A-5 (9.H/12.5)
Wash Box
GC #1
GC #2
SUBTOTAL
Desize Saturator
Desize Washer #1
Desize Washer #2
Desize Washer #3
SUBTOTAL
Caustic Saturator
Caustic Washer #1
Caustic Washer #2
Caustic Washer #3
Caustic Washer #4
SUBTOTAL
Peroxide Saturator
Peroxide Washer #1
Peroxide Washer #2
Peroxide Washer #3
Peroxide Washer #4
SUBTOTAL
TOTAL
Surface Areaa
(m2)
9.2
9.2
9.2
9.2
9.2
9.2
9.2
9.2
9.2
9.2
9.2
9.2
9.2
9.2
9.2
9.2
Temperature
(°C)
60
60
16
87
84
82
82
49
67
73
82
49
89
90
82
76
Heat Loss
(GJ/hr)
0.108
0.108
0.216
0.108
0.354
0.319
0.281
1.062
0.288
0.066
0.153
0.200
0.288
0.995
0.066
0.384
0.392
0.393
0.219
1.344
3.62
aEach individual wash box has about 1.67 m2 of water surface. The
effective surface for evaporation is based on 49 linear meters of
exDOsed clotH —
13
-------
An estimate of the magnitude of convective and radiative heat
transfer for the dye range was obtained using the equation
q = h A AT.
The combined heat transfer coefficient, h, for radiation and con-
vection was, from Appendix B, H.3W/m2-K (2 Btu/hr/ft2/F). The
environment temperature used was 29 C.
The energy loss by radiation and convection is given in Table 3
for the wash tubs and saturators for ideal and operating temperatures.
The total energy loss by radiation and convection for observed
operating temperatures was 0.534 GJ/hr.
TABLE 3. ENERGY LOSSES BY RADIATION AND CONVECTION
GC Alternator
Desize Saturator
Steamer
Desize Washer
Caustic Saturator
Steamer
Caustic Washer
Peroxide Saturator
Peroxide Washer
TOTAL
Area
-------
Wash Tub
GC Alternator
Desize Washer
Caustic Washer
Peroxide Washer
TOTAL
TABLE 4 .
Flow Rate
(dm3/s)
4.41
3.15
4.41
3.15
ENERGY LOSS TO DRAIN
Temperature
(°C)
60
82
82
76
Energy Loss
(GJ/hr)
2.77
3.02
4.26
2.72
12.7
(12 x 106 Btu
/hr)
TABLE 5. ENERGY SUMMARY - OPEN WIDTH PREPARATION RANGE
Evaporation Radiation and
Item Loss Convection. Loss
GC Alternator
Desize Section
Caustic Section
Bleach Section
TOTAL
FRACTION OF TOTAL
all
0.21
1.06
1.00
1.34
3.61
21%
values in GJ/hr
0.026
0.149
0.289
0.069
0.533
3%
Energy
To Drain
2.77
3.02
4.26
2.51
12.56
76%
(16 x 106
Total
3.006
4.229
5.549
3.919
16.703
Btu/hr)
1.3 Continuous Dye Range
1.3.1 Process Description
The range observed in this study operates on cotton and cotton/
synthetic blended fabrics. It is designed to process over 100,000
yards of fabric at a rate of 50-100 yards per minute. Other portions
of the plant utilize continuous scouring, desizing, bleaching, and
finishing equipment in addition to the continuous dye range being
studied in this project.
The continuous dye process generally starts with fabric which
has been heat-set, desized, scoured, bleached and dryed. The dry
fabric is first passed into a dye solution and through a nip roll
at the head end of the range where dye and chemicals are padded
onto the fabric. The fabric is then dryed carefully with controlled
infrared predryers and dry cans to prevent the migration of the dyes
on the fabric. If the fabric contains polyester fiber the dye
solution will contain dispersed dyes which must be fixed onto the
polyester fiber after it is dryed. This is done in a thermosol
oven at 204 to 215 C. The fabric is passed over rollers in this oven
to give a total dwell time of 60 to 90 seconds. After leaving the
oven the fabric passes over cooling cans to cool the fabric before
it enters a chemical solution (caustic and hydro) to (chemically)
reduce the vat dye already present on the fabric. A nip roll at
the outlet side of the padder removes excess liquor from the fabric
15
-------
and it then enters a steamer at 104 C. The steamer provides the
time, temperature, and moisture that is needed for the reduced vat
dye to penetrate the cotton fiber. After leaving the steamer the
fabric is passed through four rinse boxes to remove caustic and
other chemicals from the fabric. The washing is not sufficient
to remove the reduced vat dye from the cotton fiber but acts to
remove caustic and reduction chemicals from the fiber before
oxidation. This washing process increases cleanliness and reduces
the chemical flow required in a bath which is applied next contain-
ing an oxidation agent and acid.
The oxidation may be performed with many chemicals; in the
case under study potassium iodate is used with acetic acid. This
solution completes the oxidation of the leuco vat dye to the in-
soluble form. The fabric then passes into a soap solution to aid
in the crystallation of the vat dye to give the desired color
and remove surface dye and chemicals from the fabric.
After the soap solution the fabric passes through three wash
boxes which are used to completely rinse the fabric and adjust the
final pH to 6-7 if the fabric is to be given a resin treatment. The
fabric is then dryed and is ready for finishing if desired.
1.3.2 Chemicals Used
The chemicals utilized in the dye processes are shown in
Table 6. Those judged possible to reclaim are designated therein.
TABTiF 6. CHEMICAL UTILIZATION AND PROSPECTS FOR RECLAMATION
J-—j- •'-••'- •• " ' •••_ jJ^TrS -^t=-*—•— -j*r • •—»- ••- ;—' • ~~ "^•'l ' TiffiiffiJ"Sg i i if ;T j i iffli i • 1 i. nF-j^MyrftigLTrij, I f ^TjT \ ij* **"*• ":fj^ T^ "T ! I" ^™ "•"J'TBBB
(A) Chemicals that are used for the continuous dye range
Dyes Chemicals
dispersed sodium hydrosulfite
vat caustic
sulfur salt
reactive antimigrants
wetting agents - soap
acetic acid
buffering chemicals
oxiding agent (periodate)
(B) Chemicals that may be recoverable
wetting agents - soap
salt
acetic acid
buffering chemicals
-------
The specific dye range evaluated was observed on January 2 and
on May 11, 1976. The fabric being processed was a shirting material
of 65% polyester/35% cotton and weighed 1.3 yards per pound. It was
processed at a rate of 85 yards per minute. A combination of vat and
dispersed dyes was used in the following procedure.
Process
Dye pad at 60 C
Dry
Thermosol at 210-213 C
Cool fabric
Chempad at 27 C
Steam at 104 C
Rinse at 49 C
Oxidize at 54-60 C
Soap at 88 C
Rinse at 77-82 C
Dry
Overflow Water
(dm3/s)
0.47 est
0
0.63
0.63
0.75 (4 tubs)
0.06 (2 tubs)
0.12 2
0.75 (3 tubs)
0
Observed Temperature
60
99 condensate
38 water seal
43, 44, 54, 60
70, 69
89
71, 66, 66
The
Figure 4 shows the wet processes of the dye range schematically.
cloth flow is indicated as well as the rinse, oxidation, and pH
adjustment processes required. The oxidation baths were operating
at higher temperatures and the final rinse at lower temperature than
indicated as standard procedure.
Considerable variations in procedure exists from time to time
to accommodate fabric styles and dye combinations. The department
involved handles some ninety shades per week at about 30,000 yards
per shade. The observations made in the following apply only to the
particular conditions for the run cited, but some generality is
expected to exist.
Discussion was initiated concerning the application of
hyperfiltration or ultrafiltration to the existing process for the
reclamation of chemicals. It was judged possible that some dye,
sodium acetate, and detergent could be reclaimed. Most of the
periodate is consumed in the process and is not reclaimable. It
was judged probable that hot water could be reclaimed.
On the January 2 visit, the basic range operation was observed
and samples of the drain were taken. On the May 11 visit, detailed
data for energy calculations were taken together with the gathering
of specific wastewater in drum samples for membrane screening tests
at Clemson University.
A material balance (on the chemical additions reported by plant
personnel) is given below:
17
-------
OESUPERHEAT
WATER
COOLING
WATER STEAM
OVEN I
M I
00 '
CLEAR WATER
T
M
(f
WATER
STEAMER
CONDENSATE
- '- . 3, 0.06 dm3*
0.47 dnr/s 99 c
60C
6.75 dm3* each washer
\\ \
KIOj/HOAC
1
SOAP
43C 44C 54C 60C 7OC 69C
0.063 dm ft each
CLEAR WATER
0.75 drri3* each
L I J
8
10
7IC 66C 66C
WATER SEAL
0.063 dm3/s
38C
pH=5
O.I25dm3/s
89C
pH=6 7
FIGURE ^ SCHEMATIC DIAGRAM OF CONTINUOUS DYE RANGE PROCESS
-------
% O.W.B.
Chemical Estimated
Process Used Active
Dye formulation for dye pad
Vat dye mixture 8.30 2.0
Dispersed dye mixture 3.33 1.5
Antimigrant (sodium alginate) 2.30 0.2
Salt (Nad) 0.73 100.0
Monosodium phosphate 0.25 40.0
Initial rinse, after steamer
Tubs 1, 2, 3, 4 - no chemical added
Chemical pad
Sodium hydroxide (40% solution) 3.90 40.0
Sodium hydrosulfite 3.90 100.0
Oxidation bath - Tubs 5, 6
Potassium iodate 0.25 100.0
Acetic acid (84% solution) 1.90 84.0
Final rinse bath
Tub 7 - blended nonionic surfactant 0.73
Tubs 8, 9, 10 - no chemical added,
although pH buffering additives
could be required
The first column is the percent commercial chemical based on weight
of bath (O.W.B.). The second column is the estimated actual chemical
present. The remaining material is water or commercial diluents.
Samples of the effluent were removed from each overflow on the
dye range on January 2 during which period the range operation was
similar to that on May 11, 1976.
1.3.3 Energy and Water Usage
Energy losses by evaporation, radiation, and convection, and
energy flow in the wastewater stream were examined.
The energy loss from the wash tubs by evaporation was calculated
based on the data of Appendix A-5. As shown in Table 7, the evaporative
heat loss from all wash tubs was calculated at 2.03 x 106 kJ/hr. The
evaporative heat loss from the saturated moving cloth surfaces as it
was transported from one tub to another was only about one-tenth of
the total energy loss by evaporation. About 4.30 x 106 kJ/hr of
energy was used to heat water from 18 C to process temperature and
of this amount about 1.69 x 106 kJ/hr was recovered in the plant heat
exchanger. The energy loss by radiation and convection was only about
15% of the energy loss by evaporation.
The total energy loss minus the recovered energy was found to
be 4.93 x 106 kJ/hr. The use of energy was as follows: evaporative
19
-------
NJ
O
TABLE 7. EVAPORATIVE LOSS ESTIMATE
(GJ/hr)
Wash Box
1
2
3
4
5
6
7
8
9
10
Standard Procedure
(Temperature, C)
49
49
49
49
57
57
88
79
79
79
Evaporative Loss
(GJ)
0.091
0.091
0.091
0.091
0.135
0.135
0.493
0.355
0.355
0.355
Observed Temperature
(C)
43
44
54
61
70
69
89
71
66
66
Evaporative Loss
(GJ/hr)
0.068
0.072
0.118
0.157
0.237
0.226
0.516
0.249
0.195
0.195
TOTAL
2.19
2.03
-------
losses (30%), net to heat water (65%), radiation and convection
included steamer (5%). These values are depicted graphically in
Figure 5 and tabulated in Table 8.
TABLE 8. ENERGY CONSUMPTION FOR DYE RANGE
Energy Loss
Evaporation
To heat water
Radiation and convection
Steamer
TOTAL
Less energy recovered (from heat exchanger)
NET ENERGY SUPPLIED
4.93 x 106 kJ/hr (1 hr/60 min)
Net energy consumption = 27.91 kg cloth/min
2.03 GJ/hr
4.30 GJ/hr
1.38 GJ/hr.
1.62 GJ/hr
6.63 GJ/hr
-1.69 GJ/hr
4.93 GJ/hr
2944 KJ/kg cloth
„ . . .
Net water consumption =
0.416 m3/min
= 27.89 kg/min
= O.O174
The above values are for the wet portions of the dye range.
The entire range from the predryer through the dry can postdryer
consumes an amount estimated as 7.94 x 106 kJ/hr in addition to the
amounts already delineated. Thus, the wet processes only include
approximately 40% of the total energy supplied to the range.
A detailed study on the energy consumption of the continuous
dye range is included in Appendix D.
The water consumption was calculated to be 0.0074 m3 per
meter of cloth processed. The net energy was 2945.5 kJ/kg of cloth.
1.4 Pressure Becks
1.4.1 Process Description
Water and the dry cloth are charged to the beck in sequence. The
temperature of the beck is raised gradually by adding steam though its
heat exchanger. At a certain temperature chemicals are added to the
process to effect the dyeing. The dyeing is done in a two-step operation.
The first step is to raise the temperature to an elevated level allowing
the dispersed dye to penetrate the polyester fiber. As soon as that dye
21
-------
ENERGY 14.6 GJ/hr (13.8 MBtu/hr)
WATER 6.94 dm3/* (1100PM)
PRODUCTION 1677 Kg/far
45%
18 * ,.
D*
1%
24
WM^H
%
I.R. OVEN WASHER POST
PREDRYER SECTION DRYER
DISTRIBUTION OF ENERGY ON ENTIRE RANGE
65%
to
to
ENERGY 6.56 GJ/hr(6.2 MBtu/hr)
WATER 6.94 dm3/* (HO
Hi/hr)
>M)
H.E.
HEAT >
RECOVERY
>
r
1
DISTRIBUTION OF ENERGY
ON WASHER SECTION OF RANGE
;
50'
%
5%
n
(EATING VAPOR HEAT
WATER LOSSES LOSS
FIGURE 5 ENERGY DISPOSITION ON CONTINUOUS DYE RANGE
-------
has penetrated the cloth sufficiently a patch test can be made validating
its performance. Then the temperature is adjusted to a new, usually
lower level for a certain time to accomplish the direct dyeing of the
cotton. Following the dyeing the beck is rinsed with fresh water,
the fluid is dropped, more water is added, the rinse is again accomplished,
and the cloth then is removed.
1.4.2 Chemicals Used
Chemicals used are a combination of dispersed and direct dyes,
salt, and other dyeing auxiliaries.
1.4.3 Energy and Water Consumption
Pressure becks were also studied to determine the energy lost
by evaporation, convection, and radiation during the dye cycle and
to determine the energy associated with heating water in the beck.
The observed dye cycles for these becks are given in Figure 6 (low-
pressure) and Figure 7 (high-pressure).
The low-pressure beck cycle studied involved the dyeing of
3175 meters of material weighing 1280 kg. The water requirement
to fill the beck at the start of the cycle was 13.63 m3 at a
temperature of 18 C. During the dyeing cycle there were three
periods of heating before the actual dyeing took place. Initially
the water was heated to 52 C, the cloth and a chemical charge added.
The temperature was then elevated from 52 to 85 C, the pH of the
fluid was adjusted, and a final temperature adjustment to 104 C was
made. The actual dyeing took place after the beck contents were
heated to 104 C. The beck door was open at times during the cycle
to load and unload the cloth, to add chemicals to adjust the pH,
and to check the cloth for proper color (patch test). At these
times, vapor allowed to escape from the vent amounted to a heat
loss of 0.79 x 106 kJ. Convection and radiation losses were
calculated using an overall average heat transfer coefficient of
11.3 W/m2-K. The heat transfer in this particular case by
radiation and convection was determined to be only 7% of the total
heat losses. The total energy requirement for the low-pressure
beck was calculated to be 4787 kJ/kg of cloth. The water
consumption was 81.76 m3 or 0.074 m3/kg of cloth. These
consumption rates were lower than comparable rates obtained for
atmospheric becks, due to lower liquor ratio and the reduced
evaporation from the closed beck. As shown in Figure 8, approxi-
mately 80% of the total energy was used to heat water.
The high-pressure beck dyeing cycle involved 877 kg of cloth.
The beck was filled with 10.6 m3 of water at the start of the cycle
23
-------
1.16
1.92
2.6 3.2
Drop
Fill
I
140
120
o 100
S 80
1
40
20
ft
.ood Cloth „ , . Complete
and Adjust Cool and Dye ™» gJJ
Chemicals pH Dye Patchiest Process c»"» ™
\ \ 1 1 I 11
/
. __y
/
t
i
i
i
v
\
1
Time (hours)
FIGURE 6 CYCLE PROCEDURES FOR LOW PRESSURE BECK
-------
120 -
100 -
a.
0
Peitch Patch
Tt*t Dyt fttt Dyt
M
Riteh
Tttf
*
Drop
Fill
to
CQO I Unload
Cloth Cloth
2.9 4.7 6.1 ?.§ 8.3 10.0 10.5
Time (hows)
12.0 13.0
7 CYCLE FOR HIGH
-------
at a temperature of 18 C. The dyeing was performed during three
periods, each at 121 C a higher temperature than for the low-pressure
beck. All energy calculations were performed in the same manner as
for the low-pressure beck. The total energy required to heat the
water and replace that lost by evaporation, convection, and radiation
was nearly 10.11 x 106 kJ, or 11530 kJ/kg of cloth. The total water
requirement was 69.65 m3, and amounted to a water consumption of 0.092
m /kg of cloth. Again, the consumption rates were lower than obtained
for atmospheric becks. Also, nearly 80% of the total energy is used
to heat water as shown in Figure 8.
1.5 Continuous Rope Preparation Range
1.5.1 Process Description
Cloth comes from a storage pit where it may be treated with
enzymes to break down the starch in the starch/PVA mixture of
sizing materials. Cloth comes in a rope form to the alternator
where it is circulated at high temperature for the removal of the
sizing material. It then is fed into a caustic saturator which
basically exposes the cloth to 0.4% caustic solution. It flows
from there into a J-box where it is exposed by steam at 99 C for
about one hour. It is drawn from the J-box into a caustic washer
and then goes into a second washer at lower temperature to further
wash the cloth of its dissolved materials. Then the cloth goes
to a peroxide saturator where it is exposed to a bleaching solution,
wetting agent, peroxide bleach, and a mild caustic addition. It
is allowed to steam for one hour at 99 C and then is rinsed in
clear water at 71 C and in a second washer at 27 C. The cloth is
then prepared for further finishing.
1.5.2 Chemicals Used
Chemicals used on the range are mild caustic solutions, bleach,
and wetting agents.
1.5.3 Energy and Water Consumption
The rope-form preparation range studied had process elements
shown in Figure 9. In this range the alternator performs the
desizing operation, and there is a saturator, J-box, and washer
combination for both the caustic and bleaching sections. Water
from washer #4 forms the entire water flow for washer #2. Flows
and temperatures of the drains are indicated by Figure 9. The
energy calculation methodology was similar to that of the open-
width range. Vapor loss from the enclosed washers was assumed
26
-------
Dye Cycle-Low Pressure
Energy* 4780 tCJ/Kf (2050 ity/tb.)
104 € for 90 minutes
85 C for 200 minutes
80%
13%
T%
Energy to Energy to Entrfy to
Heat Water Vapor Losses External Losses
tsj
vj
Dye Cycle—High Pressure
Energy =11500 KJ/Kg (4940 Btu/lb)
121 C for 84, 29, and 30 minutes
82 € for (80 minutes
79%
J.
13%
JZZL
Energy to Energy to Energy to
Heot Woter Vapor Losses Externol Losses
8 FOR
-------
to
00
I STORAGE WAT£R
PIT WASHERS
\
CAUSTIC
SATURATOR
WATER FROM
I WASHER NO.4
CAUSTIC
WASHER.
0.416
t-
0
r -
-
0.378
PEROXIDE
SATURATOR
Q.4I6 m3*4—
71 C
to ttothtr no. 2
0.304n?/«
49 C
FIGURE 9 SCHEMATIC DIAGRAM OF ROPE PREPARATION RANGE
-------
negligible, and the loss rate per unit area of the rope-form cloth
was taken to be equivalent to that in the atmospheric beck observa-
tions. The physical envelope of the rope was estimated for an area
basis. The steam rate to the J-box was taken from piant-supplied
data, but the value reported is much higher than expected based on
calculated estimates.
The entire system energy use required to operate 'the range was
calculated as 13.0 x 106 kJ/hr. The production rate was 174 m per
minute of cloth having 3.47 m/kg, yielding 3007 kg per hour of
production. The energy use per unit of production was therefore
4341 kJ/kg. Figure 10 shows the distribution of energy on the range
by category. As previously mentioned, the J-box may actually
require a much smaller amount than indicated. In any event, the
heating of water required at least 76% of the total energy required
for the rope range.
The mill design provides for the collection of hot drain water
to allow heat exchange with incoming water. Other sources of "waste"
energy are also used to elevate the average inflow temperature such
that two-thirds of the energy required to heat water is supplied by
energy recycle. This fractional energy recovery is indicated in
Figure 10 by shading. Only a portion of the energy recovery
originates at the particular range, and the portion is not readily
determined.
29
-------
76%
Energy = 1.3 GJ/hr (12.3 x I06 Btu/hr)
U)
o
PLANT
RECOVERY
portly
ASSOCIATED
WITH
PROCESS
f
\
**"^m
W////////////A
•~- Woters f,786_nrVmin (280 GPM)
Production = 1800 kg/min
20%
^^^^^" ^» ^F «w
r-,2% r-,2%
HEATING J-BOX OTHER VAPOR
WATER HEAT LOSS
LOSS
FIGURE 10 DISTRIBUTION OF ENERGY ON ROPE PREPARATION RANGE
-------
2.0 CHARACTERIZATION OF WASTEWATER DISCHARGES
2.1 Atmospheric Beck
The analyses of several process wastewater samples taken
directly from the atmospheric dye bath are presented in Tables 9
and 10. The highest concentrations of chemicals are present in
the dye baths, as shown by the total solids analyses (8253, 5701,
6378, and 9200 mg/J,). These solids are 90% salt, which is used to
cause the dye to adhere to the fiber. The salt may be concentrated
by hyperfiltration but not by ultrafiltration. The remaining 10%
of the solids are organic additives, such as dispersing agents and
dyes, which also may be concentrated by hyperfiltration but less
so or not at all by ultrafiltration membranes.
2.2 Pressure Beck
The analyses of process wastewater samples taken from the
pressure beck on November 3, 1975, is presented in Table 11. These
data show very high concentrations of salt being discharged by the
high- and low-pressure becks. It seems that the salt, which is over
2% concentration, is valuable if suitably processed by high-temperature
hyperfiltration membranes.
2.3 Open-width Preparation Range
Water taken from the various drains of the continuous preparation
range on November 3, 1975, and March 2, 1976, was analyzed as shown
in Table 12. Analysis of the cloth for material removal was conducted
as shown in Appendix E. Calculations using measured flow rates
showed reasonable comparisons of "removal from cloth" and "addition
to water." Appendix E also shows a breakdown of the relative con-
stituents of the sizing formulation as removed. The drains from
the alternator, desize washer, and caustic washer show large organic
loading as is expected.
2«4 Continuous Dye Range
The continuous dye range (see Figure 4) was sampled and each
drain analyzed as reported in Table 13. The first four boxes show
31
-------
TABLE 9. ANALYSIS OP SAMPLES - ATMOSPHERIC BECK
oo
to
Identification
#1 Fix,
10/31/75
#1 Dye,
10/31/75
#2 Fix,
10/31/75
#2 Dye,
10/31/75
Beck #4,
Lot 6199
Lot 6199,
Lot 6159
Lot 6159
Dye Fix
#6186, 10/31/75
Beck #4,
10/31/75
Beck #15
10/31/75
Fix #15,
10/31/75
Dye #16,
10/31/75
Beck #18
10/31/75
Dye 6186
, Dye 5685
5685,
L 94896
, Dye 6052
COD
98
400
605
2176
88
3182
996
224
512
1610
TOC
22
121
173
441
21
396
159
66
151
191
Color
133
1540
224
98
63
602
154
42
6272
455
Total
Solids
557
298
3194
8253
830
5701
6378
300
948
9200
Suspended Methylene Blue
Solids Active Surfactants Chromium
10
54
22
18
18
43
44
9
640
35
0.34
3.50
<0.10
<0.10
<0.10
<0,10
1.90
0.08
29.00
0.26
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
003
006
002
001
005
Oil
001
001
006
003
Copper
0.042
0.250
0.042
0.025
0.008
0.058
0.058
0.042
0.025
0.033
pH
6.40
6.95
6.80
7.00
6.95
8.40
5.45
4.30
7.50
7.60
All results are reported in mg/l, except pH (unitless) and dolor (ADMI).
-------
TABLE 10. ANALYSIS OF SAMPLES - ATMOSPHERIC BECK
OJ
Identification
Beck #21, Dye Fix
Lot 292, 10/31/75
Scour Dump #3,
9/15/75
Overflow Dye #38,
9/15/75
Dye to drain #38
Scour #3
Dye Start Overflow
#38, 9/15/75
Scour Overflow #3,
9/15/75
Scour Solution,
Beck #4, Lot 6279,
11/5/75
Scour Solution,
Beck #6, Lot 6274,
11/5/75
Scour Solution,
Beck #15, Lot 6269,
11/11/75
COD
49
270
530
56
614
1033
1098
1740
609
1795
TOG
41
158
196
69
300
338
455
597
324
701
Color
77
168
532
350
658
1043
938
490
210
266
Total Suspended Methylene Blue
Solids Solids Active Surfactants Chromium
876
302
4494
148
642
9376
1150
1806
1005
1252
11
8
24
10
92
40
154
40
4
66
<0.10
<0.10
<0.10
<0.10
<0.10
<0.10
<0.10
<0.10
<0.15
<0.23
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
005
002
002
006
006
001
007
004
005
019
Copper
0
0
0
0
0
0
0
0
0
0
.033
.021
.379
.053
.016
.293
.053
.053
.011
.032
pH
7.2
6.6
6.1
6.8
6.7
5.8
7.3
10.1
10.4
7.0
All results are reported in mg/£, except pH (unitless) and color (ADMI).
-------
TABLE 11. ANALYSIS OF SAMPLES - PRESSURE BECKS
Total Suspended Methylene Blue
Identification COD TOG Color Solids Solids Active Surfactants Chromium Copper pH
High-pressure beck
85°F just before
drainage, 11/3/75 2821 346 1337 21958 72 5.3 0.022 0.085 6.4
Low-pressure beck
180°F end of dye
time, 11/3/75 558 93 371 6294 28 - 0.055 0.133 6.55
Low-pressure beck,
115°F just before
emptying, 11/3/75 3221 344 763 26758 43 - 0.067 0.155 7.6
a •
All results are reported in mg/Jl, except pH (unitless) and color (ADMI).
-------
TABLE 12. ANALYSIS OF SAMPLES - OPEN WIDTH PREPARATION RANGE
u>
on
Identification
Continuous range
caustic washer
11/3/75
Continuous range
desize washer
11/3/75
Continuous range
COD TOG Color
7954 4260 1344
8336 3846 658
Total
Solids
4014
8578
Suspended Methylene Blue
Solids Active Surfactants Chromium Copper pH
74
404
1.4 0.033 0.075 12.
1.2 0.022 0.219 6.
2
4
GC alternator washer
11/3/75
Continuous range
peroxide washer,
11/3/75
Alternator rinse,
3/2/76
Caustic desize,
3/2/76
1405 611 644
3452 1753 630
4160
22200
1630
3948
2930
17625
314
100
270
2680
0.3 0.055 0.171 6.
0.044 0.200 11.
0.02 6.
0.04 - - 12.
3
0
2
45
All results are reported in lag/H, except pH (unitless) and color (ADMI) .
-------
TABLE 13. ANALYSIS OP SAMPLES - CONTINUOUS DYE RANGE
Analyses
PH
BOD
COD
Total Solids
Suspended Solids
Color
Specific Conductance
Chromium
Copper
Tub
#1
12.4
140
240
4086
46
3840
5500
0.030
0.070
Tub
#2
11.8
25
240
1306
10
880
2300
0.280
0.040
Tub
#3
11.7
35
200
1187
16
1000
2000
0.080
0.075
Tub
#4
11.5
20
740
816
22
2480
1200
0*125
1.15
Tub
#5
4.4
0
0
4035
132
840
1800
0.125
1.55
Tub
#6
4.5
0
0
5085
198
1640
2300
0.155
4.30
Tub
#7
4.5
0
0
3934
228
2835
2000
0.030
3.60
Tub
#8
4.3
0
620
992
18
640
800
0.125
0.50
Tub
#9
3.7
50
500
282
12
35
500
0.060
0.10
Tub
#10
. 3.8
50
500
100
16
25
400
0.060
0.15
Results are in mg/fc, except for pH (unitless)
(micromhos/on) .
color (APHA units), and specific conductance
-------
the washout following the chemical addition on the range. Total
solids diminish from box #1 through box #4. The pH gradually
diminishes from a high value down toward a pH of 11. COD is fairly
low compared to the amount of total solids in the stream and the
color first diminishes and then rises as the temperature of the
bath is escalated forcing more and more of the colored material
from the cloth. In tubs #5 and #6 potassium periodate is added
with acetic acid to oxidize the dye and also prepare the cloth for
final finishing and resin finishing. In tub #7 soap is added to
clean the cloth of dispersed material and the washout in tubs #8,
#9, and #10 is shown to be effective in reducing total solids even
though the water is down to about the 100 mg/fc level. Boxes #5,
#6, and #7 have very minimal overflow contributing only slightly to
the total loading from the continuous dye range.
2.5 Rope Preparation Range
The rope preparation range (Figure 9) was sampled and the
analysis is reported in Table 14. The first box shows evidence
of a large organic loading corresponding to the size removal in
the first box. Conductance is fairly low indicating the amount
of ionics is low and the suspended solids is a relatively large
value probably indicating large amounts of cotton fiber being added
to the bath. The results for the second and third wash box (follow-
ing the caustic step) also indicate considerable organic loading.
Wash box #4 which follows the peroxide saturator has much less loading
of all constituents than the earlier boxes.
TART.TC 14.
Parameter3
Sample Number
Conductance
PH
Phenol Alkalinity
Total Alkalinity
Total Solids
Suspended Solids
COD
Residual Chlorine
ANALYSES OP
Wash Box
#1
1067
896
4.0
0
107
7695
3320
27520
0
DRAINS IN ROPE PREPARATION RANGE
Wash Box
#2
1068
5400
10.9
925
2400
10438
1360
9880
too colored
Wash Box
#3
1074
1080
9.0
16
547
1769
266
2293
0
Wash Box
#4
1078
585
8.3
0
288
1007
169
656
0
All results are in mg/£, except conductance (micromhos/cm and
PH (unitless).
37
-------
3.0 POSSIBLE PROCESS MODIFICATIONS
It is recognized that certain process modifications may be
required to allow incorporation of hyperfiltration. Other methods
are foreseen which aid the economic benefits by allowing lower
cost installations or more complete chemical reuse.
3.1 Atmospheric Dyeing
Closed-cycle or substantial permeate reuse will result in
there being no cold water for the beck processes. This causes a
number of perhaps small changes in the procedure normally used in
atmospheric dyeing. The only water available to fill the beck to
begin scour is hot whereas at present it is cold. The rinse that
is conducted at the end of the scour process will be hot water
rinse rather than cold water rinse. The water available for filling
the beck prior to dyeing will be hot. The dye and all of the auxi-
liaries will be added to cloth and water which is already hot rather
than at present adding the dye and auxiliaries cold and raising the
temperature gradually. Again, the rinse will be conducted with hot
water rather than with cold as it is at present. There are several
problems, some mechanical and some having to do with the relative
application of the materials. It must be determined that the cloth
can withstand high-temperature rinses and that if any draining is
to be done at the hot condition that the cloth will not be harmed by
dropping the fluid in preparation for the step while the cloth is
hot. The dye certainly must be carefully added to the bath so that
application of dyestuff is uniform on the cloth to prevent streaking
and diminish end-to-end variations. It is anticipated that the
procedure can be used with direct dyes with some difficulty. Because
of the anticipated difficulty a full-scale evaluation was conducted
and is reported in a later section. The evaluation resulted in a
satisfactory dyeing so that the application of the dye to hot solution
is a reasonable assumption. Some of the side benefits of having only
hot water available, including that for rinsing, is that the rinsing
may be more effective, allowing the dyeing to be done in a shorter
period resulting in a greater productivity of the unit. If hot water
processing is done at a faster rate the fraction of energy which goes
to evaporation loss will be diminished also resulting in a comparative
energy saving.
Salt in dyeing is used to force the dye to the cloth. Typical
direct dye exhaustions show 70% without salt and 90% with salt. By
using a higher concentration of dye in the dyebath, the direct dyeing
38
-------
operation may be performed with no salt at a cost savings. Since the
dye may be recovered and concentrated by the hyperfiltration unit, no
dye is wasted. It is possible that the dye can be concentrated
further if the relatively large salt content is not used due to elimina-
tion of the large osmotic pressure resistance which could otherwise be
expected. The procedure should speed up production by shortening the
dye cycle.
Modifications are possible which will contribute to a decreased
volume flow of water. The rinsing procedures can be improved by
adopting a drop/fill rather than a rinse overflow operation. A
critical standard for sufficiency of rinsing is badly needed. By
careful minimization of the water flows the cost of a hyperfiltration
unit will be proportionately reduced.
3.2 Continuous Dye Range
The conventional continuous dye range in the textile industry
today uses dispersed dyes for polyester fibers and vat or reactive
dyes for cotton or rayon fibers. This applies to blended fabrics
and much of the 100% cotton fabrics which are currently processed.
Since these dyes leave little usable residue in the rinse solutions,
dye recovery is not practical. The streams examined for recovery
were the rinsing or scouring baths which follow immediately after
the fixation of the dye in the steamer. Hot rinse water, detergents,
and acetate convertible to acetic acid can be recovered. A recovery
scheme using a modified process is shown in Figure 11.
The application of an ultrafiltration membrane to the rinsing
baths after the steamer will allow water to be recovered and conserve
acetic acid in the oxidation tubs (#5 and #6)» It is also possible
to recover detergents and hot water from tubs #8, #9, and #10 with
a hyperfiltration membrane, or hot water with an ultrafiltration
membrane.
As the analyses of water samples taken from the different tubs
in the dye range show, the highest concentration of chemicals in
the rinse tubs (#1, #2, #3, #4, #8, #9, and #10) is 0.5% solids.
It is projected that the solids in Tubs #1, #2, #3, and #4 may build
up to 6000 to 12,000 mg/A in a recycle system such as that shown in
Figure 11. This higher solids level would be of value in tubs #1 and
#2 in preventing washdown of the vat dye before it is oxidized.
The countercurrent flow shown in tubs #8, #9, and #10 should
facilitate the buffering of the fabric to pH 4-6, required for
resin finishing. The buffer could be regenerated with mineral
acid as it is recycled.
39
-------
SOAP
PERMEATE
t*.
o
hKUM BELJDVr
LOW TEMPERATURE v- "p" — — ^
COUNTERFLOW WASHING moke up — -J
^ •• ., *« - ^.^ KIO- and 1
NEUTRALIZE -
(MINERAL
ACID)
^^
\
I
i
*[
i _u
]pH
^mMMM^
2
4
i
3
1
ACETJC ACID r
- x"" '—•-*— •*%, t t
45678
fell ACETIC _J
ACID "K-
l»7
UF
N I
1 s
-moke up
* ^* *
(OVERFLOW)
< i
£ N
1
* 9
1
10
-+. CONCENTRATE
s FOR POSSIBLE
' f RECYCLE TO
1 BOX NO. 7
TO WASHERS
T
CONCENTRATE
FIGURE 11 IMPLEMENTATION OF PROCESS MODIFICATIONS FOR APPLICATION TO DYE RANGE
-------
3.3 Continuous Rope Preparation Range
The best place for process modifications appears to be the
scouring operation. If 1-5% solvent (varsol-type) can be used
with the scour solution it should replace some of the alkali used
and effectively removed waxes to give a well-scoured fabric. This
would decrease the alkali consumption and make the waste more suitable
for water recovery by commercial membranes. The alkali load on the
waste treatment system would also be reduced.
The membranes presently available do not provide any significant
concentration capability for caustic. However with a saturator source
at 6% NaOH, the wash water approaches 0.4% NaOH when it enters in a
neutral condition. If 0.4% NaOH water were used for washing, the
effluent would approximately double in concentration. An exponential
type rise would continue until an equilibrium condition is reached.
At equilibrium, the caustic removed in the washers will be concentrated
as if in the makeup flow alone. Thus the level of caustic in untreated,
recycled wash water will rise despite the fact that the membrane does
not specifically separate it.
The above leads to the process modification shown in Figure 12.
The (assumed four) washers will be operated in counterflow with the
membrane applied to the effluent. The concentrate containing waxes and
trash removed from the cloth will be routed to waste treatment. The
permeate, now highly caustic, will be divided into a stream for recycle
and a stream for caustic recovery by evaporation. Makeup fluid will
be added to the last washer section in an amount which will preclude
detrimental effects of excess caustic in the bleach section. It may
be possible to reduce the caustic addition to the bleach saturator.
The savings would include (a) water and energy of recycled permeate,
(b) a reduction in the purchase of caustic in the amount of that
recovered from permeate, (c) perhaps less caustic addition in the
peroxide saturator, and (d) lower cost of waste treatment both by
reduction of neutralizing acid and by a decrease in the volume of
total waste.
3-4 Pressure Beck Dyeing
Many of the statements from the atmospheric beck dyeing section
are applicable to the pressure beck. Some cooling of the beck water
will be necessary to preclude the tendency of all reuse water to
assume the highest process temperature. The degree of cooling is
not expected to be substantially different from that in present
practice.
41
-------
CAUSTIC:
SATURATOR
STEAMER
J-BOX
,. 1
moke up (25%)
100% UF „ HF
90%
T
75%
15%
TO CAUSTIC
RECOVERY
FIGURE 12 CAUSTIC WASHER SECTION PROCESS MODIFICATION
-------
3.5 Open-width Preparation Range
The open-width range surveyed is provided with plumbing to allow
recycle of water from the bleach washers to the upstream desize washer,
but this recycle is not always maintained. It is considered possible
that a water savings may be effected through the use of higher
operating temperature. A full-scale demonstration of this concept
is included in a later section. The use of a solvent for scouring
may allow the same advantages as on the rope-form range.
43
-------
4.0 ENERGY SAVINGS POTENTIAL THROUGH RECYCLE
The dyeing and finishing of textile products is a process
which consumes a considerable amount of industrial water. The
effluent water from these processes contains residual chemicals
and energy which may warrant recovery on a purely economic basis
in addition to the resultant benefits of environmental protection
and energy conservation. This report segment presents the details
of an energy and chemical balance on selected piece dyeing processes
which represent current practice for this class of batch process.
The authors realize that considerable variations in procedure exist
among the plants; nonetheless, the results are presented in the
hope of reasonably general applicability and to explain a method
for evaluating such a paucity of field data.
4,1 Atmospheric Beck
The energy use has been reported in Section 1.1.3 and Appendix C.
Discussions of process modifications are found in Section 3.1. Here
it is assumed that rinsing may be performed with hot water, that the
fluid may be "dropped" at temperatures up to 75 C, and that chemicals
and dye may be administered at any temperature. The permeate from
the dye and scour operations may be used Indiscriminately, but the
fixing fluid permeate must be segregated. It is further assumed
that the various baths may be drop/fill diluted in contrast with
the present overflow rinse.
In addition to the assumptions stated concerning feasible
operations the following estimates of a numerical nature are assumed.
On the average, four beck volumes are used to rinse after each
process , the last volume being the "fill" fluid for the next
process. The cloth will retain 5% of one beck volume when the fluid
is drained. The practical limit is expected to approximate a 75%
fluid drop. All measures for energy recovery apply only to energy
used for heating water (about one-half the total) and the remaining
half is not subject to recycle by any water based process, including
hyperfiltration.
4.1.1 Energy Recovery by Plant-scale Heat Exchangers
As mentioned in the assessment of energy balance for the pro-
cesses, the outflowing water in every case has been warmed considerably,
44
-------
One method of energy recovery is to operate a heat exchanger with the
outflow water heating the inflow water on a plant scale. Some
incoming water leaves the dye plant on the cloth enroute to the
dryer while condensing steam usually adds a similar amount during
the water heating mode. Therefore the hot and cold flows are
approximately equal. For such a case in a counterflow (or equiva-
lently a multiplass crossflow) heat exchanger, the temperature loss
in one fluid equals the gain in the other. All of the fluid entering
the plant, including rinse water, is elevated in temperature so that
the outflow temperature will rise when energy is recycled in this
manner. Let AT represent the rise in temperature as the fluid passes
through the plant. This elevation in temperature results from one
volume of fluid each at 74 C, 96 C, and 49 C (as shown in Figure 2),
diluted with N (taken as 3) volumes of rinse water at Tj, or in equation
form conservation energy yields
AT (N + 1) X 3 = (74 - Tj) + (96 - Tj) + (49 - Tj) . (1)
When TJ = T , the process entering temperature (Tj) equals the incoming
fluid temperature (Tg). A value of TZ = TS = 18 C, and N = 3 results
in AT = ATS = 13.75. When a heat exchanger of effectiveness E is used,
the outflowing hot fluid temperature will be AT + Tj, and the
definition of E
T - T
I s = E (2)
TT + AT - T
X S
allows computation of T from equation (1) as
T (1 - E) (N + 1)
T = _§ + (74 + 96 + 49) E
N + 1 - NE 3(N + 1 - NE) ( '
Strictly speaking this equation applies only up to Tj = 49 C at which
point it implies that the water would be cooled to perform the dye
fixing at 49 C. The energy saved is that fraction not used to heat
the water from Ts to Tj or
3 (Tz - Ts)
Fraction saved = . (4)
(74 - Ts) + (96 - T ) + (49 - T )
S
Figure 13 shows the fractional savings as a function of the exchanger
effectiveness E. Present economics are such that this type heat
exchanger may be purchased for a cost of the order of a year's energy
savings. However, some fluids are not amenable to cooling due to
high fouling or congealing. Also no chemical or water reuse is
available using heat exchangers.
45
-------
100-
-------
4.1.2 Energy Recovery with Filtration Devices
4.1.2.1 Plant-scale Filtration Application - Hyperf iltration (reverse
osmosis) or ultraf iltration devices offer the potential for recovery
of useful industrial water from the effluent of the process. The
temperature of such permeate will be at or near the temperature of
the effluent. The amount of permeate produced will be less than
the effluent, and thus it will be diluted with cold water to form
plant inflow. It is assumed that the cold make up water is uniformly
mixed with the permeate for purposes of energy computation, although
non-uniform, tailored make up water addition would be potentially
advantageous. The temperature rise equation (1) is valid, and with
R to denote the fraction of water reused, the influent temperature
obeys the equation
(1 - R)TS + RCT-j. + AT) - Tz (5)
Comparison with equation (2) shows that R for the filter plays the
same role as E does for the heat exchanger. However, it has been
assumed that the fix water must be separated from the scour and dye
effluents.
4.1.2.2 Filtration Applied to Specific Processes - The foregoing has
assumed that only the permeate has been recycled. Because of higher
capital costs the filter systems are not apt to be justified on the
basis of cost savings of energy alone, but may be justified on the
basis of the sum of water, energy, chemical reclamation, and environ-
mental benefits. On this basis, it may be that all of the effluent
will not be processed, but rather only that optimum fraction which
is economical. Herein, the analysis is directed toward the concept
of closed cycle, processing all of the water. Also, the effluent
from individual processes will require segregation for greatest
recycle potential; thus, the filter systems will best be applied
to specific processes. The equations for concentration (C) and
temperature (T) of a well-stirred volume of mass (M) having inflow
and outflow at constant rate (m) are
— &
-------
Equations (6) relate the decline in temperature (T) and concentration
(C) from initial values TQ and CQ towards Tj and Cj. The value N is
the ratio of m x time (= rinse mass) to process mass M.
Suppose the rinse were performed by drain and fill. If a
residual volume exists in the amount of eM, each drain and fill of
volume (1 - e)M will result in a decrease in temperature and concen-
tration of
Tn-TI cn - CI _
The subscript n denotes the number of drain/fill cycles. Since each
drop/fill increment consumes (1 - e)M fluid, the final concentration
or temperature difference ratio is related to the total rinse volume
NM = n(l - e)M by
Tn - Ti Cp - CI
T° ' TI = Co - CI
N = (1 --e)n; n - Of lf 2... (7)
The predicted values of temperature and concentration are displayed
for both the drain and fill or continuous inflow and overflow model
in Figure 14. In each case it is assumed that the concentration of
material on the cloth is identical with that in the beck. This
assumption is valid in the limit of infinitely rapid dissolution
and diffusion rates. Based on the data of Reference 5, and the
time scales for the beck washing process, the assumption is reasonable
for most dye substances. The drain/fill procedure uses about 30%
less rinse water than the overflow procedure for a dropped volume
of one-half beck and 44% less water for a dropped volume of three-
quarters beck.. The rinse procedure is important because when the
total water use is minimized, the temperature and concentration of
the mixed average effluent will be correspondingly higher and the
filtration system costs will be lower. The prospect for energy
conservation and economic chemical recovery is similarly greater.
The present set of assumptions indicate from Figure 14 that use
of four units of rinse water in the overflow method reduces the
concentration to below 2% of the original amount. Use of 2.25 units
produces a slightly lower concentration in the drop/fill procedure
with 0.75 beck volumes dropped and restored three times.
For the scour and dye portions of the cycle the fluid required
will be one original fill, three 0.75 drop volumes after scour, and
two 0.75 drop volumes after dyeing. The third 0.75 drop volume will
be filled from the fixing bath permeate. Therefore the dropped
amount is 6(0,75) while the refill amount needed is 1 + 5(0.75)
or a ratio of 94.7%. The dropped amount may only be partially
recycled and the net water recycled is some fraction of the 94.7%
48
-------
CONTINUOUS INFLOW
AND OVERFLOW
- DRAIN AND FILL
£= RESIDUAL FRACTION
I 2 3
N = RATIO OF RINSE VOLUME TO BECK VOLUME
FIGURE 14 COMPARISON OF ENERGY RECOVERY WITH DROP-FILL AND OVERFLOW WASHING
49
-------
which leaves the beck. It is possible to increase the 94.7% value
somewhat by achieving a more complete drop at the step before fixing.
The temperature rise to T experienced by water at T used as
rinse and fill water is
{(74 - T ) + (96 _ T )}
T - T = - = - .
01 1 + 5(0.75)
At a high enough recycled fraction, Tj will achieve 74 C. For recycle
fractions greater than that the scour will be allowed to rise above
74 C or, equivalently , the vapor energy loss will be mitigated by
recycled' energy. Assuming T > 74, the rise equation is
T - T
(96 - T)
- i_
1 + 5(0.75)
Using the above equations as appropriate and assuming the makeup water
source T_ is 18 C results in
O
T - T
J[ _ 8 = R
T - T
o s
where R is the recycled fraction. The energy savings for T < 74 is
2 74
(74 - Ts) + (96 - Ts)
(T -T ) + (74 - T_)
% saved = -1 s s'
(96 - T ) + (74 - T )
s s
The computations are presented in Figure 15. At about 92% recycle
TZ = 74 C is achieved. With full recycle (94.7%) some 89% of the
energy of water heating may be recovered.
A similar calculation for the fix cycle portion has been
performed. More water emanates from the process than is added by
aslight amount. High recovery membrane units may then produce a
local excess of water. Figure 15 also shows the relation of energy
recycle to water recycle for the fix portions. It is essentially
the same relation as for the scour and dye cycle portions.
It is estimated that by recycle of 90 to 95% of the water
between 80 and 90% of the hot water energy may be recycled to the
process. The hot water represents half of the total energy require-
ment, so that 40 to 45% of the total energy is achievable by recycle,
50
-------
100
90
LI
80
70
60
50
o
UJ
cr
o
o:
UJ
z
UJ
z 40
UJ
o
IT
ff 30
FIX STEP
DYE AND SCOUR
STEPS TOGETHER
20
10
0
J.
J_
J.
50 60 70 80 90 100
PERCENT WATER RECYCLE
FIGURE 15 ENERGY RECOVERY RELATED TO WATER RECOVERY
51
-------
4.2 Pressure Beck
Evaluation of the energy recoverable in pressure beck dyeing
is subject to the same type of analysis and general restrictions as
for the atmospheric becks. In the pressure beck, however, cooling
from high temperature is done by a heat exchanger with process water.
Any attempt to use the process water soon will overflow the process.
It is interesting that the process water is not being used as makeup
for beck filling at present. Such use could be arranged easily by
employing a temporary storage tank. Certainly the use of membranes
to recover one unit of energy is not economically competitive with
the use of a tank to store heated process water to recover one-half
unit of energy,
The pressure beck uses steam to heat to 110 C or 121 C and
process water cools .it to around 85 C. Several dye "adds" are
made with excursions to dye temperature and back to 85 C. The
energy needed to heat water from 18 C (average annual temperature)
to 85 C is the maximum recoverable amount. The total energy to
.heat water is that to raise water from IB C to JJ.O or 121 C plus
that required for excursion from 85 C to 116 or 121 C for each
dye add. Assuming that N dye adds are made and that 90% energy
reuse (95% water reuse) is achieved the energy savings is estimated
as
% savings - °'^ + 30, high pressure (121 C) dyeing
0.9 (67)
% savings = 98 + 31N low pressure (110 C) dyeing
Table 15 shows the expected energy recovery possible through the
recycle of spent dye baths. The table is based on the estimated
water heating requirement of 80% of the total energy use for the
pressure becks. In each of the cases the recycled energy is the
same amount, but as the number of "adds" is increased the total
energy increases so that the recycled percentage diminishes.
TABLE 15. ENERGY RECOVERY ESTIMATES FOR PRESSURE BECK DYEING
-Percent Energy Recovery
High-pressure Dyeing " Low-pressure Dyeing
Number of Dye Adds _ (121 c) _ _^__ _ (110 C)
- 5— — ~ — _
1 34% 41%
2 27% 34%
3 23% 29%
There is one possible reservation associated with the recovery
of fluid for pressure beck dyeing. The material survey from the
52
-------
bath indicated that salts were added to the dye fluid in the amount
of about 2%. Recovery of 95% of the water as implied above will
raise the salt concentration to a relatively high value. At such
a high value, the osmotic pressure of the concentrate will begin
to cause lower fluxes or higher operating pressures. Also dye
precipitation and finally salt precipitation may occur each with
attendant problems.
4.3 Dye Range and Preparation Range
The energy recoverable from drain water on the range is,
except for direct heat losses to the environment, exactly proportional
to the water recycled. In situations where chemical concentrate recycle
is practical, the recovery will be high as demanded by the chemical
bath. In most cases the chemical will be required to be added at or
above the concentration used at present to insure no overflow of
the saturator. In situations where chemical concentrate is not recycled,
the recovery level may not be as high. The dye range is expected to be
a high recovery (98%) situation whereas the preparation range is only
projected to recovery levels of 90% due to the very high solids loading
and low probability for chemical reuse.
4.4 Summary of Energy Recovery Estimates
Table 16 shows in summary the expected savings for each process
when water recycle is practical. The number of dye adds has been
estimated as shown in Figures 6 and 7 to be 1 and 3, respectively,
for the low- and high-pressure becks. Recovery of energy from concen-
trate has not been credited in Table 16, and realistic volume recovery
levels have been estimated in agreement with the test results in
following sections.
53
-------
TABLE 16. POTENTIAL FOR ENERGY SAVINGS BY WATER REUSE - SUMMARY
Ln
ox
Atmospheric beck
Pressure beck
low-pressure
high-pressure
Dye range3
Open-width preparation
range
Rope preparation range
Total Energy
kJ/kg cloth
4.47 x 104
4.79 x 103
1.15 x 104
3.95 x 103
9.3 x 103
4.34 x 103
Hot Water Energy %
kJ/kg cloth
2.29 x 104
3.83 x 103
9.11 x 103
2.55 x 103
7.08 x 103
3.29 x 103
Volume Recovery Savings Potential
estimated kJ/kg cloth
95b
95b
95b
95
82
80
2.05 x 10**
1.96 x 103
2.6 x 103
2.43 x 103
5.84 x 103
2.64 x 103
Savings
%
45
41
23
62
62
61
Includes only "wet" portions of the range
The savings potential for batch processes is not simply the volume recovery - Hot Water Energy
product
-------
5.0 SCREENING TESTS
Samples obtained from the various effluent sources were subjected
to screening tests for preliminary evaluatipn. The data were used
to guide in the selection of promising membrane/effluent combinations
for in-depth testing, to indicate problems with the particular fluids,
and to allow cursory economic-based estimates.
5.1 Facility
The screening tests were conducted in a mobile (trailer) labora-
tory on loan to Clemson by ORNL. The trailer was located at the
university during the testing activity. The laboratory was equipped
with two independent systems shown schematically in Figure 16. The
low-pressure system utilized a multistage centrifugal pump capable of
delivering 3.4 m3/hr (15 gpm) at 3.4 MPa (500 psig). Pressure, flow
rates, and temperature were controlled manually. The internal pump
parts limited the temperature of operation to 71 C (160°F),
The high-pressure loop utilized a positive displacement diaphram
feed pump. The circulation velocity was maintained with a canned-
rotor centrifugal pump capable of delivering 23 m3 /hr (100 gpm) at
1.03 MPa (150 psid) differential. Pressure was controlled automatically.
Temperature and circulation flow rates were controlled manually. The
system is constructed of 300-series stainless steel. Complete details
of the facility are given in Reference 4.
5.2 Procedure
The fluid was shipped from its source to Clemson in barrels
and was run as quickly as practical. The following procedure was
adopted for normal operating conditions.
(1) Check flux and rejection (conductivity) of each membrane
in a 100 rng/A NaCl solution. Drain system.
(2) Clean prefilters and collect a one-gallon sample of raw
feed from the waste to be run. Adjust pH of the waste if
Dahleimer, J. A., D. G. Thomas, K. A. Kraus, and J. R. Love,
"Applications of Hyperfiltration to Treatment of Municipal Sewage
Effluent," FWQA Report ORD-17030EOH01/70, 1970.
55
-------
tn
CT>
PEED
IN
FEED
IN
CONCENTRATE RETURN
PRODUCT RETURN
\ HEAT
\EXOJ
-*"*
EXOANgER
R.O.MOOULE
RO. MODULE
PRESSURIZIM6
CIRCULATING PUMP
LOW PRESSURE LOOP
CONCENTRATE RETURN
PRODUCT RETURN
R.aMODULE
R.O. MODULE
2h
BY PASS
PRESSURIZING
(POSITIVE DISPLACEMENT>
CONCENTRATE
—r-^^nxJ—»
2r*
PftOOUCT
BLEED
CONCENTRATE
'BLEED
VALVE
iTINO PUMP
HIGH PRESSURE LOOP
FIGURE 16 SCHEMATIC LOOPS IN MOBILE HYPERFILTRATION LABORATORY
-------
Pressure
800
250
75
Temperature
60 - 70
60 -
60 -
70
70
PS.
4.5 - 8.5
4.5
2.0
- 12
- 12
Velocity
10 - 15 fps
4 -
20 -
5 fps
30 fps
% Recovery
70
75
75
- 80
- 90
- 90
necessary after taking samples.
(3) Pump pH-adjusted waste through prefliters. One 55-gallon
barrel will be processed. Set up a 55-gallon barrel for
collection of composite permeate.
(4) Start system and take initial data at 0% recovery. Data
should be taken twice daily thereafter until the end of
run.
(5) At the end of the run perform a four-point temperature
excursion while recycling permeate to feed tank.
(6) Collect the last gallon of permeate before shutdown.
(7) Collect two gallons of composite permeate water.
(8) Collect two gallons of final concentrate.
(9) Flush system with tap water. Samples should be logged in
the sample logbook and labeled appropriately. Barrels of
composite permeate water should be clearly identified for
future reference.
Rejections monitored include conductivity and turbidity or color.
The typical operating parameters and prefiltration methods are
given in the table below.
Membrane
Zr-PAA
Zr-sodium
silicate
Union Carbide
Summary of samples taken:
1 gallon of raw feed
1 gallon of treated feed (if applicable)
1 gallon of final permeate
2 gallons of final concentrate
2 gallons of composite permeate
The screening test fluids were subjected to ultrafiltration or
hyperfiltration according to the most favorable estimate for applica-
tion. Three types of membranes were selected for test purposes.
(1) Union Carbide 3 NJR tubular ultrafiltration (UC)
(2) Zr-sodium silicate tubular ultrafiltration (Zr)
(3) Zr-PAA tubular hyperfiltration (Zr-PAA)
These membranes are listed in order of increasing rejection and
generally reverse of the permeate production rate. All fluids were
n°t subjected to all membranes. Table 17 indicates the combinations
tested.
Figures 17 through 21 show the results of testing in the screen-
In9 program. The data include rejection of conductivity and rejection
°r turbidity or color on cloudy or colored solutions as appropriate.
57
-------
on
CO
DESIZE WASHER
UNION CARBIDE MEMBRANE
ALTERNATOR OESIZE WASHER
PEROXIDE WASHER
too
75
u 50
K
>« 25
C
-25
-50
goo
0
u.
S600
x400(
3
_l
"•200
0
ro
y eo
a. 50
* 40
30
(
20
C
•
-
(
© /^
N^ /•)
"
0 CORRECTED TO 82'C
•
> © ©
© (
I I 1 1 1 1 I
•
-
O /
^\ '-,
-
-
0
' ©
1 1 1 1 1 1 w 1
20 40 60 8
> <
•
<
1
)
<
0 C
'
•
•
©
>
_
CONDUCTIVITY
CORRECTED TO 60*C
•
j
_
© (
0
-
(
-
-
•>
till
5 10 IS 20 2
>
1
)
3
i
S (
r v
V V
• ~
5 0
© ©
© COND.
V TURB.
II 1 l l I 1
CORRECTED TO 82*C
•
5 ©
© ©
1 1 I I I 1 i V
© G I
©
>
• —
.
ii l i i i »
> 5 10 15 20 25 30 35 4
1
>
)
)
0 (
-
V
•
P° 0 .
-------
£.oo[
J«
90
80
4OO
a
u.
(9
X
3
100
65
55
35
PEROXIDE WASHER
o
0
0
CORRECTED TO 76°C
°
0
SCAN
ZR PA A MEMBRANE
o .
©
V* I
5670
TIME (HOURS)
ALTERNATOR
WITH UNION CARBIDE PREFILTER
O
0
° O
CORRECTED TO 60° C
0
0 0
o
o o
0
1 1
10
FIGURE 18 SCREENING TESTS - HYPERFILTRATION ON OPEN WIDTH PREPARATION RANGE
-------
DYE EFFLUENT
MEMBRANE ZR-PAA
SCOUR EFFLUENT
MEMBRANE ZR-SODIUM SILICATE
DYE EFFLUENT
100
. 75
hi
j5
25
0
s
U.
S 100
3 so
<
u
l-
u
IT
K
0 20
U
o 75
50
0.
* 25,
Id V
0
r •- s B :
(NEW MODULE
) 0 s>03 ^INSTALLED
0
SOLID- COLOR REJ.
OPEN -CONDUCTIVITY REJ.
11 1 • I | ft
CORRECTED TO 65"C
)
(NEW MODULE
fINSTALLED
__ 171
0 Q
©% J3
Q 1 NEW MODULE
0^ T INSTALLED
Q
• ,T
> G
(
<
V T *'
0 0 0 0
O CONDUCTIVITY
V COLOR
•
CORRECTED TO 65°C
"
-
50 o 0 o-
.
0 Q ©
o"
5
* i ^^^^ 1 l ^
<
<
(
'
3CiJ *
-
CORRECTED TO 65°C
)
™
" TD "^
0
0 0"
O
O 0
0
^^
•\ CD
*
0 10 20 30 40 50 60 70 80 Q
10 20 30 40 50 60 70 80 0
TIME (HOURS)
5 10 15 20 25 30 35
FIGURE 19 SCREENING TESTS - HYPERFILJRATION ON ATMOSPHERIC DYE BECK
-------
ZR-PAA MEMBRANE
1-
> 100
t-
u
2 80
o
o 60
i 40
* 20
140
120
£ 100
S
80
X
3 60
b.
40
20
0
65
£ 55
a 45
III
33
25
UNION CARBIDE PREFILTER
: 1 e • a :
V MODULE 19
O MODULE 14
D MODULE 13
M
El n
9 9 90;
,
.
-
v T T
o
0.5
WASHER 3
UNION CARBIDE PREFILTER
e e
2.5 3 0
TIME (HOURS)
-1
o o o
' '
V MODULE 19
O MODULE 14
Q MODULE 13
E
V
o
FIGURE 20 SCREENING TESTS - ROPE PREPARATION RANGE
-------
ZR-PAA MEMBRANE
100
TUB 1
STD
SOLN
HOT WATER
COMPOSITE
STD
SOL'N
TUB 8
STD WATER SEAL STD
SOL'N STEAMER SOL'N
wy p
90
_. 60
Ul 5
* 70<
SO
40
160
140
2 <
o 120
100
r
x80
£ 60
40
20
o 60
40
a. r
* 201
u '
.
V
0
©
•
• • . • 1 . • 1 1
.
'CORRECTED TO 43»C
)
•
.
I
0
v 0
V
i i i . i i . i .
B a
*
9
a
(
D ^
.
1*.
O '9
V 14 .
o •
CORRECTED TO 60" C.
)
o © g
o
1 1 1 1
m
B E 0
]
88
•
.
V V
00
•
.
pc
o
o ^5 5 5 v *
0 °
. V f
CORRECTED TO 7I»C
V
_ o
,. ** © ^
G
o
0 GO-
H •
i— i m m
I'l l_J
. mm a
i
(
i
3r*
o
3Ei
«
r
) 7
- fO ~
5 "
- CORRECTED
TO 38 "C
I
• ® ® V
D 0
m D
3
, _
i- •
-
i-
j.
.
.
- -
-^ •
© •
„
m m
3
'-
M
0 Z3456789IOOI23450I234367890I234
TIME (HOURS)
FIGURE 21 SCREENING TESTS - HYPERFILTRATION ON DYE RANGE
-------
The data obtained from the temperature scans was used to adjust the
measured permeate flow to the level expected at each process tempera-
ture. The data show the actual temperature of operation and the
flux corrected for temperature to the level indicated on the figure.
In each fluid category the time variations observed are shown as the
fluid was processed toward the 70-90% recovery level. Therefore
the feed solution is becoming more concentrated as time progresses.
TABLE 17.
Fluid Source
Open-width Preparation Range
Alternator washer
Alternator washer
Desize washer
Peroxide washer
Peroxide washer
Atmospheric Dye Beck
Scour effluent
Dye effluent
Dye effluent
Rope Preparation Range
Washer #2 (desize)
Washer #3 (scour)
Dye Range
Tub #1
Hot water composite
Tub #8
Water seal on steamer
SUMMARY OF TESTS UNDERTAKEN
Membrane
UC
Zr-PAA
UC/Zr
UC
Zr-PAA
Zr
Zr
Zr-PAA
Zr-PAA
Zr-PAA
Zr-PAA
Zr-PAA
Zr-PAA
Zr-PAA
Pretreatment
None
UC as prefilter
None
None
pH neutralized
None
None
None
UC as prefilter;
pH neutralized
UC as prefilter;
pH neutralized
pH neutralized
pH neutralized
None
pH neutralized
The operation of the ultrafilter on the open-width preparation
range fluids was reasonably uneventful as shown in Figure 17. Fluxes
are reasonably high and the clarity of product water was excellent.
Some conductivity rejection was registered on the alternator due to
separation of the CMC size agent in the water. The negative rejection
shown for the second desize washer is not understood and is perhaps
spurious. The rejections recorded on the peroxide washer fluid are
low but consistent. It is not known what constituent may be responsi-
ble for conductivity rejection.
Operation of the Zr-PAA membrane on the peroxide washer effluent
(Figure 18) showed high fluxes and reasonably high rejections. Simi-
larly the operation on the ultrafiltered alternator feed was uneventful
and showed a normal flux level.
Operation of the hyperfilter on atmospheric dye waste effluent
(Figure 19) showed high color separation and low conductivity (salt)
rejection. A second membrane was installed which had improved salt
63
-------
rejection. Even during the short test period there was an obvious
flux decline in the first membrane.
The zirconium-sodium silicate membrane showed high color rejection
and high conductivity rejection for the scour fluid. The flux level
was fairly low but stable. When subjected to dye fluid, the (salt)
rejection was low as expected at the higher ionic strength, and the
flux showed a rapid decline.
Figure 20 shows the results of testing on the rope preparation
range test fluids. All monitored parameters appeared to be stable.
Both the level of rejection and flux were moderate during the test.
The dye range fluid tests are summarized in Figure 21. Operations
were within the normal range expected. The rejections were systematically
lower in tub #1 and on the water seal, probably due to the greater ionic
strength of feed. Operation on the fluid from tub #8 had an immediate
restorative effect on the rejection which had declined during exposure
to the hot water composite. Operation on tub #8 showed a decline in
permeate flux which could be minor adjustment or the start of a long
term decline, longer testing durations are required to be more
definitive. The flux levels (excluding the low temperature water
seal) are of favorable magnitude.
All fluxes shown are labeled "flux corrected to xx'C" where the
value xx is the temperature at which the processing is anticipated.
The operation of the test was, in general, not conducted at constant
temperature. At one point in the run a scan of temperature, in steps,
was imposed and flux data collected. This curve was used to correct
the flux data at actual test temperature to that designated for the
process. The following procedure was used. Using the temperature
scan data, the best value a was selected according to a plot of
In J = A - a/T
The value of a is the slope of the line on a plot of InJ vs. T"1.
Using the value of a for a given fluid and membrange, the value of
J(To) which is the flux corrected to temperature T is determined from
J(T0) = J exp {+ a £ - -^] }.
T TO
Here J is the observed flux at temperature T. Presumably the value
of J(TQ) only depends upon pressure, time, etc., without dependence
on temperature. Common values of 2200 to 2900 (K"1) are observed.
Use of such a procedure allows a plot of data to reveal trends
which are not due to temperature effects. For example, assuming a
value of o = 2500 to fit the data, an observed flux of 1 x 10~5 m/s
at 41 C (314K) corrected to 66 C (339K) would be
J (339K) - 1 x 10~5 exp [2500 (— —) ] = 1.8 x 10"5 m/s.
314 339
64
-------
The corrected flux is that flux which would be expected in a duplicate
experiment run at temperature TQ.
65
-------
6.0 FIELD TESTING
Because experience had indicated the need to process waste as
it emanates from its particular source, a small movable test unit
was constructed. The unit was designed to provide some automatic
process controls while others were omitted. The operating parameters
pressure, temperature, pH, and velocity could each be controlled
independently. We selected to provide automatic control for pH and
pressure while depending on hand-operated valves to maintain relatively
fixed levels of velocity and temperature. The unit was designed to
include an operation package and an instrument package. The operational
package was set up to perform all basic functions without the presence
of the instrument package. Also the operational package was supplied
with safety cutoffs for out-of-range conditions.
A schematic of the skid-mounted test unit is shown in Figure 22.
Waste feed is pumped through a float valve which maintains the level
of fluid in the reservior. It is mixed in the reservoir with concen-
trate returning from the module and pH correction fluid (as applicable)
from the reagent tank. The mixed fluid enters a small boost pump
which forces flow through the particulate filter and supplies the
net positive suction head of the high-pressure pump. High-pressure
flow passes through the heat exchanger where its temperature is
moderated by cooling water. The flow then is split between the primary
module flow and a bypass flow which allows velocity control in the
module. A turbine meter in the module flow circuit allows measurement
of flow. The rejoined flow then normally passes through the automatic
pressure control valve and is split into a small concentrate flow and
the concentrate which returns to the reservoir. The module produces a
permeate flow of a few tenths of dm^/min and the through flow ranges
upward from 10 dm3/min. Therefore, the module exit concentration
is about only 1% greater than the inlet concentration. The module
exit flow and the bypass flow and the feed tank all have essentially
the same concentration at any time. Changes in concentration are
only produced slowly by control of the relative amount of concentrate
and permeate released.
The dimensions of the unit are: length = 2.3 m (7 feet); width =
1.07 m (3.5 feet); and height approximately 1.52 m (5 feet) to the
highest point. The dimensions have been minimized for easiest
incorporation into on-site locations. Besides the waste, the unit
requires approximately 8 liters per minute (2 gpm) of coolant flow
and 230 volt, 3 phase, four wire electrical power of 50 amperes.
66
-------
en
Concentrate ~ *~&
Level
Contro
Waste -^p— Cxj-
Feed .
Overflow J
Feed — "
Tank
i
A
1 T
I
•****?•
Air
Blee
1
jProduct
£ Module
tea gent
Tank — s^. •
L_J a
r^
T *WI^
]Pump *
^
j Booster Particulate
I Hump Miter
Lege
Dra i n
d
« Tap
H20
Heat
Exchange r
\ '
t i
Drain
Low
Point
Drain
sure
I
[7
nd
Legend
n Automatic Hi
^ Pressure Valve
Pressure o
Control Valve ***
Regulating Valve
M
No Flow- Full Flow Valve XI
FIGURE 22 FLOW SCHEMATIC OF SKID MOUNTED UNIT
-------
The modules consisted of a bank of ceramic tube modules and a
module incorporating stainless steel tubes. The modules have
dynamically formed ZrO-PAA membranes.
The basic ceramic tube module is shown in Figure 23. Each
element of the module consists of seven tubes individually potted
into a plug which will thread into a pipe tee. The free ends of the
tubes are sealed, and the potted ends permit flow into a filtrate
chamber. This flow in turn can be readily collected with adjacent
filtrate flows. Flow passes along the tubes within a pipe fitted
to the tee outside of the ceramic tubes. The feed can flow in
either direction. One candidate method of joining the elements is
shown in Figure 23. This method is reasonably economical and
relatively easy to assemble. An assembly of 30 elements is equiva-
lent to about 0.87 m2 of membrane area.
The stainless steel module assembly is shown in Figure 24. In
this arrangement the pressure is applied within the tubes and the
permeate is collected by the shell. Feed flow enters one tube, is
routed to another tube by appropriate cutouts in the flow routing
sheet, and finally exists the opposite end. Each tube is sealed
by an "O" ring on a tube end piece welded to the filter support
tube. Flat gaskets prevent flow between adjacent channels, while
flow out of the system is prevented by an "0" seal on the faces
of the routing sheet. The three-piece bundle end is bolted together
(not indicated by the figure), and -the ends are joined by rods which
are in slight tension. The membrane area is 0.95 m2 for the module.
An ultrafiltration skid was constructed having the same general
description as the hyperfiltration skid. The pump which was found
unsatisfactory for hyperfiltration service was used with a one horse-
power motor. A variable area flowmeter was located downstream of the
test section which was bypassed by an appropriately valved line allowing
flow control up to 42 dnr/min (11 gpm) through the test section. Pressure
readouts were located upstream and downstream of the test section. The
feed was filtered through a 25 micron cartridge filter and added to a
stainless steel tank normally maintained at a 76 dm3 (20 gallon) level.
All metal parts were 300 series stainless steel with connections made
of industrial rubber hose.
No heat exchanger was provided since at the lower pressure level of
operation the energy buildup was slight. However, this lack of energy
buildup proved to be some problem since in low throughput cases the
performance data were obtained at low temperature making interpretation
difficult. The throughput could be increased by having permeate flow
all pass to the drain or increasing the flow of concentrate to the
drain. With higher throughput rates the feed barrel would be exhausted
soon after a feed flow interruption. The source fluid was interrupted
frequently on the dye range in association with shade changes. The
automatic low level cutoff would shut down the unit and, when
unattended, the operation would be discontinued.
68
-------
Filtrate
j
c
Potting Material
7 Tubes With
Sealed Ends
JLLS / / / / t ' ; 1 / / i I t i i
Pipe Section
Feed in (Out)
CERAMIC TUBE MODULE ELEMENT
Feed Out (In)
Product
Collection
Manifold
In
Coupling
FIGURE 75 CERAMIC TUBE MODULE ARRANGEMENT
Method of Joining Elements
to Assemble Module
-------
Product
Collection
Shell
(Reference)
Flat
Gaskets
iz*
f.-.v.^'-.--.^
\LZ7f7_L
IT \\\m\
Tube Header
Sheet
End Cap
Flow Kouting
Sheet
Basic Hexagonal
Arrangement
13 Tubes
FIGURE 24 STAINLESS STEEL TUBE MODULE
-------
As shown in Figure 25, the three modules were located in series,
each having similar flow requirements, in order to use one flow meter.
The UC 3 NJR and ZrO permeate flows are less than 1% of the feed flow
so that each module was exposed to the same fluid. The permeate
from the Union Carbide and ZrO dynamic membrane were normally
diverted to drain, while the larger Abcor module permeate flow was
returned to the feed. The latter permeate return arrangement was
decided after experiencing considerable difficulty in maintaining
feed sufficient to keep up with the permeate flow. During attended
runs the permeate flow was diverted to drain as required to establish
the desired operating conditions.
6.1 Operation of Atmospheric Beck Dye Fluid
The skid unit was located adjacent to a dye beck together with
a feed reservoir. The feed reservoir was comprised of four inter-
connected vessels each formed by welding a pair of 55-gallon drums
end-to-end. The drums were painted with epoxy paint to alleviate
corrosion and the insertion of corrosion products into the feed
stream. The normal procedure for filling the feed reservoir was to
pump beck fluid from a location near the drain as the beck was
undergoing rinse/overflow. On some occasions the operator pumped
undiluted dye liquor and in others, only a "topping off" of the
feed reservoir was required. In both of these situations a more
concentrated feed was delivered to the membrane unit. New fluid
was added at essentially each dyeing over the test period.
The tubular stainless steel module was installed on a skid
and tested for operational acceptance. Table 18 shows a skeleton
of data obtained. After two weeks of operation (233 hours) the
flux had degenerated to below 20% of its original value. The module
was removed and replaced with a ceramic tube bundle with the resolve
to carefully monitor for the presence of metal ions which could be
responsible for flux decline.
Starting from January 26, 1977, twelve 7-tube ceramic bundles
were operated for 502.4 hours which was 46.5% of the total time
between January 26 and March 11, 1977. Figure 26 summarizes the
test conditions for the period. Figure 27 shows flux versus time
where the flux has been corrected to 66 C using a procedure outlined
in the screening test section.
The unit was operated to provide approximately half concen-
trate and half product for a period of about 100 hours. The
temperatures and thereby the permeate flux varied considerably,
while the conductivity rejection remained fairly constant at
about 80%. After the 100-hour point, the concentrate flow was
reduced allowing the membranes to be exposed to more concentrated
material as evidenced by the higher total solids and concentrate
conductivity levels in Figure 26. The curve of Figure 27 as expected
71
-------
25 micron
filtered
feed
I
CP
BYPASS
UCSNJR)—1 zro |—IABCOR]—'—cxH
~~M~~^ ^
NORMALLY
TO
DRAIN
NORMALLY
RETURN
TO
FEED
TANK
FLOW
FIGURE 25 ULTRAFILTRATION SKID SCHEMATIC FLOW DIAGRAM
-------
-J
w
TABLE 18. MEMBRANE OPERATION ON DYE WASTE FROM ATMOSPHERIC BECK
PressureTemperatureFluxConductivity
. MN/m2 (°C) m3/m2-s Rejection
11/29/76 Tubular stainless steel module installed.0.95 mz
NaNO^ solution prior to exposure to waste.
5.5 (800 psig) 50 2.03 x 10~5 (43 GFD)
12/01/76 Dye waste solution.
5-9 75 1.23 x 10~5 (26 GFD) 67
12/16/76 - <38 0.24 x 10~5 (5 GFD) 90
01/12/77 Ceramic 0.307 m2 module installed.
NaNO3 solution
5-5 43 2.03 x 10~5 (43 GFD) 87
Dye waste solution initially.
5-5 52 2.6 x 10~5 (55 GFD) 58a
a Feed conductivity 16,000 ymho/cm.
-------
5000
zoo 2500
OQ,
1001-
I-HUJ
000
50
o
00,
o°«oo'
oo<> o oo
o
o o oo
250001-
10 20000
E
-------
Flux corrected to 66 C
•°6 '-5t
X
^
*• 1.0+
O
t 0.5-
o
o
orffc o
p o o
100
200 300 400
Operating Time (hours)
500
FIGURE 27 PERMEATE FLOW RESULTS - ATMOSPHERIC DYE BECK
75
-------
shows little scatter due to temperature variation but does indicate
the severe flux decline. The membrane appeared to stabilize in flux
until 100 hours, whereafter the flux declined steadily to a level of
about 2.8 m3/m2-s (6 GFD).
Seventy-two samples were taken during the period January 26
to March 11, 1977. These consisted of twenty-four each of concen-
trate, product, and feed. Analyses of these samples is shown in
Table 19. The data in the table correspond to the period of testing
shown in Figure 26. The samples 2022, 2023, and 2024 were extracted
at about 109 hours, i.e., near the onset of the flux decline period.
The only apparently significant data relative to the flux decline is
that the turbidity of both feed and concentrate had jumped to "opaque"
from previously being in the instrument's range. Neither total solids,
calcium, or magnesium registered a significant increase near the point.
A separate decline test was performed on February 15 through
February 18, 1977, using a previously unexposed membrane designated
bundle L. The result is shown in Figure 28. Flux is corrected to
66 C using data from the temperature scan test performed on February
7, 1977.
Because of the correlation shown in Figure 29 between the low
flux levels and the elevation in chromium and copper in the test
fluid, mechanisms for the flux decline were sought. Laboratory
testing confirms that copper will produce flux declines but not
similar in concentration effects to those shown in Figure 29. During
the period of testing the cause of flux decline was not isolated
and remains to be studied. A later section of this report shows a
contrasting relative effect of metals on flux to Figure 29.
A number of test dyeings were conducted using concentrated dye fluid
from the hyperfiltration unit. The concentrated material was a mixture of
a number of runs and was used in test dyeings matching several shades.
Measurements of the dye formula required to bring the dyeing to shade were
compared with standard formulations. In some cases less dye was required;
in other cases, more. Little or no economic advantage is apparent; however,
reuse of the concentrated dye has a decided environmental advantage.
In conducting tests for reuse the following procedure was followed.
The dyes were added to a beaker containing 1% Duodye and the solution
mixed well. The fabric was then introduced into the beaker and the
solution was heated to 93 C for one hour. The fabric was then removed
from the bath, rinsed, and dryed.
Samples of concentrate dyebath wastewater obtained from the skid
hyperfiltration unit were evaluated in the plant laboratory for dye
content and value for producing commercial shades. Table 20 shows
data for the amount of dye required to adjust the concentrate to give
a commercial shade. However, the average amount added was less than
normal. Most important though was the fact that the dye concentrate was
determined to be reusable in dyeing a variety of test shades.
76
-------
TABLE 19. ANALYSTS OF SAMPLES FROM HYPERFILTRATION OF DYE BECK FLUID
Sample
#
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
Date
Taken
1/26/77
1/26/77
1/26/77
1/27/77
1/27/77
1/27/77
1/28/77
1/28/77
1/28/77
1/31/77
1/31/77
1/31/77
2/02/77
2/02/77
2/02/77
2/03/77
2/03/77
2/03/77
2/04/77
2/04/77
2/04/77
2/07/77
2/07/77
2/07/77
2/08/77
2/08/77
2/08/77
2/09/77
2/09/77
2/09/77
2/15/77
2/15/77
2/15/77
2/16/77
Type
F
P
C
F
P
C
P
C
F
F
P
C
F
P
C
F
P
C
F
P
C
F
P
C
F
P
C
F
P
C
F
P
C
F
Conductivity
(uroho/cm)
1940
520
5000
1500
8600
2330
12000
3500
_
1820
7600
2820
1780
6450
2150
590
_
2050
435
3800
1380
1120
5950
1400
1520
5900
_
_
9100
830
7600
9100
% Rejection
(Conductivity)
_
89.6
81.9
80.5
76.0
—
72.4
..
_
88.5
81.2
74.2
-
_
89.0
—
Turbidity
(FTU)
1.20
0.64
31.00
0.51
4.10
0.37
2.20
2.70
0.42
2.10
5.20
0.36
1.70
0.56
0.27
opaque
0.24
opaque
opaque
0.27
opaque
opaque
0.21
opaque
opaque
1.30
opaque
opaque
pH
6.48
6.61
6.42
6.80
6.58
6.78
7.05
ft qc;
U • J ^
fi Q9
u • 7^
7 ^n
/ « j \j
6QO
• j\j
f. qc.
V . .7 ^
7.39
7 nfi
/ • VJ O
600
• oo
6.65
fi QT
\J . ^ .j
6.59
6.73
6.85
6.70
7.00
7 "3R
/ • «J O
7.22
7.51
7 IT
f • .L J_
7.51
7.60
-------
TABLE 19. (continued)
oo
Sample
#
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
Date
Taken
2/1 6/77
2/16/77
2/18/77
2/18/77
2/18/77
2/28/77
2/28/77
2/28/77
3/01/77
3/01/77
3/01/77
3/02/77
3/02/77
3/02/77
3/03/77
3/03/77
3/03/77
3/04/77
3/04/77
3/04/77
3/07/77
3/07/77
3/07/77
3/08/77
3/08/77
3/08/77
3/09/77
3/09/77
3/09/77
3/10/77
3/10/77
3/10/77
3/11/77
3/11/77
3/11/77
Type
P
C
F
P
C
F
P
C
F
P
C
F
P
C
F
P
C
F
P
C
F
P
C
F
P
C
F
P
C
F
P
C
F
P
C
Conductivity
(pmho/cm)
1200
9800
9500
3600
24500
8400
3800
24000
8190
4150
25000
8400
4800
27200
8300
5400
30000
8100
6100
28000
7050
5550
28900
4750
5450
26500
3600
4700
28000
3350
4800
27000
3120
5000
25000
% Rejection
(Conductivity)
87.8
85.3
84.2
83.4
82.4
82.0
78.2
80.5
79.4
83.2
82.2
80.0
Turbidity
(FTU)
0.88
opaque
opaque
0.47
opaque
0.36
opaque
opaque
0.53
opaque
opaque
n
-------
VD
Sample
#
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
2025
2026
2027
2028
2029
2030
2031
2032
2033
2034
2035
2036
2037
2038
410 ym
87.9
92.8
81.5
82.8
94.9
84.1
92.5
75.2
83.1
86.9
93.3
87.0
89.2
96.2
87.9
93.8
94.9
86.5
91.9
98.2
87.1
93.6
97.0
82.9
93.1
99.0
74.1
-
—
-
4.0
95.0
35.0
4.7
96.5
Color
500 ym
87.3
96.8
88.3
89.1
96.2
92.1
97.5
90.5
95.2
92.9
95.8
93.5
93.2
97.2
93.4
91.9
92.9
91.8
91.9
99.8
88.5
95.0
99.0
83,6
95.2
100.0
76.0
—
—
—
0.5
96.0
50.0
1.5
98.9
600 ym
98.0
95.7
93.8
95.2
98.9
95.1
99.2
95.8
98.5
96.5
96.2
97.3
97.0
99.3
99.0
95.5
98.4
96.9
93.9
100.0
94.0
97.5
100.0
89.0
97.3
100.0
83.4
—
_
_
43.5
92.1
68.5
42.3
99.8
TABLE
Total
Solids
1055
308
3203
5300
809
5277
1246
7044
1983
5366
939
5303
1701
955
4490
992
139
2839
1331
243
2696
673
535
3697
684
744
3760
_
—
_
6450
462
5168
6741
449
19. f continued)
Copper Iron
(mg/£} (mg/S,)
-
_
—
—
—
_
—
_
_
—
_
—
_
—
<0.20
<0.20
0.00
<0.20
<0.20
<0.20
<0.20
<0.20
0.00
_
_
_
0.42
0.20
3.30
0.26
0.18
-
_
—
_
_
—
—
_
—
—
_
_
_
—
1.20
0.00
0.24
0.20
0.04
0.28
0.16
0.08
0.51
^
0.04
0.00
0 . 23
0.05
0.01
Hardness
Ca & Mg
(mg/Z)
21.0
1.6
29.9
42.3
3.3
42.3
7.9
70.7
17.9
58.1
2.9
57.8
16.7
4.0
60.6
16.5
2.5
40.0
18.6
1.0
79.0
11.9
1.6
63.7
11.9
2.1
68.1
_
20.9
4.5
76.3
21.1
4.6
Calcium
(mg/£)
5.4
nd
0.6
14.0
nd
14.6
2.1
34.4
8.0
31.0
2.1
30.8
8.4
2.6
30.5
~m
10.4
0.4
48.5
8.0
2.6
51.1
12.6
1.2
53.1
1.2
_
1.5
Alkalinity
(mg/2, CaCOs
19.3
8.7
56.2
15 8
^ *J • \J
59 4
— ' ^ • T
18.8
58. 8
23 9
*» •** • ^
48.6
17.2
42 4
T fc« • *T
14 9
-L-* • _y
15.5
32 1
J £* • -i-
21.3
8 7
\J • /
20.2
18.5
18.5
34.6
17.9
7 7
* * /
50.3
15.6
19.6
48.7
115.5
16.9
65. 5
118 9
^Jm \J * J
19.8
-------
00
o
Sample
ft
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
2050
2051
2052
2053
2054
2055
2056
2057
2058
2059
2060
2061
2062
2063
2064
2065
2066
2067
2068
2069
2070
2071
2072
410 ym
11.5
4.1
98.5
00.0
7.0
97.1
00.0
7.5
96.3
00.0
11.1
0.5
00.0
0.3
90.4
00.0
0.3
80.3
00.0
5.9
96.2
00.0
17.9
96.9
00.0
32.4
97.1
00.0
35.4
95.5
00.0
36.3
97.2
00.0
Color
500 ym
13.9
0.5
99.1
00.0
2.1
97.5
00.0
2.5
97.8
00.0
5.2
93.0
00.0
00.0
96.1
00.0
00.0
82.2
00.0
1.8
96.8
00.0
9.8
97.3
00.0
21.9
98.8
00.0
24.7
97.8
00.0
27.9
98.3
00.0
600 ym
55.0
39.1
99.5
12.0
41.0
100.0
7.0
41.0
99.3
6.3
47.7
96.1
5.0
29.9
99.2
1.7
28.9
94.8
2.6
38.7
97.9
1.9
58.1
99.5
2.9
70.7
99.7
2.3
72.3
97.8
2.2
74.1
99.3
1.9
TABLE
Total
Solids
(mg/Jl)
6824
6454
2042
17524
5675
2335
17189
5670
2385
18695
5710
2815
20546
5690
3272
22429
5680
3867
20910
4813
3521
23116
3180
.3359
21398
2182
2860
21711
2011
2899
21606
1794
2909
19948
19. (continued)
Copper Iron
(mg/4) (mg/A)
2.61
1.05
0.20
3.60
1.50
0.21
2.64
1.30
0.21
4.32
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
0.27
0.09
0.01
0.40
0.08
0.01
0.50
0.47
<0.04
0.08
Hardness
Ca & Mg
(mg/Jl)
106.8
14.0
2.3
110.0
27.7
3.3
110. 0
114.5
5.0
28.8
23.7
3.7
122.5
23.6
5.1
134.9
23.1
10.9
127.7
20.8
3.5
148.5
14.1
3.0
132.5
3.4
9.3
140.2
8.5
2.8
143.8
8.0
5.6
139.0
Calcium
(mg/Jl)
-
_
1.4
-
-
0.8
—
-
1.1
-
—
1.0
-
-'
1.7
-
—
2.4
—
-
2.1
-
—
1.6
30.9
9.2
2.8
53.9
6.6
1.5
61.3
3.8
2.8
61.7
Alkalinity
(mg/Jl CaCO^)
88.5
112.6
36.6
241.6
105.2
48.5
242.6
104.6
42.7
271.9
107.2
47.4
294.0
105.2
57.3
322.8
110.3
65.7
310.0
94.5
71.0
357.5
66.5
59.7
327.5
47.6
51.1
334.6
43.6
51.9
325.2
40.3
52.5
298.9
-------
00
IOO
80
60
40
20
0
2.5
o
7 2.0
evil
IOL 1.5
X
3 1.0
u.
0.5
Flux corrected to 66 C
a
o
cP
00
OD
OO
10 15 20 25 30 35 40 45 50 55 60 65 70 75
Operating Time (hours)
FIGURE 28 RESULTS OF FLUX DECLINE STUDY WITH ATMOSPHERIC DYE BECK FLUID
80
-------
30
CD
NJ
10
o>
20
o:
LJ
o
o
o
o
10
O CHROMIUM CONCENTRATfON
COPPER CONCENTRATION
A FLUX CORRECTED TO 66° C.
100
500
200 300 400
OPERATING TIME, HOURS
FIGURE 29 COPPER AND CHROMIUM EFFECT ON FLUX BEHAVIOR
-------
TABLE 20. DYE ADDITIONS FOR REUSE TESTS
Chemical
2/09/77
Direct dye
Direct dye
Direct dye
Salt
1/22/76
Direct dye
Direct dye
Direct dye
Salt
1/30/76
Direct dye
Direct dye
Direct dye
Salt
3/11/77
Direct dye
Direct dye
Direct dye
Salt
3/04/77
Direct dye
Direct dye
Direct dye
Salt
3/08/77
Direct dye
Direct dye
Direct dye
Salt
2/17/77
Direct dye
Direct dye
Salt
Dye Name
Superlitefast Yellow EFC
Superlitefast Blue 8GLN
Direct Fast Red 8BLN
Superlitefast Blue 8GLN
Superlitefast Yellow EFC
Direct Fast Blue 4GL
Superlitefast Blue 8GLN
Superlitefast Yellow EFC
Direct Fast Blue 4GL
Diazol Light Orange 5JA
Direct Fast Red 8BLN
Superlitefast Blue 8GLN
Diazol Light Orange 5JA
Direct Fast Red 8BLN
Superlitefast Blue 8GLN
Diazol Light Orange 5JA
Direct Fast Red 8BLN
Superlitefast Blue 8 GLN
Superlitefast Rubine WLKS
Superlitefast Yellow EFC
Dye Used Based on
Test Dyeinq
0.0420
0.0050
0.0004
25.0000
25.0000
0.0040
0.0500
25.0000
0.1500
0.1250
0.0300
50.0000
0.1350
0.1500
50.0000
0.1350
0.1500
50.0000
0.1000
0.0100
25.0000
% Fabric Weight
Normal Dveincr
0.0070
0.0500
0.0030
25.0000
0.2250
0.0020
0.0400
25.0000
0.2250
0.0020
0.0400
50.0000
0.0550
0.0850
0.0110
50.0000
0.0550
0.0850
0.0110
50.0000
0.0550
0.0850
0.0110
50.0000
0.4500
0.0180
50.0000
6.2
Operation on Dye Range
The skid was operated on the continuous dye range during the
Periods indicated in Table 21. The fluids in both washbox #1 and
washbox #8 were selected for testing. In the interim period from
April 20, 1977, and May 9, 1977, a new pump was installed in the
skid due to the numerous difficulties encountered with the previous
Pump. The refurbished skid was returned on May 9 and was operated
°n the first dye washer {washbox #1) and on the soaper washer
(washbox #8).
83
-------
TABLE 21. SUMMARY OF TESTING OF HYPERFILTRATION ON DYE RANGE
Source
(see Figure 4)
Washbox #1
Washbox #1
Washbox #8
Membrane
SS Zr-PAA - 0.95 mz
Replace Pump
Ceramic Zr-PAA - 0.012 m2
SS Zr-PAA - 0.95 m2
Period
4/1-4/20
5/9-5/23
5/27-6/9
Hours
120
220
132
Samples
4001-4029
4030-4070
4071-4087
pH Adjustment
pH = 11 to pRi
pH = 11 to pH-1
not adjusted
o 7
CD
-------
Conditions during operation on the fluid from washbox #1 are
shown in Figures 30 and 31. Unexpected conditions resulted due to
the pH adjustment and the periodic shade changes on the range. The
skid was protected against running below about pH=2 (operator setting)
and acetic acid was used to neutralize the feed to 7 < pH < 8 during
operation, in anticipation of achieving optimum rejection. When shade
changes occurred on the range, the washboxes were purged, thus
interrupting the supply of feed to the skid. The feed' would be
processed normally until the low level cutoff interrupted the operation.
The unit when stationary was such that acid added to the feed tank did
not register a pH change since the pH indicator was located in the flow
circuit, not in the tank. Acid would be continually added during the
inoperative period. Upon restart of the range and restoration of feed,
the operator would start the skid. Very quickly the excess acid would
cause the pH to register below the acceptable limit (usually 2) and the
unit would shut down. New fluid was added to the feed tank after
draining the system and operation could be resumed.
The foregoing is detailed to suggest a cause for the flux
behavior shown in Figure 31. The most probable explanation is that
the membrane exposure to acid solutions "loosens" its structure allowing
a temporary flux increase. We observe, during formation, that sudden
PH steps, from 3 to 7 for example, result in a greater flux than a
gradual transition of unit pH steps. Usually the rejection is observed
to be lower also following large pH steps. One other explanation is
that flux reducing multivalent metal ions may be released from the
membrane during excursions to low pH. Separate laboratory tests fail
to confirm that this mechanism is probable, at least with copper.
Recovery of flux after exposure to copper is not readily done by
lowering of pH alone.
It is somewhat speculative what level of flux may be attained
without the intervening pH excursions, but we judge from the data
of Figure 31 that 2 m3/m2-day is realistic at 65.5 C. However, the
probable temperature of operation is only 45 C so that a correction
to 1.25 m3/m2-day as a design flux is reasonable. All testing was
Performed between 4.8 and 5.2 x 106 N/m2 (700 to 750 psi) and the
Permeate flux is expected to vary linearly with pressure.
The data indicated in Figure 29 for the atmospheric dye situation
and that in Figure 31 compare the response of flux to the metal ion
concentration. In the atmospheric beck dye case, the metals
are part of the dye molecules and also may be present in the diluent.
In the dye range case the source of chromium ions is not in the
Particular dyes used but may have been leaching from the stainless
steel during low pH conditions. In any event the presence of comparable
°r greater metal ion concentration did not produce a comparable flux
decline.
Following operation on the first washbox the unit was connected
to waste from washbox #8, termed the "soaper" because it is the
first washer after saturation with soap or detergent. The membrane
85
-------
30000
> E 20000
PO =>,
OO
10000
t
0
100
1-h-UJ
ooo
838
8* 50
OD
cn
o o
o
O
O
O O
00
10
6 30000
o>
V)
-. 20000
o
CO
<
h-
O
10000
CONCENTRATE
>
PERMEATE
STAINLESS STEEL ZR-PAA
I i I i I 1
CERAMIC ZR PAA
20 40 60 60 100 120
OPERATING TIME-HOURS
KOC
I
I
20 40 60 80 100 120 140 160 180 200
OPERATING TIME-HOURS
FIGURE 30 FIELD RESULTS - HYPERFILTRATION ON DYE WASHER
-------
00
SELAS ZR
O CHROMIUM CONCENTRATION
COPPER CONCENTRATION
FLUX CORRECTED TO 65.6° C
too o ioo
OPERATING TIME, HOURS
200
FIGURE 31 COPPER AND CHROMIUM COMPARISON TO FLUX
-------
was operated for 132 hours from May 27, 1977, to June 9, 1977. The
fluid had pH of approximately 3.8 and was not adjusted. Because of
this low pH operation the rejection of ionic material (conductivity)
was only from 30 to 50%. Figure 32 shows the conductivity and total
solids variation observed during the operating period. The unit was
operated at a volume recovery level of up to 0.9 during the testing.
Near the end of testing, even though the conductivity rejection was
low, the total solids rejection was approaching 90%. Most of the
rejected material probably was comprised of soap carryover.
Figure 33 shows the flux response to temperature as darkened
symbols together with the bulk of data taken on the soaper feed. As
with most other feeds the relation of log flux to reciprocal tempera-
ture is a straight line with negative slope. The flux temperature
trend applied to each point predicts the performance to 76 C and
is plotted versus time in Figure 34. Even though the fluxes do vary
20% the overall trend is quite stable and fluxes of at least 4 m3/m2-
day are expected.
A variety of tests were conducted on water from the hyperfiltration
unit. The tests of Table 22 comprise the basis for reuse evaluation.
Obviously these tests must also be combined with experience for the
determination of reusability of the various streams. The tests were
evaluated carefully for a navy shade for which the unit was pre-flushed
and operated to yield 91% recovery. The permeate produced a slight
coloration and some dulling effect on the blend of cotton/polyester.
The effect was very slight on cotton material. In the filter test
of concentrate the color was strong on the filter and the filter
plugged, but the permeate produced no color on the filter material
or plugging. The concentrate produced a bird egg blue (less than
10% of original shade) when used alone for a test dyeing. Test dyeing
using the recovered dye in the concentrate augmented by additional
dye produced a non-distorted color which was fast. Test dyeings with
permeate water were conducted and showed no distortion and dye fastness.
Testing was also carried out on permeate and concentrate from
operation on the soaper washer during several shades. As before, the
permeate showed no color on the filter and this time no effect on
cloth exposure. The concentrate contained colored material which
stained the filter. Test dyeings on permeate were conducted with
good dye fastness and no distortion. Some loss of brightening compared
to soda ash and wetting agent was detected, though very slight. The
concentrate and permeate were used for the soap test of Table 21. The
permeate ribbon did not sink by five minutes, the concentrate ribbon
sank in 8.5 seconds, and the soap standard solution produced sinking
in two to five seconds. Foaming of eight-ounce samples in a blender
produced foam volumes of: permeate - 473 ml (16 ounces); concentrate -
621 ml (21 ounces).
The laboratory personnel offered their summary judgments that
(1) the permeate water was adequate for reuse in dyeing or washing,
(2) some dye recovery was evident though the amount was too small to
88
-------
1500-
"JDUCTIVIl
EJECTION
PERCENT
60
20
0
L-_-xT^^-^>— —o—^
-
cc
5000L
10 4000 -
•v
o»
£ 3000
2000-
20 40 60 80 100 120
OPERATING TIME-HOURS
140
FIGURE 32 MEMBRANE EXPOSURE - SOAPER FEED ON DYE RANGE*
-------
0.80
0.70
0.60
-
•o
w
0.40
«T
x 0.30
0.20
o
o
Q 0
.,0-
SS ZR-PAA
PRESSURE 4.9MN/IH2
05
J
O NUMBER SEQUENCE
• TEMP, SCAN AT END
OF TEST
O I
2.9 3.0 3.1 3.2
|/T°KxlO*3 RECIPROCAL TEMPERATURE
FIGURE.33 FLUX AND TEMPERATURE - SOAPER FEED ON DYE RANGE
90
-------
10
$ 8
T*
I
cvj
•I7
u.
Id 5
t-
u
UJ
tt 4
O
u
FLUX CORRECTED TO 76°C
USING DATA TREND
J I I I
20 40 60 80 100 120 140
OPERATING TIME HOURS
FIGURE 34 FLUX RESULTS - SOAPER FEED ON DYE RANGE
-------
TABLE 22. DESCRIPTION OF DYE LABORATORY REUSE TESTING
Test
Exposure
Procedure
Interpretation
vo
NJ
Test dyeing color match
Fastness
Soap
Filter
Foaming
Expose cloth to solution,
determine coloration of
sample.
Add dye to solution to
determine amount required
to produce standard shade on
cloth sample.
Evaluate color fastness to
both light and wash influence.
Weighed ribbon exposed to
solution. Time required to
sink (cease floating) measured.
Millipore filter examined for
coloration following solution
passage.
Add solution to blender and
operated blender.
Excessive coloration of sample
indicates excessive impurities
for quality dyeing.
Indicates potential for dye material
recovery.
Indicates whether any reclaimed dye
in the color match tests meets fast-
ness standards.
Soap or detergent will promote
wetting and will reduce the time to
sink.
Coloration of filter proportional
to amount of dye agglomerates
present.
Measure foam volume produced.
-------
be economical, and (3) some soaper feed dragout was recovered, also
too slight to be economical. We feel that the judgment of the
economical is premature in that recovery to 90% means that the
recovered chemicals are still volumetrically dilute. We believe the
lab personnel judgment to be made on the basis of concentration, not
on the rate of flow of recovered mass. One factor not considered in
the dye recovery is that of the difficulty in segregating shades
during normal operation involving shade changes perhaps every four
to six hours. Since the soaper feed is not involved with this
problem, its concentrate recovery should be easier technically. The
buildup of color in recycled soaper concentrate could present a problem.
Methods such as filtration, ultrafiltration, or simply blowdown could
be used to separate or reduce offending color from recycled soaper
feed if it becomes a problem.
6.3 Operation on Preparation Range Fluids
The hyperfiltration unit was operated on each of the three
drains on the open-width preparation range. At the time of operation
of the unit, the range was operated on polyester/cotton material
having warp size composed of a starch and polyvinyl alcohol (PVA)
mixture. This size contrasts with the formulation current with the
energy survey and analysis of the drain. Otherwise the operation
is similar, though the chemical additions are also milder. Overall
the movable unit was on site from June 13, 1977, through August 26,
1977, and operated for over 360 hours combined on the three streams.
A minimum of information was gathered on the alternator (water
washer) since pilot units are operating on such effluent and one
full-scale recovery unit is in operation. However after forty-odd
hours of operation, conditions were varied to provide data on the
dependence of water flux on velocity and pressure. Figures 35 and 36
show the results obtained for high and moderate recovery situations.
In the first part of Figure 35 four data observations at various
velocity conditions confirm the expectation that flux varies with
velocity to a power essentially 0.8 in turbulent flow. The single
Point falling off the trend is considered to be a non-equilibrium
observation. During the operation on the range, it was common for
the permeate flow to be unsteady. This unsteadiness is presumed to
be due to a delicate balance between gravity and surface forces
within the cavity of the module. Some other flow observations are
much higher that could be consistent with all other observations at
such conditions. The suspect data have been included even though
they are felt to be observed in error.
In the situation where a high solids content of poorly
diffusing solute is present to the extent that velocity effects
are apparent, the flow resistance of the membrane is augmented by
the resistance of the layer or rejected material. When the rejected
93
-------
1.0
0.8
FLUX 0.6
y
0.4
0.2
m /m day
J«V
0.8
T = 74C
P = 2flMN/m2 (400psi)
pH=6
90 < RECOVERY < 94
01 L
234 56
VELOCITY m/sc
L5
FLUX ,.o
mvm2 day
0.5
0
-
o
0
o
g Vel=5.5 m/s
I 0 T = 74C
0 500 psi (ref)
1 1 L_LL i i
0 t 2345,6
PRESSURE MN/m2
FIGURE 35 HELD RESULTS- HYPERFILTOATION ON PREPARATION
RANGE WATER - HIGH CONCENTRATION
94
-------
RECOVERY =63%
ZO-PAA MEMBRANE
4
-J-
FLUX 3
^/m2 day
T=82C j
T=77C
Vel
V5.5m/s
O 4.9 m/s
03.6m/s
i
T5.5m/8
• 3.6 m/8
- FLUX TREND
AT ZERO
RECOVERY
5.5m/s, 82C
V
5.5m/s,77C
•
3.6m/s
500 psi (ref)
/»wv
234
PRESSURE-MN/m2
FIGURE 36 FIELD RESULTS - HYPERFILTRATION OF PREPARATION
RANGE WATER WASHER - MODERATE CONCENTRATION
95
-------
layer is sufficiently great in influence, the flux becomes "diffusion
limited" and becomes independent of the membrane permeability. At
such a condition there is a declining flow return as the pressure is
increased indicated in the lower plot of Figure 35.
Initially the flux at 2.76 MN/m2 (400 psi), 71 C, 5.1 m/sec was
1.89 x 10~5 m/s (40 GFD) and the unit was set up to achieve a particular
recovery. It is not known to what recovery the data of Figure 35
correspond, but the levels of 90 to 94% are estimated to bracket the
exact value. Water from the permeate was decreed qualified for reuse
by plant personnel.
Normally one expects that (neglecting osmotic pressure for high
molecular weight material) zero pressure should produce zero flow.
However, the intercept in Figures 35 and 36 both extrapolate to zero
flow with some 0.41'MN/m2 (60 psi) applied. Since the pressure drop
in the module was 0.54 MN/m2, the average membrane pressure was less
than the applied pressure at the inlet which is used as the independent
variable. Tn light of the pressure drop, a 0~41 MN/m2 offset is
reasonable.
Water from the desize washer was used as feed for approximately
160 hours without a significant permeate flow reduciton. The flow
had pH % 11 to 12 and was not adjusted even though previous opinions
favored pH control. Figure 37 shows the progression of both concen-
trate (feed to membrane) solution and the rejection data. A
significant variation in raw feed to the unit was observed in the
range from 1% to nearly 2% total solids for an average of 1.38%
[13,800 mg/A). The feed data are shown darkened in Figure 37. The
rejection of conductivity was surprisingly high at the high pH level
but did not result in elevated electrolyte as the rejection dropped
quickly as the concentration increased. The rejection of total solids
was high allowing concentration levels over 9% solids. At the
9% level the fluid viscosity was high enough that the feed to the
pump could not be maintained fully and further concentration was not
attempted.
Figure 38 shows the dependence of membrane flux on the fluid concen-
tration. At the conditions of 4.13 MM/m2 (600 psi), 71 C (160°F), and
5.2 m/s (17 ft/sec) velocity, the observed flux dropped from about
3.3 x 10~5 m/s (70 GFD) to about 0.5 x 10"5 m/s (10 GFD) at 9% solids
corresponding to 88% volume recovery. Most of the data scatter appear
to evolve from the compensation procedure used to adjust for the
temperature effect. Points #5 and #7 were corrected more than 40 C
and the temperature trends observed were not uniform. Despite the
uncertainty in data interpretation the early and late data, separated
by a hundred hours, do not indicate any membrane flux decline.
The effect of pressure and velocity on membrane flux was investi-
gated briefly at one concentration. The results are shown in Figure
39. Data obtained at two velocities are shown together with the
trend of the membrane as expected on water and on the desize washer
96
-------
X
E
0
J
to
i_
i\J\J
80
1
£0
gi= 60
°H
Q; UJ
uj 12
°-S 40
20
0
~ 100000
^
^^
^ 80000
E
"" 60000
H
p 40000
^
0 ° o % o
° T77
r v V
V
v v
pH=||-l2
V
_i I i I i i i
0 20 40 60 80 100 120 140
0 OTOTAL SOLIDS
VCONDUCTIVITY
• TOTAL SOLIDS OF
FEED MATERIAL
0 o
0
f\
o
§ S § 2000(
o
20
W
I
i
±
±
40 60 80 100 120 140
OPERATING TIME HOURS
W AVERAGE FEED
±
FIGURE 37 FIELD RESULTS - HYPERFILTRATION OF CHEMICAL DESIZE WASHER FLUID
-------
3.5
3.0
2£
PERMEATE2'0
FLUX
m^/ni2 doy
1.5
1.0
0.5
0
_
- «DATA CORRECTED FOR
9* VARIATIONS IN TEMP. AND
PRESSURE TO THAT EXPECTED
AT 71 C (160 F),
4.13 MN/m2(600psi)
4 'NUMBERS INDICATE SEQUENCE
.. 3.
1 • 6
K> ^
8
\
\
N
\
N
N
}
\
.
7
AVERAGE FEED
•
2
i i i i i i i i i i
34 5678
SOLIDS CONCENTRATION %
10
FIGURE 38 CHEMICAL DESIZE FLUID - FLUX VERSUS SOLIDS
98
-------
FLUX
2.0
day
1.0
0
ZO- PA A MEMBRANE
FLUX CORRECTED TO 71C (160 F)
5.3 PERCENT SOL IDS
_L
_L
J_
J_
1234
PRESSURE, MN/m2
FIGURE 39 CHEMICAL DESIZE FLUID - FLUX VERSUS PRESSURE AND VELOCITY
99
-------
at zero recovery. Clearly at the zero recovery a slight diffusion
resistance is encountered and at the 5.3% solids level the resistance
is significant. The indicated velocity effect is greater than
anticipated. Usually flux varies with the 0.8 power of the velocity
to a useful approximation. Although the viscosity of the fluid has
not been determined, it is felt that the flow may be near or in laminar—
to-turbulent transition. The data were run much cooler than 71 C so
that the -transition would be postponed to lower velocities at higher
temperature.
Operation of the hyperfiltration unit was moved to the caustic
washer. Even though the caustic water was not exposed to the membrane
during the screening tests, the deisze washer experience was encour-
aging. Also the peroxide washer which was used in the screening tests
does not have a drain but is coturterflowed to the caustic section
and thus was unavailable for processing. The operation was conducted
for 100 hours without a significant change in permeate flux, although
a slight increase was registered. At the end of the exposure period,
some data on the effect of pressure, temperature, and velocity were
determined. The data (Figure 40) show at 3.1 MN/m2 that drops in
flux occur as the velocity is changed from 5.1 to 3 ra/s. The best
interpretation is that the flux declines from 3.5 x 10"5 to 2.35 x
10~5 m/s (75 to 50 GFD) with the velocity variation. The single low
value with a velocity of 5.1 m/s may be due to non-equilibrium in the
relatively short accomodation period allowed. The flux thus interpreted
varies nearly with v*"^*8 as predicted for diffusion phenomena. Two points
were observed with quite high flux levels. Compared with the trend
of the membame on water, it appears that some observation difficulty
may have been encountered. The lower plot of Figure 40 shows reason-
able agreement between observations on the caustic fluid and that
expected on water.
i
The data of Figure 40 are interpreted to indicate that at 87%
volume recovery and below (lower concentration) the flux of permeate
is essentially as it would be on water. This trend only holds for
relatively high velocity operation; at lower velocities, and probably
only at the higher concentrations, the flux declines about as v 0-8^
The data show nearly linear flux vs. pressure but in the velocity
dependent region a declining return is predicted. It seems to be
reasonable to expect an average flux of at least 2.35 x 10~5 m/s (50
GFD) and more probably 3.5 x 10~5 m/s (75 GFD) to be possible for a
90% recovery system.
Table 23 shows the results of analysis of spot samples from
operation of hyperfiltration membranes on the caustic washer. The
rejection of COD and total solids is high as expected. The presence
of oil and grease and suspended material in the permeate may be due
to intrusions into the sample container. Also the pH values noted
are below 13 and are not considered accurate perhaps due to delay in
measurement.
100
-------
pH~!3, T = 76C
87% VOLUME RECOVERY
ZO-PAA MEMBRANE (100 HOUR EXPOSURE)
8
6
J
FLUX
m3/m2doy 4
2
0
(
1.5
1.0
0.5
'V 0
-0.5
-1
o 5.1 mps
a 3 mps o
/^ — MEMBRANE ON WATER
- S***
/ .xSOOpsi (ref)
.xi i iti i i
) 1 23456
PRESSURE MN/m2
—
- --^,^ CHARACTERISTIC
^*""3-S-^*t/ TREND ON WATER
°^"^-cX AND DILUTE FLUIDS
-
A 1 1 1 1
.0028 .0029 .0030 .0031
RECIPROCAL TEMP '
FIGURE 40 CAUSTIC WASHER FLUID - PARAMETRIC FLUX DEPENDENCE
101
-------
TABLE 23. INDUSTRIAL PLANT ANALYSIS OF SPOT SAMPLES
(caustic washer)
Parameter
Sample/Date
PH
Total Solids (mg/£)
Suspended Solids (mg/Jl)
Volatile Solids (mg/Jl)
Oil and Grease (mg/Jl)
COD (mg/Jl)
Permeate
8/19
12.6
7300
100
5200
0
5560
Concentrate
8/19
12.7
21,900
3400
7500
0
31,200
Permeate
9/14
12.5
6400
0
4600
0
2124
Concentrate
9/14
12.6
2100
200
7500
0
21,744
Permeate
9/16
11.7
13,400
0
6800
100
4575
Concentrate
9/16
11.3
92,900
6500
19,800
0
172,800
o
10
-------
6.4 Ultrafiltration of Dye Range Fluids
The previously described ultrafiltration test unit was supplied
with filtered feed from washer #1 of the dye range. The temperature
was not controlled and tended to respond to process temperature
depending on the rate of throughput. Figure 41 shows the results
of operation. There was no noted effect of concentration in the
range of 0 to 96% recovery which were run. The membranes were
operated in the normal flow and pressure ranges except for the SS
(stainless steel ZrO) dynamic membrane which normally operates at
higher pressures and has lower permeability. The Abcor SWM ultra-
filter was operated at pH = 11 outside its intended pH range of up
to 10.5 which may account for the drop in permeation rate noted.
The SS module also experienced a flux decline while the Union Carbide
module increased by a significant amount, not entirely due to
temperature changes.
Table 24 shows the results of sample analysis for the rejection
of the ultrafiltration membranes on the dye washer fluid. Except
for the initial conditions, the dynamic membrane rejected in the
highest amount. Materials present in the stream are sodium hydroxide,
sodium hydrosulfite, and reduced dye. Figure 42 shows the relative
light transmission recorded with each membrane including the dynamic
hyperfiltration membrane. All membranes rejected the colored material
well. The dye laboratory tests indicated sufficient color removal
for reuse of each permeate.
Date
6/7
6/8
6/10
6/13
6/24
TABLE 24.
Rec.
50%
TOTAL SOLIDS
ABCOR-SWM
19
45
16
30
26
REJECTION ON DYE WASHER
ZrO Dynamic
27
55
35
43
34
UC3NJR
28
33
16
21
21
The membranes were next exposed to soaper washer effluent. Figure
43 shows the conditions and permeate results which were observed. NO
analysis was made supporting the activity except the reuse evaluation.
Reuse of the water was satisfatory and there was no concentration of
surfactant observed. The Union Carbide membrane flux diminished
somewhat compared to operation on the dye washer. Both other membranes
registered increases in flux - the dynamic membrane because of the
temperature increase, and the Abcor membrane presumably because of
relief from the previous fluid conditions. The soaper washer produces
apparently ideal conditions for all membranes applied to it. The
mild acetic acid and oxidant together with extremely dilute trash
content all presumed to contribute to this situation.
103
-------
o
*»
48
40
4
2
6
x
I
a
0.092-.099 MM/HI2 (79-80p»0
D.03I-.035 MN/ma (45~90psi)
FLJOW RATE l-l.im8/!*
o° o ^
oo o oo oo
o
o
o a
tr
d
a
a
a
o
- vv v
0 CARBIDE
O ABCOR
V SS
O
V
o
w
1 1
0
V V
40
80
120
160
200
240
280
OPERATING TIME. HOURS
FIGURE 41 FIELD RESULTS .- ULTRAFILTRATION ON DYE WASHER FLUID
-------
PERMEATE
Concentrate ^ ABCOR UNION CARBIDE ZrO Dyn ZrOPAA Dyn
4IOnm
100
500 nm
100
600 nm
FIGURE 42 RELATIVE LIGKT TRANSMISSION - MEMBRANES ON DYE WASHER
105
-------
PRESSURE
INLET, .49-.55MN/m2
OUTLET,.28-33MN/rri2
CARBIDE
O ABCOR
SS
20 40 6O 80 100 120 140
OPERATING TIME, HOURS
FIGURE 43 FIELD RESULTS - ULTRAFILTRATION ON SOAPER WASHER DYE RANGE FLUID
106
-------
The ultrafiltration skid followed the hyperfiltration skid to
the preparation range where membranes were applied to fluids in the
first two washers. In contrast to the operation of the hyperfiltration
membrane on the alternator (water desize) washer, only slight velocity
dependence to the flux was observed. Part of the reason may be that,
at the low operating temperature, the flow may be laminar due to
the increase in viscosity with decreasing temperature. In laminar
flow the effect of velocity on diffusion resistance is much less
pronounced than in turbulent flow. Because of this low temperature
condition the data of Figure 44 may be significant underestimate of
the permeate rates which would occur at 80 C (expected observation).
The data obtained for the hyperfiltration membrane shown in an
earlier section should be essentially identical to that obtained for
ultrafiltration at similar conditions. However, the flux (hot fluid)
for hyperfiltration was higher.
Rejection of total solids (average of two determinations) for
the three membranes operated on water washer fluids was: Abcor -
48%, Union Carbide - 57%, Zro dynamic - 71%. Despite the relatively
low values all permeate water was found to be satisfactory for reuse
by laboratory tests. Laboratory analysis performed by the cooperating
industrial plant of the relative contents of the permeate samples
are shown in Table 25.
TABLE 25. PREPARATION RANGE WATER WASHER:
COMPARISON OF ULTRAFILTRATION PERMEATES
Item* ABCOR Permeate
COD
Suspended Solids
Total Solids
Volatile Solids
Oils and grease
PH
2446
0
3100
1200
0
6.0
ZrO Dynamic Permeate
1960
100
1700
800
0
7.7
UC3NJR Permeate
1599
200
2400
800
0
6.4
* Units in ppm except pH which is unitless.
Samples 8/8/1977.
The skid operation was moved to the chemical desize washer. Only
the Union Carbide membrane was used because of the high pH in excess
of the recommended level of Abcor operation and the low total flow
rates produced by the SS dynamic membrane. The data are shown in
Figure 45. After a brief period the flux dropped to the line shown
in the figure where it held constant for about 100 hours. Two complete
concentration runs were performed and the operation was performed
while attempting to maintain an elevated temperature. The data taken
are sketchy, but are presented' in the interest of completeness. As
in the water desize washer, the ultrafiltration results for flux are
less than those measured for hyperfiltration. The region of concentration
107
-------
1.4
MEMBRANE
FLUX
I.O
0.8
0.6
0.4
T = 3IC
PRESSURE IN = 052 MN/m2 (75 psi)
FLOW = 3 to 8GPM (NO SIGNIFICANT
DEPENDENCE)
3-0.63 em ID
(UNION CARBIDE
3NJR TUBES)
ABCOR SWM MODULE
—£L
1.3cm ID Zro Tuba
20 30 40 50
RECOVERY (VOLUME)
FIGURE ¥» FIELD RESULTS - ULTRAFILTRATION OF WATER DESIZE FLUID
108
-------
UC 3NJR MEMBRANE
10dm3/min flow (2.7 6PM)
0.7 MN/m pressure (100 psi)
50-70 C TEMP
4.- x- INITIAL OPERATION
3.-
PERMEATE
FlUX
mvm-day
I.
0
0 10 20 30 40 50 60 70 80
PERCENT RECOVERY
FIGURE 45 FIELD RESULTS - ULTRAFILTRATION OF CHEMICAL DESIZE FLUID
109
-------
below about 35% recovery is indicated to be membrane permeability
limited. At higher recovery levels, the concentrated material
adjacent to the membrane increases the resistance to flow and the
steadily decreasing flux.
110
-------
7.0 ECONOMIC ESTIMATES
The following section describes an effort to forecast the general
level of costs associated with the application of pressure driven
membranes. There are acknowledged weaknesses in the information and
its applicability will not be uniform to all sites in the industry.
There is a fundamental procedure which has been followed in achieving
the estimates made herein. First, the field data have been used to
arrive at a design of a membrane unit which will achieve a stated
recovery goal. The designs account for pressure drop and where possible
arrange modules to form a tapered system. In the tapered system the
velocity is maintained within a certain range by dividing the flow into
an appropriate number of parallel elements. The designs of hyperfiltration
systems always optimize at the lowest velocity comensurate with high
rejection. Usually the velocity is decreed rather than designed. In
ultrafiltration (high solids cases) applications there is frequently a
balance between cost of membrane and cost of pumping. At low velocity
the pumping cost is diminished but the flux decreases causing an
increased membrane size. In the following designs the optimum has
been sought for such cases and the velocity or flow is stated for
situations where no optimum occurs.
7.1 Individual Designs
7.1.1 Water Washer on Preparation Range 151 dm?/min (40 GPM, 75%
recovery)
Abcor ultrafiltration modules have been arranged as (number in
parallel x number in each leg) accordingly to the following schedule of
segments: (8 x 25)+(5 x 20)+(3 x 37). The flux data at 0.48 MN/m2
(70 psi) have been adjusted in proportion to pressure. Three repres-
surizations have been required. It is believed that a significant
design improvement would be realized by allowing higher pressure to be
utilized. The membrane area could be reduced considerably and the
number of pumps decreased. The flux schedule shown below has been used
with an assumed value of 3.44 m2 (37 ft2) of area per module.
% recovery flux m3/m2-day (GFD)
0 0.204 (5)
30 0.163 (4)
50 " 0.122 (3)
75 0.081 . (2)
111
-------
Pressure drops proportional to the square of flow have been used with
AP = 7 kN/m2 (1 psi) at 0.0011 m3/min (3 GPM) flow rate. Flow rates are
maintained between 0.0011 and 0.016 m3/min for the design stated.
Power consumption of 45 kw is estimated.
Union Carbide ultrafiltration modules in 46.5 m2 (500 ft2) elements
have been employed. These packages have been operated with an assumed
recirculation flow of 0.0038 m3/min (1 GPM) to each of the 1910 elements.
The following fluxes per module have been used in agreement with the
data taken.
Module Flux m3/m2 day (GFD)
1 1.22 30
2 0.896 22
3 0.652 16
4 0.469 11.5
5 0.367 9
The power consumption of each module is estimated at 28 kw, for a total of
140 kw.
ZrO membrane performance was measured at low pressure of 0.48 MN/m
(70 psi) which has been assumed to be scaled proportional to a pressure
of 4.8 MN/m2 to yield a flux of 0.406 m3/m2-day (10 GFD). A single
pass system would require at least 4015 m2 (43,200 ft2) of area and
will require a minimum of 40 kw of pump power.
The ZrO membrane data suggests a region from 0 to 50% recovery of
constant permeability at 3*3 m/s greater velocity; J = 1.8 m /m -day
(44 GFD) at 1.66 MN/m2 (240 psi). From 50 to 75% recovery the data
suggests J = J (v/v )**0.8 (double asterisk is Fortran symbology for
raised to power 0.8). Here Jo = 1.63 m3/m2-day (40 GFD) at VQ = 3.3 m/sec
(11 fps). A pressure drop of 9100 N/m2 per m of length at 3.3 m/sec
velocity has been used. The design resulting is expressed as (number
of parallel elements x length of element) in the groupings (5 x 36.6 m) +
(3 x 49 m) + (2 x 527 m) + (1 x 152 m). The first two groups reduce the
flow to the 50% recovery work and the last two provide the final concen-
tration. The overall results is 65 m2 (700 ft2) and membrane area and 40
kw pump power required.
7.1.2 Chemical Desize Washer (0.151 m3/min (400 GPM) 80% recovery)
The Union Carbide data taken are not conclusive but based on the
best interpretation the following design is made (6 m/s velocity).
Module Flux m3/m2 day (GFD) % Recovery
1 1.83 (45) 45
2 1.02 (25) 62
3 0.49 (12) 70
112
-------
Here 3 modules can be used to achieve 70% recovery of feed water as
permeate.
The ZrO-PAA membrane data projected to 82 C and 3.6 m/s velocity are
as follows.
% Recovery Flux (average) m3/m2 day (GFD)
0 to 40 2.04 , 50
40 to 60 1-69 41.5
60 to 70 1.43 35
70 to 75 1.25 30.6
75 to 80 1.00 24.6
Using these flux data permits estimation of 79 m2 (750 ft2} of membrane
area and three pressurizations using a power of 50 kW. It would be
attractive to pursue a lower velocity design to allow less pumping.
7.1.3 Caustic Washer (0.151 m3/min (4O GPM) @ 90% recovery)
Only the ZrO-PAA membrane was operated on this feed and at the
elevated pH the results are not expected to be due to the second layer
(PAA). The data suggest little effect of velocity and that the flux J
is 9.7 x 10~7m/N-day times the pressure (e.g. 6.72 m3/m2-day at
6.94 MN/m2). A system having a velocity about 3 m/s velocity requires
two pumps and has area of 36.5 m2 (393 ft2). A lower velocity
could possible be adopted to eliminate one of the pumps but no data
were taken in the 2 m/s range to justify such a design. A power
requirement of 47 kW is estimated.
7.1.4 Dye Washer #1 (0.163 m3/min (43.2 GPM) @ 90% recovery)
The Abcor unit showed 1.01 m3/m2-day (25 GFD) flux production, and
pressure drop is not high in this case, so a simple estimate of 60
modules results. Power of 7.5 kW is sufficient for pumping.
The Carbide membrane produced 4 m3/m2-day (100 GFD) on this feed
such that one module will essentially recover 90%. Power of 28 kW is
required.
7.1.5 Soaper Washer on Dye Range (0.163 m3/min (43.2 GPM) @ 90%
recovery)
The Abcor module produced 3.3 m3/m2-day (81 GFD) at expected
operating conditions. An assembly of 19 modules requiring 3.5 kW
power is judged adequate.
. 113
-------
The Carbide unit also produced 3.3 m3/m2-day flux but is expected
to produce double this at operating pressures of the design. A single
500 ft2 module is only operating to 70% of its capacity. Power of 28
kW is required.
The ZrO membrane flux adjusted for pressure effects in operating
compared to testing is 8 m3/m2 day so that 26 m2 (280 ft2) of membrane
is estimated. Power of 43 kW is also estimated.
The ZrO-PAA membrane produced a flux of 4.2 m3/m2-day (103 GFD).
A membrane having 51 m2 (595 ft2) is anticipated. Power required is 43
kW. A higher recovery fraction could be achieved with this membrane
with the possibility of retrieving a portion of the $65 per day soap cost.
7.1.6 Atmospheric Dye Becks (0.319 m3/min (83 GPM) 95% recovery)
The flux data on the atmospheric dye becks showed about 200 hours
of operation at high flux followed by a severe flux decline to 20% of
the initial value. The drastic nature of the decline leads to the
probability that there is an element of feed which causes the decline
which may be avoided. Therefore, two designs are advanced, one using
the low flux and the other using the relatively vulnerable high flux.
The high flux case is only meaningful if a problem elimination or
cleaning methodology is worked out.
The low flux is about 0.306 m3/m2-day (8 GFD) while the high flux
is 1.63 (40 GFD). The resulting designs at 2.1 m/sec velocity require
1352 m2 (14,250 ft2) and 265 m2 (2850 ft2) respectively. Operating
power of 91 kW is estimated, though the high flux design will require
less.
7.2 Costs of Systems
The following groundrules have been used in the process of evaluating
expected costs. Abcor has provided a budgetary estimate for the systems
described, including motors, rudimentary controls, and mounted on an
equipment skid. The Union Carbide system cost has been estimated by
the authors at $50,000 per module and it is presumed that this would
include drives and rudimentary controls. The ZrO and ZrO-PAA systems
costs have been estimated at $1290 per m2 ($120 ft2) except for very
large units of $1070 per/m2. This estimate has been based on tube prices
in appropriate quantities ($70 per ft2), the cost of assembly for heat
exchangers ($10 per ft2), the expected cost of membrane application to
assembled modules ($5 per ft2), costs of pump stations with controls of
$10,000 flat costs, and a factor of about 25% for profit and indirect
costs.
114
-------
Annual replacement costs for Abcor units are estimated based on
two year expected life and represent about 12% of initial cost. Replacement
of tubes for the tubular membranes is expected to cost 3% of the purchase
price, though little data exist to support the estimate. Abcor estimates
$500 to $2500 annual cost for cleaning. The tubular membranes likely have
a cleaning requirement which has not been estimated. Operating costs due
to pump power have been calculated with $.02 Jew hour on a 6000 hour year
basis (250 days/yr).
Prefiltration has shown itself to be a major concern with spiral
membranes. An amount from $2500 to $5500 annually has been added to
Abcor's estimate to account for this cost. No similar estimate is
appropriate for the tubular membranes.
Water and energy savings have been estimated on a 6000 hour per
year estimate. Water costs of $0.055 per m3 (25$ per k gal) and
energy costs of $3.00 per MBtu have been used. These figures result
from an informal survey of local industry and are arbitrarily selected
because of the rapid changes. Only energy added above 40.5 C (105°F)
has been claimed as conserved to account for energy recovery in the
plant. This is with the single exception of the beck recovery in
which all energy relative to 18 C (65°F) is claimed as savings.
For all systems the cost of installation has not been added. The
authors feel (with little data to cite) that historic estimates of
installation ratios (60 to 100%) are excessive for the compact equipment
with relatively high costs that membrane systems represent. Many of
the small systems described herein may well be installed for under
20% of the purchase price. The layout of a user's facility, his
relative local power accessibility, and other factors will play a
prominent role in the installation costs. Therefore no installation
costs are included.
No operator costs have been assigned to the systems. All systems
can be made to fail safe with allowance for the user's range or facility
to continue normal operation. Thus no operator is required and
malfunctions may be treated by the vendor or perhaps by the plant's
maintenance department.
In a naive way, the cost of yearly amortization has been estimated
to allow a crude estimate of the net annual cost or savings for a
prototype installation. The present worth factor of 4.35 has been used
to amortize capital costs at a 10% interest rate over six years. That
is, the capital cost divided by 4.36 is taken as a yearly cost of
capital. No effect of tax is evaluated.
7.3 Results of Economic Survey
Following the groundrules mentioned in the preceding section, each
of the designs has been evaluated. Tables 26, 27, and 28 show the results
115
-------
TABLE 26. RESULTS OF ECONOMIC ANALYSIS FOR PREPARATION RANGE
Water Washer - Preparation Range
System
Abcor
UC/GC
ZrO
ZrO-PAA
Capital
Costs
$460,000
250,000
432,000
84,000
Cost of
Yearly
Amort iz at ion
$105,500
57,300
98,600
19,300
Annual
Replacement
Costs
$60,000
7,500
12,900
2,500
Annual
Operating
Costs
$12,900
16,800
4,000
3,480
Annual
Water
a) Savings
$2,700
2,700
2,700
2,700
Annual
Energy
b) Savings
$20,250
20,250
20,250
20,250
Net
Annual
(Cost)
Savings
($155,000)
(58,650)
(92,550)
(2,330)
Chemical Desize Washer - Preparation Range
System
UC/GC
ZrO-PAA
Capital
Costs
$150,000
120,000
Cost of
Yearly
Amortization
$34,400
27,500
Annual
Replacement
Costs
$4,500
3,600
Annual
Operating
Costs
$10,100
6,090
a)
$2
2
Annual
Water
Savings
,538
,900
Annual
Energy
b) Savings
$21,437
24,500
Net
Annual
(Cost)
Savings
($25,025)
(9,790)
Caustic Washer - Preparation Range
Cost of
Capital Yearly
System Costs Amorti z ation
ZrO-PAA $47,000 $10,800
Annual
Replacement
Costs
$1,400
Annual
Operating
Costs
$5,220
Annual
Water
a) Savings
$3,200
Annual
Energy
b) Savings
$27,500
Net
Annual
(Cost)
Savings
$13,280
a) Savings @ 25
-------
TABLE 27. RESULTS OF ECONOMIC ANALYSIS FOR DYE RANGE
Dye Range, 1st Washer
Capital
System Costs
Abcor
CU/GC
ZrO
ZrO-PAA
Dye Range
System
Abcor
CU/GC
ZrO
ZrO-PAA
$ 90,000
50,000
134,400
223,920
Cost of
Yearly
Amortization
$20,600
11,500
30,826
51,360
Annual
Replacement
Costs
$10,500
1,500
4,020
6,720
Annual
Operating
Cost
$4,400
3,357
3,800
3,800
Annual
Water
a) Savings
$3,500
$3,500
$3,500
$3,500
Annual
Energy
b) Savings
$3,000
3,000
3,000
3,000
Net
Annual
(Cost)
Savings
$(29,000)
(9,800)
(32,145)
(55,380)
, Soap Washer
Capital
Costs
$50,000
50,000
33,600
65,400
Cost of
Yearly
Amortization
$11,500
11,500
7,700
15,000
Annual
Replacement
Costs
$10,500
1,500
1,000
1,900
Annual
Operating
Cost
$3,000
3,360
3,800
3,800
Annual
Water
a) Savings
$3,500
3,500
3,500
3,500
Annual
Energy
b) Savings
$22,600
22,600
22,600
22,600
Net
Annual
(Cost)
Savings
$ 1,100
9,700
13,600
5,400
a) Savings @ 25
-------
(Dye Fluid Only)
Capital
System Costs
TABLE 28. RESULTS OF ECONOMIS ANALYSIS FOR ATMOSPHERIC DYE BECKS
29 Dye Beck System
Net
Cost of Annual Annual Annual Annual Annual
Yearly Replacement Operating Water Energy (Cost)
Amortization Costs Costs
ZrO-PAA 1.4xl06
£ Capital
03 System Costs
Zro-PAA $285,000
$321,100
Cost of
Yearly
Amortization
$65,366
$42,000
Annual
Replacement
Costs
$8,550 .
$11,000
Annual
Operating
Costs
$11.000
$7,125
Annual
Water
a) Savings
$7.125
$96,250
Annual
Energy
b) Savings
SQfi.s^n
$(270,7:
Net
Annual
(Cost)
Savings
si s.d^n
a) Savings @ 25$/kgal
b) Basis compared to 65 F cold water
-------
of the evaluations for the preparation range, dye range, and atmospheric
dye becks respectively. For each case the capital cost, yearly
amortization cost, annual replacement costs, annual operating cost,
water savings, energy savings, and net cost or savings are evaluated.
Some of the situations evaluated are winners and others are losers.
The water washer application on the preparation range appears to be
unattractive, but if size recovery is practical this situation would be
faced with a possible credit of possibly $500,000 for reclaimed size.
It is well known that J. P. Stevens Corporation has pioneered in this
activity with Union Carbide technology.
The chemical desize washer carries such a heavy loading that
permeation is very slow causing relatively expensive water recovery.
No chemical recovery is probable and the application has little merit.
The caustic washer appears to be an attractive investment for the
recovery of hot water. In addition, the possibility of caustic recovery
and reduction in waste aklalinity are attractive growth benefits which
will accompany recycles.
The dye range washer water is not hot enough to justify membranes
for water reclamation alone. Only if the reuse of acetate as described
in the process modification section of this report can be demonstrated
and proves valuable will there be an economically based acceptance of
such a unit.
Contrary to the dye washer, the soaper washer on the dye range is
predicted to be an economically viable installation for all membranes
evaluated. All membranes cost a similar amount and have similar benefits
on this particularly "easy" feed stream. The spiral (Abcor) membrane
lifetime and associated replacement costs appear to be the most signifi-
cant cost factor different between the installations. As has been
mentioned previously, laboratory testing showed that the ZrO-PAA
membrane concentrated effective wetting agent toward possible recovery
of this material.
The application to a 29 beck (7 foot size) plant of 83 GPM is
economically attractive only if the optimistic level of the flux
indicated by testing is achieved. The optimistic flux is about
5 times as large as the observed long term value resulting in a
capital installation 20% as large. The optimistic flux will only
be achieved if the substance responsible for the flux decline be
identified and an appropriate remedy be found. In this program
the substance was not identified.
7.4 Economic Impact on Industry
In a study conducted for (ERDA)DOE, General Electric [5] determined
the number of units and energy consumption of classes of textile
119
-------
finishing equipment. Table 29 is extracted from that source, showing
the number of units and respective fraction of energy use of each
type unit. Used as a guide, this information indicates the equipment
studied herein directly represents 48 percent of the wet finishing
processes on an energy basis.
TABLE 29. ENERGY AND
Process
Open-width Preparation
Rope Preparation
Thermosol Dye Ranges
Atmospheric Becks
Pressure Becks
Number
408
389
132
3918
740
NTTMRTCR OP Rp!]-(p.r"|nRn
Annual Energy
1.48.x 10lb J/yr
Io37 x 1016 J/yr
0.26 x 1016 J/yr
1.16 x 1016 J/yr
0.32 x 1016 J/yr
PROCESSES
Consumption
(1.4 x 10l5 Btu/yr)
(1.3 x 1013 Btu/yr)
(0.25 x 1013 Btu/yr)
(1.1 x 1013 Btu/yr)
(0.3 x 1013 Btu/yr)
TOTAL
STUDY TOTAL- ALL FINISHING
4.59 x 1016 J/yr (4.35 x 1013 Btu/yr)
9.63 x 10 16 J/yr (9.1 x 10 13 Btu/yr)
From General Electric Survey, approximately 1972 information (Reference 5)
Table 30 shows the impact if membrane recycle were adopted for
the five processes studied herein. A 21% water reduction and a 27%
energy reduction is expected. The capital required would be something
over % billion dollars, and net operating return of about 60 million
dollars annually would be achieved. The demand for electric power would
increase by about 120 Mw (one eighth of a modern power plant typical
installation). All these factors are considered to be reasonably
positive in nature as regards to the possibility and gross economic
feasibility of membrane installation.
120
-------
TABLE 30. IMPACT ON INDUSTRY OF FULL MEMBRANE IMPLEMENTATION
Process
Open-Width Preparation
Rope Preparation
Dye Range
Becks
TOTAL
Water Reduction (Annual)
54.4 x 10°m?
51.8 x 106m3
18.5 x 106m3
35.2 x 106m3
(14,4 x 10^ Kgal)
(13.7 x 106 Kgal)
x 10G Kgal)
x 106 Kgal)
(4.9
(9.3
Energy Use Reduction (Annual)
1.06 x 10ib J (10x 10iz Btu)
1.02 x 1016 J (9.7 x 1212 Btu)
0.26 x 101G J (2.5 x
0.54 x 1016 J (5.1 x
1012 Btu)
1012 Btu)
1.6 x I08m3 (42.3 x 106 Kgal) 2.88 x 1016 J (27.3 x 1012Btu)
OR
38% of industry total or 30% of industry total
112. x 109 gal* (4.24 x 108m3)
*SIC 22 total, 1972 Census of Manufacturers
-------
8.0 FULL SCALE PROCESS MODIFICATIONS
Two full scale process modifications related to membrane applications
have been evaluated. Both of these have been introduced in the earlier
section on possible process modifications. The first, for atmospheric
dye application, is critical to the effective recovery of energy by
membranes. The second offers a definite economic benefit by itself and
a substantial benefit in reducing the cost of a hypothetical membrane
unit. In most situations involving textile washers there are two
independent parameters-water temperature and flow rate. The flow of
water causes the washed off substances to be carried away from the
cloth by dilution and mixing. Temperature affects solubilities,
diffusion, and viscosity each tending to affect the rate of solute
exchange from the cloth to the water. Conceivably the adequate removal
of material from the cloth may be achieved at many combinations of
temperature and flow. The experiment reported moved from a condition
of relatively high flow, low temperature to a condition of lower flow,
higher temperature. In both conditions, the cloth was adequately prepared
in that the plant's stringent quality measures were satisfied.
8.1 High Temperature Dye Addition Experiment
The purpose of the trial dyeing was to evaluate the influence of
using simulated recycle hot water on a dyeing process conducted at 82 C.
All of the dyeing steps were conducted at 82 C with dyes and salts added
separated, each over a one-hour period to determine the feasibility of
reuse of hot water in the dye cycle. The following procedure was used
for the dye cycle:
(1) Water was introduced into the beck to the desired level.
(2) The cloth was then added to the beck.
(3) Wetting and buffering agents were then added to the beck.
(4) The temperature was then raised to 82 C.
(5) After the temperature of 82 c was reached, the dye
solution was added slowly over a one-hour period. The
beck solution behind the baffle was manually stirred as
the dye was introduced into the beck.
(6) At the completion of the dye addition, the cloth was
allowed to dye for twenty minutes before the salt was added.
(7) Salt (Nad, saturated) was added to the dye bath over a
one-hour period at 82 C with stirring to give a slow and
even application of the salt solutions.
(8) After the addition of the salt was complete, the bath
was cooled to 60 C.
122
-------
Dye Formula 501
Lot 647
41 yards
Pattern 112-0
Color 133
Wt/Yd 1.31
Weight 53.71
Pieces 1
Level 7
Dye/Chemical Name Amount
Amaterge T 1.0750 pounds
Direct Dyes 0.0714 pounds
Fix GD 1.0750 pounds
Salt, Liquid 3.5000 pounds
The foregoing procedure produced a dyeing in a critical shade
which met normal quality standards. This test dyeing followed an
unsuccessful prior attempt in which the dye and chemicals were added
about three times as fast and the stirring behind the baffle was not
done. In the earlier attempt some variations in color were produced.
The demonstration is not absolutely convincing but does indicate the"
feasibility of this critical process.
8-2 Open-width Preparation Rangt;
Following the energy survey of the open-width preparation range
two actions were implemented by the plant. Flow meters were purchased
to document the actual flow. A flow-temperature modification at full
scale under standard processing quality requirements was performed.
Also, water exiting from the peroxide washers was counterflowed to
the caustic washer. It was inserted in the third of the four washers
(counting in cloth direction). The plant was able to maintain the
required high quality after the flow change even with less chemical
additions.
The motivation for manipulating the flow variables on the range
is (1) to conserve water and possibly energy by operating at lower
flow and higher temperature and (2) by reducing the water flow to
enable the most economical hyperfiltration installation. Generally
membranes are more productive at high temperature and a smaller
system will be required for less flow capacity.
The previous and modified conditions on the range are shown in
Table 31. The modified flows are considerably lower than the actual
previous rates and are allowed by increasing the temperature. The
last column reflects the change in energy flow to drain at each
section. Only the water desize wash requires more energy at the new
condition while the others require much less. Because the cloth exposed
to atmosphere is hotter, a greate evaporation rate calculated by the
123
-------
same procedure as for Table 2 is anticipated as indicated by Table 31.
However, the net change favors operation at higher temperature. Also
the water borne energy was previously 76% of the total while after
the change only 52% of the energy is exhausted with the hot water.
The reason for this is that the energy to drain has been diminished
and the other energy requirements have been increased.
TABLE 31. FLOW AND TEMPERATURE MODIFICATIONS
Previous Flow @ T Modified flow @ T Change in E
(dm3/s @ °C) (dm3/s @ °C) GJ/hr
Water Desize Wash
Chemical Desize Wash
Caustic Wash
Bleach Wash
4.41 @ 60
3.15 @ 82
4.41 @ 82
3.15 @ 76
2.59 @ 93
1.39 @ 88
1.58 @ 91
1.58 @ 93
+0.202
-1.571
-2.51
-0.97
Estimated increase in vapor loss = + 3.32 GJ/hr
estimated net energy change = - 1.53 GJ/hr
Previous total use 16.9 GJ/hr (reference)
Energy reduction 9%
Flow reduction 47%
124
-------
APPENDIX A
FUNDAMENTAL EVALUATION OF EVAPORATIVE AND CONVECTIVE HEAT LOSS
IN FULL-SCALE TEXTILE EQUIPMENT
The magnitude of heat loss by convection and evaporative
mechanisms in typical textile equipment is not documented in the
literature so it was decided to perform some basic field experi-
ments to determine the magnitude of such losses. Because of the
nature of the experiments and the instruments used it is difficult
to assert any high accuracy estimate; however, the data are felt
to be useful to estimate the amount of energy loss by the combined
mechanisms. One similar experiment was run during the course of
our work, the data from which are supplied herein for comparison.
The first experiment was conducted on dye becks from 3 to 12
feet in width shown in schematic cross-section in Figure Al. During
normal dyeing procedures the beck temperature decline on cooling
was recorded. The simple equation which follows was used to
calculate heat loss rate.
Q = -me dT/dt
Q is the heat loss rate by all mechanisms
m is the mass of water, cloth, and affected beck portions
c is the specific heat of mass m
dT/dt is the temperature derivative with respect to time or the change
in temperature with time.
Generally, the losses external to the beck are very small, much smaller
than the error inherent in the procedure, and were summarily neglected.
Procedurally, the time derivative is difficult to calculate
accurately. We used three methods to estimate its effect, each of
which has weaknesses.
Method A) Fit an exponential curve to the data.
Method B) Fair a curve through the plotted data and extract
tangents at various points.
Method C) Use straight line segments between successive data.
Having tried all the above and comparing the results we have
preceded to interpret the data as described in the following.
The data were examined for effect of beck width and no systematic
effect was found. Therefore all data were normalized by the beck width.
It is considered intuitively satisfying that width should offer no
variation, since both the cloth surface and water surface do not vary
125
-------
DOOR
COLD
WATER
SUPPLY
CLOTH
BECK VENT
THROUGH ROOF
DAMPER
ROTATING
FRAME
DOOR
OVERFLOW
WATER LEVEL
DRAIN
DRAIN PORT
FIGURE Al SCHEMATIC DIAGRAM OF AN ATMOSPHERIC BECK
12*
-------
with beck width normally. Theoretically the evaporation rate will
vary in proportion to the vapor pressure at the liquid surface
temperature since the vapor pressure in the incoming air is small.
Upon dividing the heat loss rate per width of beck by the vapor
pressure at the observed beck temperature, the data are as presented
in Figure A2. There is a tendency for the magnitude of this "mass
transfer coefficient" to decrease somewhat as the temperature
increases. The decrease may be due to any one of several factors
and considering the experimental uncertainty may not actually exist.
The data from Reference 6 are seen to agree with those observed.
The above data were observed under the conditions with the
vent damper open and the door to the beck closed. Opening of
the door (with damper open) causes an estimated 20% increase in
less rate. The closing of the damper may result in a 50% decrease
in loss rate below that presented. Our data with damper closed
are much fewer than with damper open and the exact decrease in
values so crudely determined is difficult to assert with confidence.
That there is a change is undoubted, however. Figure A3 shows the
observed variation of temperature in time for three runs. In
one, the damper was open and the door closed from the highest
temperature. In another the damper was open and the door open
and the door closed at about 86 C, at which time a noticeable
decrease in temperature decay occurred. Finally a third run with
door and damper closed was changed to both open and a more rapid
drop ensued. These data show conclusively that the positions
have a strong effect.
The second experiment was conducted in a wash box on a
continuous range depicted schematically in Figure A4. Cloth carry-
ing water from a similar box at 89 C enters the box and flows in
a serpentine fashion through the wash water. Water is added to the
box and direct steam injection provides heat to control the tempera-
ture. Two experiments were run. In one, the water supply temperature
was adjusted to exactly equal the outflowing water temperature and
the steam was shut off. Equilibrium was reached at 63 C at which
condition the difference in energy content of the cloth and water
at 89 C entering and 63 C existing supplies the losses from the cloth.
This energy loss rate is assumed to correspond with the 63 C level
although a considerable portion is lost from the cloth in transit
between the upper and subject box. In the second experiment the water
supply was shut off and temperature raised by steam addition. When
the steam was terminated, a record of temperature decline was made.
The heat loss is then calculated from
Q = -Me dT/dt + (n^c + m c ) (89 - T)
6Woodall, L. C. and E. F. Godshall, "Energy Economics in a
Dyehouse," presented at Clemson University, Energy Conservation
in Textile Industry, January 12-13, 1977.
127
-------
2 00.914 Q = HEAT LOSS BY EVAPORATION
E A 1.52 KJ/hr
A 2.13
V 2.54 6u= BECK WIDTH -- m
I-
£12
\
Z
^ 8
j
i ^1
b
J
J 4
o
A)
3 2
•V
^ °
T 2.54REF(6)
0 3.66 Psat=SATURATlON PRESSURE
Pa=N/m2
0
a
a
w
A A (!* A A
^^VotfcS
° °V0^ ^^A A A
— ' 1 1
>U 80 ~90
100
BECK TEMPERATURE ~C
FIGURE A2 CORRELATION OF EVAPORATIVE LOSS DATA
FOR ATMOSPHERIC BECKS
128
-------
100 h
o OPEN DAMPER, CLOSED DOOR
D OPEN DAMPER, OPEN DOOR
DOOR CLOSED AT •
CLOSED DAMPER .CLOSED DOOR
BOTH OPENEDAT A
20 30 40
TIME , MINUTES
50
FIGURE A3 TEMPERATURE TRANSIENTS SHOWING EFFECTS
ON PORTAL CONDITIONS
129
-------
WATER IN
AT
FROM 89 C
WATER
OUT
DIRECT STEAM
INJECTION
FIGURE A4 SCHEMATIC OF WASHER USED FOR HEAT LOSS STUDY
1-30
-------
Q is the heat loss at temperature T
Me the sum of mass and specific heat products for the water,
contained cloth, etc.
m, c, is the mass flow rate and specific heat of the cloth
m2C2 is the mass flow rate and specific heat of the water carried
on the cloth
dT/dt is the time rate of change of temperature T
89 is the upstream source temperature.
The results of these experiments are shown in Figure A5. The loss
rate very nearly coincides with the vapor saturation pressure curve
as indicated by the correlating equation fit to the data obtained.
For estimation purposes, the area of liquid and cloth (one
side) have been used to normalize both the data for the wash box
and the beck. This calculation results in a single equation
Q = 2 x 10~4 A Psat
Q is the loss rate in kJ/s (kW)
A is the area of liquid and one side of the cloth involved in m2
Pgat is the vapor pressure at a given temperature in N/m2 (Pa).
The equation agrees with both the beck and wash box experiences and
has been used to estimate losses occurring from wet cloth surfaces
in rope form, approximate division of losses within components, and
the impact of modifications. Use of the equation in situations
having significant inhibitions of air circulation is considered
imprudent.
131
-------
UJ
<
Of
CO
§
UJ
ISO
160
140
120
100
80
60
4O
ZO
0
CORRELATING EQUATIONS'
0= 0.0021
SAT
TRANSfENT
TEST OA1A
STEADY STOTE TEST
20 40 60 8O
TEMPERATURE. °C
IOO
FIGURE A5 VARIATION IN OBSERVED HEAT LOSS WITH TEMPERATURE
132
-------
APPENDIX B
CALCULATION OF EXTERNAL RADIATIVE AND CONVECTIVE HEAT LOSSES
IN TEXTILE EQUIPMENT
Heat loss by convection and radiation external to equipment
has been calculated using models and procedures in common use.
Radiation from an object at Tj^ to its surroundings at T2 obeys a
relation of the form
q = aF^I^4 - T^)
q is the heat flow per unit area
a is the Stefan Botlzmann constant 5.67 x 10~8 (Wm"2^1*)
Fe is a factor used to account for the radiation emission properties
of the involved surfaces
Fa is a view factor (fraction of total view of 1 occupied by 2) .
For simplicity F and F are assumed to be unity. The assumption
may be poor for Fe if the material is a bare metal in which case
the calculation will over-estimate the actual loss. Under such an
assumption the above equation may be factored and arranged as
= hr (T! - T2>
2 2
hr = a(Tx + 1^) (T + T )
For values of T and T the range 300-3 50K, h is well approximated
at 5.7 W/m2 C Tl Btu/hr-ft2-F) and this value is adopted.
In free convection in air from vertical plates, horizontal
cylinders, or horizontal plates1 gives equations from correlations
of data as
h = (0.95 + 1.43)(Tx - T2)1/3 W/m2-K
In the range up to T - T2 = 50 C, values of h are approximately in
the range of 5 W/m2-K. Values for forced convection are by comparison
expected to be small except perhaps in cases of high localized heat
transfers accompanying high inpinging air velocities.
7J. L. Holman, Heat Transfer, McGraw-Hill, 1976 (4th Edition).
133
-------
Our calculations are based in the assumption of a total radiation
plus convection loss described by
q - h(Tx - T2)
h = 11.34 W/m2-K (2 Btu/hr-ft2-F)
Inasmuch as the losses calculated are expected to be above the actual
amount and the calculated losses are generally quite small in
comparison with evaporative losses, this procedure is deemed sufficient.
134
-------
APPENDIX C
ENERGY USE IN ATMOSPHERIC BECK PROCESSES
DESCRIPTION OF FACILITY AND DYEING PROCESS
Dye House Facility
The facility observed employed five different sizes of atmospheric
becks ranging in capacity from approximately 1000 to 2200 gallons in
the dyehouse. At the same that this work was begun, there were 32
operating dye becks in the dyehouse, 10% of which would be undergoing
maintenance at any one time. The becks are essentially manually
operated, though temperature recorder-controller units regulate the
rate of temperature rise during the heating-up events and maintain
the temperature at a nearly constant value during elevated temperature
operation. The becks are installed so that the access doors are at
a convenient height with respect to the operation floor level; that
is, the becks are in a pit which dips to about four feet below the
operating floor level (see Figure Cl). In order to avoid severe burns
to the operating personnel, the becks were vented, and fans circulate
air through the becks and out through the building roof. The design
is such that air circulates through a given beck only when the beck
door is open.
Dyeing Process
Dyeing in an atmospheric beck is done as a batch process with
both natural and synthetic fibers being dyed in lots from 50 yards
('v-GO Ibs) to 1600 yards (^2000 Ibs). The becks are supplied with steam,
water, and chemicals at the proper times during the dye cycle to
accomplish the scouring, dyeing, and dye fixing for the cloth. A
large amount of cloth is loaded into a beck so that almost all of the
cloth is immersed in the water. Several endless loops of cloth are
used, with a small amount of cloth in each loop passing over the
rotating frame (see Figure Cl). This rotating frame causes the loops
of cloth to be continuously circulated through the chemical solutions
during a dye cycle. Circulating the cloth in this manner serves
two purposes: First, it causes the cloth to be uniformly exposed
to the process chemicals; and second, it keeps the solution stirred
for more even distribution of chemicals and temperature within the
bath.
135
-------
DOOR
COLD
WATER
SUPPLY
CLOTH
nfn
I
I
BECK VENT
THROUGH ROOF
DAMPER
ROTATING
FRAME
DOOR
OVERFLOW
WATER LEVEL
TO DRAIN
DRAIN PORT
FIGURE Cl SCHEMATIC DIAGRAM OF AN ATMOSPHERIC BECK
136
-------
The normal dye cycle has three distinct phases: first, the
cloth is scoured (cleaned and bleached); second, the dyeing is
accomplished by exposing the cloth to a high-temperature dye bath
for a relatively long period of time; and third, the cloth is
exposed to chemicals which "fix" the dye so that the colors are
fade resistant and so that the dye does not migrate in the wet cloth
during the remaining processing within the plant. Correspondingly,
there are three times when hot liquid is discharged from the dye beck
into the plant waste stream. These discharges are accomplished by
adding cold water to the beck (see Figure Cl) so that the liquid
rises and overflows into the sewer. The cold water rinse both cools
the beck contents and flushes the processing chemicals out of the cloth.
Normally, a beck is completely drained only after flushing for a long
enough time for the water to become clear. Two typical cooling rates
are shown in Figure C2. For the first 17 minutes (approximately) a
low cooling water flow rate assured that the cloth cooled slowly
enough to avoid wrinkles; the cooling water flow rate was then
increased to finish the cooling and washing process. From the
curves of Figure C2 cooling water flow rates of 34 gal/min and
225 gal/min were calculated for the two stages of this process.
For those fabrics which do not wrinkle easily when cooled rapidly,
the cooling water is admitted to the beck for the whole process at
the higher rate of 225 gal/min. In either case, the total amount
of water required is the same if the same final temperature is
reached. In the cases illustrated in Figure C2 the becks were
drained after their contents were cooled to about 100°F.
Plots of temperature versus time for typical complete dye
cycles are shown in Figures C3 and C4. Figure C3 is for a cycle
which includes the scour, dye and dye fix phases, while Figure C4
is for a cycle having only the dye and dye fix phases. The scour
phase was used only for automotive fabrics, which constitute about
one-third of the batches of cloth dyed. The various events of the
cycle are indicated in Figures C3 and C4, and are listed here:
(1) Cold water (at the average daily atmospheric temperature
is admitted to the dye beck and is heated to between 90°F
and 120°F.
(2) The cloth is loaded into the beck.
(3) Scouring (cleaning) chemicals are added to the water, and,
while the cloth is circulating through this solution, the
temperature is increased to, and maintained at, 160°F for
about one-half hour.
(4) The scouring chemicals are flushed out of the beck and
rinsed from the cloth. (This requires a large amount of
"cold" water.)
(5) Fresh water again fills the beck and is heated to between
90°F and 120°F.
(6) The dye chemicals are added to the water.
(7) The solution is heated and maintained at about 190°F for
sufficient time to color the cloth to the desired shade.
137
-------
ISO
160
BECK TEMP DURING
COOLDOWN
A BECK 8
O BECK 9
BECK CAPACITIES-GAL.
NORMAL OVERFLOW
BECK 8 1428 2180
BECK 9 1632 2490
u.
o
140
120
100
OVERFLOW BEGINS
0
2
8
10 12 14
TIME-WIN.
16
18
20
22
24
26
FIGURE C2 TYPICAL BECK COOLING CURVES
-------
<£>
210
190
CYCLE EVENTS
DYE
3 4
DYE TIME (HR)
14
FIGURE C5 TEMPERATURE CYCLE FOR ATMOSPHERIC BECK WITH SCOUR
-------
DYE
TYPICAL DYE CYCLE
WITHOUT SCOUR
6 AND 7
CYCLE EVENTS (TABLE I AND TEXT)
2 4 6 8 10 12 14 16
TIME - HOURS
FIGURE C4 TEMPERATURE CYCLE FOR ATMOSPHERIC BECK WITHOUT SCOUR
-------
Sometimes it is necessary to add more dye during this
event. These "adds" are made at operating temperature.
(8) The dye solution is flushed out of the beck and rinsed
from the cloth. (This, again, requires much cold water.)
(9) Cold water is admitted to the beck again and is heated to
between 90°F and 120°F.
(10) Chemicals used to make the dye "color fast" are added to
the water.
(11) This solution is heated to about 120°F and maintained at
that temperature for about 15 minutes.
(12) These dye "fixing" chemicals are flushed from the beck
and carefully rinsed from the cloth. (Again, much cold
water is used.)
(13) The cloth is unloaded from the beck and sent to the dryer.
ANALYSIS OF ENERGY AND WATER CONSUMPTION IN THE DYEHOUSE
A detailed study of the energy and water usage in the dyehouse
was undertaken to determine the potential for reclamation. During
the course of this study two different methods were used to calculate
the energy and water consumption per pound of cloth processed. In
the first method an energy and water mass balance is made on the dye-
house without reference to the individual dyeing cycles. The dyehouse
is taken as a black box, and the total water and energy use is calcu-
lated from the inlet and discharge water flow rates and temperatures,
and the steam consumption in the dyehouse. The second methods is based
on the analysis of selected dye cycles to determine the energy and
water consumted per pound of cloth processed during that cycle.
Three cycles, two representing the average cycle (with and without
scour) and the other a typical cycle (with scour), were studied in
detail for this purpose.
The following outline will be adhered to in the analysis. First
the two methods will be discussed in detail with specific references
to the data used. Then the results will be compared and discussed.
Finally, losses from the becks during the dye cycle will be examined,
and suggestions will be made that may cut down the losses and reduce
operational expense.
Method I
The first method utilizes the lump analysis techniques. An energy
balance is made on the dye as shown in Figure C5.
STEAM CONVECTION AND RADIATION
COLD WATER DYEHOUSE WARM WATER
COLD AIR WARM, MOIST AIR
Figure C5. Schematic Energy Diagram.
141
-------
O + Q =0 + O
steam water in ^losses xwater out.
Thus knowing the inlet (filter plant) and discharge (composite
waste) water temperatures and flow rates and the steam consumption,
we can determine the total energy loss from the dyehouse. Furthermore,
from heat transfer considerations the convection and radiation losses
can be evaluated independently, making possible the determination of
evaporation losses from the becks (an essential point in evaluating
beck performance). The relative magnitude of the losses can be
studied to determine if energy-savings measures that might involve
capital expenditure can be justified. The energy content of the
composite waste can then be examined to evaluate the potential for
energy reclamation. The results of the calculations are given in
Table Cl.
TABLE Cl. TOTAL ENERGY QUANTITIES FOR THE MONTH OF MAY, 1975
(106 Btu)
Losses
°-steama AQwater outb Qevaporation ^radiation & covection
10428 4136 6116 176
Assuming 75% boiler efficiency and that 90% of the steam production is
used in the dyehouse.
Energy potentially available for reclamation is 4136 106 Btu.
The measured data include water inlet temperature, water discharge
temperature, and the water flow rate as determined by adjusting the
filter plant output to the amount that goes to the dyehouse. The
average temperature rise of the water in passing through the dyehouse
was 14.5°F during the three-day test period.3 The cloth production by
the dyehouse during that time was 320,000 yards, which at an average
weight of 1.26 Ibs/yd was 403,200 pounds of cloth. Taking into
account the fact that redyes for this type of production are in the
range of 30-40%, the dyehouse processed about 544,000 pounds of material,
The plant operated for 22 days in that month, and the dyehouse used
10,428 x 106 Btu's of steam and 29.2 x 106 gallons of water. Hence
the energy and water consumption per pound of cloth processed can be
calculated in the following way:
This temperature rise can vary from time to time as a result of
variations in the color and types of cloth dyed, and because the
steam and water valves are.controlled by the operating personnel
rather than an automatic controller.
142
-------
Energy used per pound of cloth processed: 10/428 x 10 = 19,170 Btu/lb
544,000
29.2 x 106
Water used per pound of cloth processed: 544 QOO = 53-7 gal/lb
Method II
The second method uses a typical (with scour) and two average
(with and without scour) cycles" to determine energy and water consump-
tion per pound of cloth processed in the dyehouse. As mentioned earlier,
two distinct dyeing cycles are used. The dye cycle withour scour
corresponds to about two-thirds of the total number of batches of cloth
dyed. The cycle with scour, now used only for automotive fabrics,
takes up the remaining one-third.
All the available temperature charts from the dyehouse of the
week of September 29, 1975, were examined. Averages were obtained
of the duration and operating temperature of each of the events
during a dye cycle, and with this information the average cycles
with and without scour were constructed. The typical cycle was
selected arbitrarily and happened to be a cycle with scour. The
temperature histories of the event of the average cycles with and
without scour and of the typical cycle are given in Tables C2, C3,
and C4, respectively.
Using this data the energy and water consumption per pound of
cloth processed were calculated for each one of these selected
cycles. The results of these calculations are given in detail in
Tables C2, C3, and C4 for the typical cycle, and average cycles with
and without scour, respectively. Also included in these tables are
the energy consumption to heat the contents of the beck from the
inlet to the operating temperature, the energy consumption to maintain
at operating temperature (to compensate for losses), the total energy
consumption, the hourly rate of total energy consumption, energy
content of the water in the beck, and the total water usage for
each event as well as the cumulative quantities for the cycle itself.
Table C5 compares the energy and water consumption per pound
of cloth processed for the selected cycles. Recalling that the
overall ratio of cycles without scour to cycles with scour is 2:1;
we may better represent the data by defining the "average" cycle as
"average" cycle - 2/3 (average cycle without scour) +
1/3 (average cycle with scour).
These cycles were assumed to be run in a 7-foot (width) beck with
an operating capacity of 1428 gallons (11,900 pounds) of water.
143
-------
TABLE C2. TYPICAL DYE CYCLE
Event 1 & 2
Duration min 91.5 .
Temperature °F 115
Energy Consumption
to heat from 65 °F
to operating temp B^u
(pVCpAT = QH) x 106 0.594
Energy Consumption
to maintain at
operating temp
Btu
(pVCp||AT - QM) X 106 0.192
Total Energy
Consumption Btu
(QH + QM + QT) X 106 0.786
Rate of Energy millons
Consumption at T °f fitu
(QT/At)a Per hour 0.15
Energy Content of
water in beck
about 65 °F
(including steam Btu
condensate) x 106 0.622
Water Flow Rate
Low/High gal/min 225
Water Used gal 1493
(hot)
Cumulative Energy = 16.093 x 10& Btu
Cumulative Water = 21339 gallons
Cumulative Hot Water = 4748 gallons
Cumulative Cold Water = 16635 gallons
345
37.5 30.00 .99.0
165 - 125
0.622 - 0.713
0.286 - 0.276
0.908 - 0.989
0.86 - 0.21
1.304 0 0.753
35/225 225
72 3520 1507
(hot) (cold) (hot)
QH = 3.600 x 10b Btu
QM = 12.409 X 106 Btu
6 & 7 8 & 9
552.0 45.0
208
1.042
11.597
12.639
1.41
1.944 0
35/225
125 6895
(hot) (cold)
Energy 16'093
ijjjciyy __ .
654.5 # of cloth processed ^^
Water 21339 -
fiR4 t;
10 &
21.0
125
0.713
0.058
0.771
0.21
0.753
225
1507
(hot)
x 106
.5
11 12
42.0
-
-
-
—
—
0
35/225
6220
(cold)
-24460 BTU/#
cloth
32.6 gal/# of clot!
a To heat at 3°F/min requires ^2.1 x 10 Btu/hour.
-------
TABLE C3. AVERAGE DYE CYCLE WITH SCOUR
tn
Event 1 S 2 3
Average
Duration min 43.2 85.8
Temperature °F 115 160
Energy Consumption
to heat from 65°F
to operating temp Btu
(pVCpAT = QH) x 106 0.595 0.560
Energy Consumption
to maintain at
operating temp
(pVCp^At - QM BtU 6
P Pdt V x 10 0.091 0.251
Total Energy
Consumption Btu
(QH + QM = QT) x 106 0.685 0.811
Rate of Energy millions
Consumption of Btu
(QT/At)a per hour 0.15 0.55
Water Flow Rate
Low/High gal/min 225
Water Used gal 1428 146
(hot) (hot)
Cumulative Energy = 13.947 x 105 Btu
Cumulative Water - 26720 gallons
Cumulative Hot Water = 4729 gallons
Cumulative Cold Water = 22035 gallons
QH = 3.477 x 106 Btu
QM = 10.426 x 106 Btu
For a week 109,936 # of cloth processed (183
600.74 ft/lot
13947 x 106 K
Energy 600_74 - 23,200 Btu/# of cloth
Wai-^-r .2.§7?-° - 44.5 rra linns/* of r-ln1-h
4 5 6 & 7 8&9 10 & 11 12
60.0 73.8 437.4 165.0 160.8 68.4
115 205 - 120
0.594 1.119 - 0.654
0.155 9.577 - 0.352
0.749 10.696 - 1.006
0.15 1.4 - 0.20
35/225 225 - 35/225 225 35/225
7345 1493 1627 7345 1512 7345
(cold) (hot) (hot) (cold) (hot) (cold)
a ,.
To heat at 3°F/min requires ^2.1 x 10b Btu/hour.
lots)
-------
TABLE C4. AVERAGE DYE CYCLE WITHOUT SCOUR
Event
Average
Duration
Temperature
1 & 2 3 4
min
oF
5
73.8
115
6 S 7
437.4
190
8 & 9
165.0
10 & 11
160.8
120
12
68.4
Energy Consumption
to heat from 65°F
to operating temp Btu
(pVCpAT = QH) x 106
Energy Consumption
to maintain at
operating 'temp
0.594
0.933
0.654
(oVCp^_?_AT ~ OM1
dt
Total Energy
Consumption
(QH + QM = QT)
Rate of Energy
Consumption
(QT/At) a
Water Flow Rate
Low/High
Water Used
Btu
x 106 -
Btu
x 106 -
millions
of Btu
per hour -
gal/min -
gal -
0.155
0.749
0.15
225
1493
(hot)
8.004
8.937
1.2
-
110
(hot)
0.352
1.006
0.20
35/225 225
7345 1500
(cold) (hot)
-
_
-
35/225
7345
(cold)
Cumulative Energy = 10.692 x 106 Btu
Cumulative Water = 17794 gallons
Cumulative Hot Water = 3055 gallons
Cumulative Cold Water = 14690 gallons
QH = 2.139 x 106 Btu
QM = 8.511 x 106 Btu
10.692 x 106
Energy —fnn ^A = 17,800 Btu/# of cloth
aTo heat at 3°F/min requires ^2.1 x 10G Btu/hour.
_
oU\J • r 4
_ 2Q.fi
600.74
of
-------
Using this definition the energy and water consumption per pound of
cloth for the "average cycle becomes
Energy consumption per pound of cloth processed = 2/3 (17,800) +
1/3 (23,200) = 19,600 Btu,
Water usage per pound of cloth processed = 2/3 (29.6) +
1/3 (44.5) = 34.6 gallons.
TABLE C5. COMPARISON OF WATER AND ENERGY USE FOR THE SELECTED CYCLES
Energy consumption per Water usage per Ib
Ib of cloth processed of cloth processed
(Btu) (gallon)
Typical Cycle
(with scour)
Average Cycle
(with scour)
Average Cycle
(without scour)
24460
23200
17800
32.6
44.5
29.6
Comparisons of the Results Obtained by the Two Methods
Table C6 compares the results obtained by the two methods. These
two methods for the calculation of energy consumption per pound of
cloth processed agree substantially. The discrepancy in the water
usage figures can be explained by the fact that the rate and duration
of the water usage in the cooling and rinsing phases of the cycle
is not predetermined or automated; it is entirely left to the
operator's intuition. For the lack of better information, we assumed
that each cycle used cooling water at the high flow rate of 225/gal
min for 30 minutes. These assumptions are partially justified by
Figure C2 and the dyeing process description.
TABLE C6. COMPARISON OF THE ENERGY AND WATER USAGE PER POUND OF CLOTH
PROCESSED AS DETERMINED BY THE TWO METHODS
Energy consumption per Water usage per Ib
Ib of cloth processed of cloth processed
(Btu) (gallon)
Method I (overall balance) 19170 53.7
Method II ("average" cycle) 19600 34.6
Losses
Losses from the becks are of two kinds: evaporation, and
radiation and convection. The nature and significance of each of
these will be discussed below.
147
-------
200
195 -
u.
i
UJ
tr
UJ
o.
Z
UJ
190 -
LA FRANCE PLANT
BECK COOLING CURVES
NO STEAM OR WATER FLOWING
CLOTH CIRCULATION IN DYE BATH
BECK NO. 28, JAN. 29, 1976
O DOOR AND DAMPER CLOSED
D DOOR CLOSE. DAMPER OPEN
< DOOR AND DAMPER OPEN
TIME-MIN.
148
-------
Evaporation Losses — During the dyeing cycles steam is bubbled
directly into the dye solutions to heat the beck contents to, and
maintain them at, the required temperature. Since the becks operate
at atmospheric pressure and near the boiling temperature for several
hours per dye cycle, evaporation results in a large energy loss.
This vapor escapes from the becks either by leakage or as a result
of forced ventilation. Determined from the air flow rate and
temperature and specific humidity data, this vapor loss amounts
for more than two-thirds of the energy supplied to the dyehouse.
The loss was confirmed by the overall dyehouse energy balance (Table
Cl) and can also be calculated from a heat loss/temperature plot (such
as Figure A2) after the steam flow is stopped and before the cooling
water flow begins. The rate of evaporation losses at any temperature
can thus be determined from Figure A2 since the energy loss from the
becks is mainly due to evaporation from the high-temperature contents.
This loss can be considerably reduced either by replacing the
atmospheric becks with pressurized units or by installing energy
recovery equipment.
Radiation and Convection Losses — The radiation and convection losses
are so significant (Table Cl, relative to the evaporation losses) that
purchase of equipment to reduce them may not be justified.
DISCUSSION OF RESULTS AND CONCLUSIONS
The results presented in Tables Cl, C2, C3, and C4 show two things
of great importance in this report. First is the large waste of energy
because the dye becks are open to the atmosphere and operated for
rather long periods of time at temperatures about 180°F. Evaporation
of some of the hot water and the discharge of this vapor to the
atmosphere accounts for about two-thirds of the energy supplied in
steam to the dyehouse. This energy loss is not associated with the
wastewater stream discharged to the sewer, but it is a tremendous
amount of energy which may be conserved. The numbers in Table Cl
indicate that, for fuel costing $2.75 per million Btu (delivered to
the dyehouse), this energy loss amounts to more than $750 per day.
The second important factor is the large amount of energy discharged
to the sewer in wastewater. Table Cl describes this as energy potentially
available for reclamation, and Tables C2, C3, and C4 describe similar
values as the "energy content of water in beck above 65°F" (including
steam condensate). Though less than the evaporation loss, the loss in
hot water discharged to the sewer is substantial. If the recoverable
chemicals and water are included in the consideration, the daily cost
of this wastewater becomes more nearly equal to that of the evaporation
loss. The potential thermal energy recovery as indicated in Table Cl
amounts to about $500 per day savings. This quantity is of primary
interest in the present project.
149
-------
APPENDIX D
ENERGY USE ANALYSIS OF A CONTINUOUS DYE RANGE
ENERGY CONSUMPTION BY PROCESS
This Appendix discusses the energy use of the observed dye range.
Energy balances applied to the process equipment were made to determine
the energy loss by evaporation from the wash tubs, the energy lost to
the drain, and the energy recovered by means of the recovery heat exchanger.
Data
Data were required for water flow rates and temperatures for ten
wash tubs used in the dyeing range. These data, given in Figure Dl were
obtained as follows:
(1) Water flow rates were pro'vided by plant personnel. A
rotoameter-type flow meter was installed on the supply
line to one wash box to confirm the reported flow rate.
(2) Water temperatures were measured using a hand held mercury-
in-glass thermometer, a bimetal thermometer, or an iron-
constantan thermocouple connected to a temperature recorder.
Cloth temperatures were measured by holding the thermometer
against the wet, moving cloth in places where there was
sufficient water to give flooding of the sensing element.
Evaporative Heat Loss
Evaporative heat loss is a significant contribution to the total
energy losses for the dye range. This latent heat loss is associated
with (1) the evaporation from the water and cloth surfaces at the
wash tubs, and (2) the evaporation from cloth surfaces as the
cloth is transported between wash tubs.
The latent heat loss was measured in field tests as a function
of water temperature as shown in Figure D2. A curve proportional to
saturation pressure of water is fit through the data. This curve
was used as the basis for the latent heat loss calculations for each
wash tub.
150
-------
CLOTH*-
OVERALL RANGE
DYE
PAD
1. R.
PREHEATER
THERMOSOL
OVEN
COOLING
CANS
CHEM
PAD
M
STEAMER
7.5 6PM f
140° F CONDENSATE '
IOGPM WATER SEAL
2IO°F IOGPM
100 °F
WASHERS
(SEE BELOW)
I
DRY
CANS
CONDENSATE
WASHER SECTION
WASH
TUB
1
WASH
TUB
2
WASH
TUB
3
WASH
TUB
4
rt \\
II.8GPM II.8GPM ll.8GPMIt.8GPM
MO°F I!2°F I30°F I4I°F
OXID
5
moN
6
I
2GPM
I92°F
WASH
TUB
8
WASH
TUB
9
WASH
TUB
10
2
I58°F
FIGURE Dl SCHEMATIC DIAGRAM OF DYE RANGE
I.8GPM
I60°F
*
II.8GPM II.8GPM
I50°F I50°F
-------
Point 3
CURVE PROPORTIONAL
TO VAPOR PRESSURE
140 150 160 170
TEMPERATURE — *F
180
190 200
FIGURE D2 ENERGY LOSS BY EVAPORATION FOR DYE WASHERS
-------
The specific data were obtained from two experimental procedures.
The first procedure involved equating the temperature drop of the
saturated cloth in moving from one tub to another at the rate of
energy loss by evaporation. The temperature of the wash tub was
adjusted by controlling the proportion of hot and cold water supplied
to wash tub #8 so that the water entering and exiting the tub was at
the same temperature (145°F). The energy in flow rate associated with
the cloth entering from wash tub #7 can be equated to the loss from
the wash tub. The observed temperature of cloth entering the tub #8
was 182°F. The loss of temperature from 192°F to 182°F was associated
with evaporation between tubs, and tub #8 was credited with receipt
of cloth and water at 182°F. Point 1 in Figure D2 was obtained by
this procedure.
The second procedure was concerned with equating the cooling
rate of a heated wash tub (#8) to the latent heat loss at the water
and cloth surfaces within the wash tub. Steam was admitted to the
tub to achieve a steady elevated temperature. Then the steam flow
was terminated allowing the mass of fluid in the tub to cool. The
total energy decline rate thus determined with no water throughflow
can be equated to the loss by evaporation. Points 2 and 3 in
Figure D2 were obtained by this procedure.
Using the heat loss schedule in Figure D2, an evaporative
loss prediction for each wash box is shown in Table Dl. There are
entries thereon for operation at both standard procedure temperature
and at the observed process temperature.
TABLE Dl. EVAPORATIVE LOSS ESTIMATE
Wash Box
1
2
3
4
5
6
7
8
9
10
Standard Procedure
Temp (F)
120
120
120
120
135
135
190
175
175
175
Evaporative
Loss/Btu/hr
86 x 10^
86
86
86
128
128
466
335
335
335
Observed
Temp (F)
110
112
130
141
158
156
192
160
150
150
Evaporative
Loss Btu/hr
64 x 103
68
112
148
224
213
487
235
184
184
Total
2071 x 103
1919 x 103
A certain amount of the total evaporative loss emanates from the
section of cloth between the wash boxes. Covers on lids on the boxes
would have no effect on this loss. Table D2 shows the predicted amounts
of such loss for each washer connecting section for the observed operating
153
-------
temperature schedule. The basis for the calculation assumes an equal
loss per unit of exposed wet surface and having the temperature trend
of Figure D2. Such an assumption yields results which agree well with
attempts to measure the temperature of the moving cloth surface.
TABLE D2. EVAPORATIVE LOSS ON CLOTH BETWEEN TUBS
Section Following
Wash Box #1
2
3
4
5
6
7
8
9
10
Total
Evaporative Loss Btu/hr
6 x 10*
7
12
15
23
22
51
24
19
19
2 x 105
Energy to Heat Water_
An estimate of the amount of energy used to heat water was obtained
from the equation
O (Tdrain ' 65)
where qarain is heat leaving the drain, m is the water flow rate, c is
water specific heat, T
-------
Item
Wash Box 9
10
Steamer Condensate
Steamer Water Seal
TABLE D3. (continued)
Observed Observed
Flow Ib/hr Temp F
5897
5897
5000
5000
150
150
210
100
Energy to heat
water Btu/hr
5.01
5.01
7.25
1.75
Total 40.6 x 105 Btu/hr
oven cooling water 3750 140 2.81 x 5 Btu/hr
Also shown in Table D3 are the contributions to an energy balance
of the steamer condensate flow, water seal flow, and the cooling cans
which follow the thermosol oven. The cooling cans represent an
energy recovery from the cloth leaving the oven and thus do not con-
tribute to the energy required for the range. The values for the
water seal and condensate do represent heating requirements. The
magnitude of the condensate flow of 5000 Ib/hr implies that heat is
lost sensibly in an amount of approximately 5 x 106 Btu/hr. This
number is much larger than we believe can occur with air cooling and
radiation from the steamer surface. Order of magnitude calculations
suggest about 1.5 x 105 Btu/hr should be used to heat the cloth and
water from 80°F to 210°F and perhaps 1 x 105 Btu/hr should be lost
by direct losses to the air and surrounding environment. The flow
measurement of the condensate drain may not have been accurate at the
10 gallons per minute rate shown in Figure Dl but is about twenty times
as great as predicted above. Consequently, we allow the following
options: (1) the observed flow was not a representative average
condition, (2) the observed flow has other sources in addition to
condensate, or (3) some very much larger than estimated heat loss
mechanisms occur.
Table D4 shows a predicted heat loss rate for radiation and con-
vection from the outer surface of the equipment to the surroundings.
The prediction, from experience, is known to be complex and subject
to considerable uncertainty. However, an upper bound calculation
may be performed and predicts a fairly small loss compared with other
energy items. This upper bound has used conventional convection heat
transfer correlations at maximum probably air velocities and maximum
radiative heat transfer coefficient. Such a procedure results in a
composite upper bound heat transfer from the source to 85°F at an
overall combined heat transfer coefficient of 2 Btu/hr-ft2-°F (see
Appendix B).
TABLE D4. ENERGY LOSSES BY RADIATION AND CONVECTION
Item
Wash Box 1
2
Energy Loss Btu/hr
0.53 x 10H
0.57
155
-------
TABLE D4. (continued)
Item
Wash Box
Steamer
Energy Loss Btu/hr
3
4
5
6
7
8
9
10
0.96
1.19
1.56
1.51
2.29
1.61
1.39
1.39
15.3
28.3 X
104 Btu/hr
Energy Recovery
On a plant scale, energy recovery is effected by collecting the
hot drains into a feed stream to the hot side of a heat exchanger.
A hot water supply is provided at 150°F. About 35% of the hot water
is heated in the heat exchanger to approximately 110°F, then is
heated by steam to 150°F. The other 65% is heated by steam only.
For an incoming temperature of 65°F, the heat recovery fraction of
the hot water may be estimated. If mf is the flow rate of 150°F
water, the total energy added to the hot stream is
QTotal = ** CP (15° " 65) '
while the energy supplied in the heat exchanger is
QH/X = 0,35 mfcp (110 - 65) .
The fraction of thte total amount supplied is the ratio or
_ 0.35 (110 - 65) = 0.185.
Qrotal ~ <150 - 65>
While the ratio is valid for the entire plant energy recovery, it
does not adequately represent the particular dye range of interest.
To properly credit the energy recovered from the dye range,
its contribution to the total recovery is estimated. The water
from the condensate drain, wash tubs #8, #9, and #10 compose the
hot drain from the range. All fluid collected from the various
hot drain sources is cooled to approximately 95°F at the heat
exchanger exit. Therefore, neglecting heat losses in transit from
the individual drain to heat exchanger entrance, the energy recovered
from each stream is
™icp Ti
156
-------
Here m.^ is the mass flow of drain i at temperature T- and is the
more specific heat value. The sum of these energy recovery amounts
for each "hot" drain is that amount credited to the dye range heat
recovery. The amount is 1.6 x 106 Btu/hr for the above listed group
of component drains. This represents 40% of the energy used to
heat water on the range.
Results and Discussion for Energy Description
The emphasis here has been to identify the modes of energy loss
and to determine the amount of heat which is recovered for the dye
range. Energy losses by evaporation and radiation and convection
and energy flow in the wastewater stream were examined.
The energy loss from the wash tubs by evaporation represents a
relatively large quantity which could feasibly be conserved with an
alternative equipment design. As shown in Table D2, the evaporative
heat loss from all wash tubs was calculated at 1.92 x 106 Btu/hr.
The evaporative heat loss from the saturated moving cloth surface
as it was transported from one tub to another was only about one-tenth
of the total energy loss by evaporation. (About 4.06 x 106 Btu/hr
of energy was used to heat water from 65°F to process temperature
and about 1.6 x 106 Btu/hr was recovered in the plant heat exchanger.)
The energy loss by radiation and convection was only about 15% of the
energy loss by evaporation.
The total energy loss minus the recovered energy was found to be
4.66 x 10 Btu/hr. The use of energy was as follows: evaporation
losses (30%), net to heat water (65%), radiation and convection
including steamer (5%). These values are depicted graphically in
Figure D3 and tabulated in Table D5.
TABLE D5. ENERGY CONSUMPTION FOR DYE RANGE NO. 8
Energy Loss - Evaporation 1.92 x 105 Btu/hr
to heat water 4.06 x 106
Radiation & Convection 1.30 x 105
Steamer 1.53 x 105 _
TOTAL 6.26 x 106
Less Energy Recovered (from H/x) -1.6 x 106 _
Net Energy Supplied 4.66 x 10 b
4.66 x 106 — ( * hr )
Net Energy Consumption = _! - hr fin m-i^ _ 1262
c, c Ibm cloth Ibm cloth
bl. b - ; -
mm
157
-------
CONTINUOUS DYE RANGE
45%
!3.8xl06BTU/HR
103 GPM
3690LB/HR
ENERGY
WATER
PRODUCTION
DISTRIBUTION
ENTIRE RANGE
OF ENERGY
I6°X
13%
24%
I.R. OVEN WASHER POST
PREDRYER SECTION DRYER
65%
an
00
6.2xl06BTU/HR
103 GPM
DISTRIBUTION OF ENERGY
WASHER SECTION OF RANGE
H.E.
HEAT
RECOVERY1
30%
5%
HEATING
WATER
VAPOR
LOSSES
HEAT
LOSS
FIGURE D3 DISTRIBUTION OF ENERGY ON DYE RANGE
-------
The above values are for the wet portions of the dye range.
The entire range from the pre-dryer through the dry can post-dryer
consumes an amount estimated as 7.5 x 105 Btu/hr in addition to the
amounts already delineated. Thus, the wet processes only include
approximately 45% of the total energy supplied to the range.
Conclusions for Energy Description
Evaporation from the water and cloth surfaces of the wash tubs
amounted to 14% of the total energy lost in the dye range. Since this
evaporation is a direct loss to the plant and represents a significant
item for energy savings it should be minimized. Proper equipment
design that provides for covered wash tubs would decrease evaporative
energy losses.
Almost two-thirds of the wet portion energy is lost to the drain
and represents the largest area for energy and cost savings. The
potential for energy conservation by recycling water is significant.
If only 80% of the energy now used to heat water could be recycled
in the dye range, the energy savings would be about 3.2 x 106 Btu/hr.
As the system now.operates only 1.6 x 106 Btu/hr is being recovered
to heat incoming fresh water.
There are two relatively inexpensive energy savings devices
apparent from the analysis conducted associated with this report.
These are the addition of tub covers and the rerouting of one of
the drain lines. Our analysis results in an evaportion heat loss
of 1.92 x 106 Btu/hr from the tubs, with 1.0% of this or 0.2 x 106
Btu/hr associated with the cloth between tubs. Covers over the
tubs would reduce the heat loss up to 1.7 x 10 6 Btu/hr. We are not
able to estimate the effect of an imperfect cloth seal on the approach
to this value stated.
The water drain from the cooling cans following the thermosol
oven constitutes nearly the flow supplied to one wash box and was
observed to be at 140°F. This water is clean and has some head
allowing it to be routed to provide flow to one of the tubs. A
saving of 40,000 Btu/hr is expected to result (or 80,000 Btu/hr
if the flow does not contribute to the heat recovery system) .
The steamer condensate represents an additional energy source.
Depending on its quality with respect to the quality required for
wash water, it could be pumped into a tub directly at a savings
potential of over 300,000 Btu/hr. We tend to believe this number is
not representative of steamer actual condensate rate, so the matter
should be investigated further.
Based on data received from plant personnel, the use of the
Max Trac urethane roller prior to the dry cans would decrease the
energy flow an estimated 1,800,000 Btu/hr. This estimate assumes
159
-------
a water content of 0.7 pounds for the standard roller versus 0.3
pounds for the improved roller per pound of cloth. At a cost of
$7000 for the roller and $2 per 106 Btu for energy, the roller
should pay for itself in under 2000 hours. In future installations
the superior roller will allow the use of half the number of dry cans
or fewer and better drying control with less dye migration or anti-
migration chemical. Thus, a new installation should be both cheaper
overall and have substantially lower energy requirements. In present
installations the urethane roller probably should be implemented
along with a reduced number of active cans. The unused cans should
be unheated and removed from the cloth path. This would provide for
a minimum of direct heat loss from the cans.
160
-------
APPENDIX E
ANALYSIS OF MASS REMOVAL FROM CLOTH IN PREPARATION
It is desired to remove as much as possible of the warp
sizing from material in preparation while using a minimum of water
and equipment. The presence of substantial amounts of size in the
water wash and chemical desize wash of the open width preparation
range has been shown in the text of this report. Herein an
investigation into the completeness of the removal from the cloth
is considered.
The cloth considered was a 65/35 polyester/cotton blend
with a size composed of PVA, CMC, and sizing wax. Other substances
may also be present: spinning oils, natural waxes, and dissolved
or removed motes. The total amounts removed by a boiling exercise
for 15 minutes in detergent were 9% by weight of fabric.
A simple exercise at graduated boiling intervals up to 5
minutes showed that essentially all material may be removed in
30 seconds to one minute at boiling conditions. By using a mild
Freon extraction, the sizing wax could be selectively removed from
the cloth, leaving the CMC and PVA. Thus it was possible to
separately analyze for the removal of wax and PVA/CMC.
Cloth samples were removed following the firs t washer (water
washer) and following the chemical desize washer. Selective
weighings show that the removal of wax are 12 and 25% respectively
in these two steps. Also 58 to 39% of the PVA/CMC are removed by
these two steps. Thus the sizing wax persists to the caustic
washer and only 3% of the PVA and CMC persist to the caustic washer.
161
-------
APPENDIX F
OPTIMIZATION COMPUTER PROGRAM
This facet of the Energy Project is concerned with the
development of a computer program to evaluate the relationship
between operating parameters and net cost for a hyperfiltration
system. For known resource values and given operating conditions,
this program was conceived to seek out the greatest profit or least
loss to the operation. The validity of the discovered optimum must
be scrutinized in light of the assumptions made to model mathematically
the hyperfiltration process. It is the purpose of this section to outline
those assumptions, in particular the definition of the assumed hyper-
filtration system, the economic simplications, and the method of
characterizing the membrane performance. A brief discussion of the
means of optimization and an overview of some of the results will
also be presented.
The hyperfiltration system which has been modeled is schematically
illustrated in Figure Fl. Depicted is an industrail process which
produces chemically contaminated wastewater that may include chemicals,
detergents, dye and other wastes not necessarily limited to those
generated by the manufacture of textile goods. This wastewater is the
input to the hyperfiltration system. The desired outputs may include
reuseable water, reclaimed chemicals, and heat. At the heart of the
assumed hyperfiltration systems is a tubular module that separates
soluble chemicals into dilute (permeate) and concentrated streams.
The module consists of a series of tube bundles acting as porous
supports for the membrane which is deposited on-the tubes. Two
pumps are in the hyperfiltration systems, one for primary flow and the
other to recirculate concentrate flow so that is might be brought to
a higher concentration. An evaporator completes the array of equipment
in the hyperfiltration system. The evaporator further concentrates
the solutes in the stream so they may achieve value as reclaimed
chemicals. The hyperfiltration system assumed in this study then
consumes contaminated water, electrical power for pumping, and thermal
energy for evaporation; it produces clean water, reclaimed chemicals,
and recoverable heat.
The portion of the computer program that assesses the net cost of
operation is an attempt to characterize only the five most prominent
expenditures or benefits. The costs are divided into two groups, capital
costs and operating costs. Capital costs are limited to expenditures
for the tube module and the pumpage. The cost of the evaporator is
not included in this study. The cost of dollars per tube includes
162
-------
INDUSTRIAL
PROCESS
WATER CONTAMI-
NATED WITH CHEM-
ICALS (DYE,SALT,ETC.)
PRIMARY FLOW
RECIRCULATION FLOW
MODULE
FLOW
TUBE MODULE
PRODUCT (CLEAN) WATER
CONCENTRATE FLOW
EVAPORATOR
RECLAIMED CHEMICALS
FIGURE Fl SYSTEM DESCRIPTION
-------
the averaged cost of the tube, membrane, housing, and fittings for the
module and is an input quantity, i.e., $45/tube. The cost for pumps
is assumed proportional to the power required, i.e., pressure rise
multiplied by primary flow plus recirculation pressure rise multiplied
by recirculation loop flow. The constant of proportionality is input,
i.e., $X/kilowatt.
Operating costs are figured as dollars per day. The optimum cost
function, which is in dollors, is calculated on the assumption of a
300-day operating period. Therefore, an optimum cost of zero indicates
that after 300 days of operation the initial cost and operating costs
were exactly offset by savings of chemicals, water, and/or heat. A
negative optimum reflects a benefit to the operator, and a positive
number reflects a deficit.
Of the three operating costs, the first is for the consumption of
electrical power by the pumps. This cost is proportional to the number
of kilowatt hours required for a day of pumpage with the constant of
proportionality input, e.g., .013 dollars per kilowatt hour. The other
two operating costs reflect savings through clean water and recoverable
heat in the product flow stream and reclaimable chemicals in the con-
centrate flow stream. For savings associated with the product stream,
a credit is assigned for both clean water and its associated thermal
energy, and the credit is negated if the product water contains excessive
concentrations of solutes. Recoverable thermal energy here refers to
fluid having temperature above 293°K. The savings associated with re-
claimable chemicals in the concentrate flow is examined on a solute-by-
solute basis and summed over all the solutes. If a solute in the stream
has a concentration equal to or higher than the desired concentration for
reclamation, then that solute is credited in dollars per kilogram with
the: full value of reclamation. If, however, a particular solute has a
lower concentration than that desired for reclamation, it must be
evaporated until the concentration reaches a reclaimable value. The
dollars per kilogram credit is accordingly reduced by the cost of energy
input for evaporation. If the cost of energy needed for evaporation to
concentrate a particular solute is more than a solute is worth, then
the solute is not reclaimed and the savings for that solute is zero.
This analysis assumes that there is no cost of separating solutes and
that no heat is reclaimed from the concentrate stream even if no evapora-
tion is needed. The investigation at Clemson has assumed up to six
solutes in the primary flow.
The mathematic modeling of the performance for the membrane is
accomplished by accounting for the influence of temperature and pressure
on the membrane flux and the influence of velocity and pressure on the
intrinsic membrane rejection. Flux is the term used to describe amount
of product flow per area of membrane. A flux representative of real
membranes examined at Clemson operating at 349°K and 6.5 x 106 Pa (950
psi) is input at the beginning of the program. This flux is modified
by a pressure function and a temperature function. The temperature
modification is of exponential order with higher temperatures yielding
larger fluxes. The pressure modification is linear with higher pressures
164
-------
also yielding larger fluxes. The temperature of the primary flow is a
constant that is input at the beginning of the program and does not
change from section to section in a single design or from design to
design. The pressure used to evaluate the pressure modification is
that at the inlet of each section of tubes. This pressure changes from
section to section and from design to design.
Rejection is a term that describes how efficiently a membrane
separates solute and solvent. A membrane rejection is input at the
beginning of the program for each of the solutes in the primary flow
stream. These rejection values, characteristic of membrane behavior
at high operating pressures, are modified by a pressure function and a
velocity function. The pressure function is such that at high operating
pressures the rejection values tend to approach the input values, and
rejection drops off slightly as the pressure tends toward zero. The
velocity modification assumes one of two values for each velocity,
depending on whether the Reynolds Number indicates a turbulent or laminar
condition for the primary flow through that section of tubes. The
turbulent case results in smaller fluxes at high velocities and larger
fluxes at lower velocities than the laminar case does. The rejection
is reduced in both cases for high velocities due to concentration
polarization. The pressures and velocities used for the rejection
value modifications are those existing in each section and therefore
are subject to change from section to section. The nature of the
pressure modification of solute rejections has its analytical basis
in the Onsager equations, the coupling between solute and solvent
fluxes, and the distribution coefficient.
Optimization
Optimization is initiated by defining certain input parameters.
The cost function to be optimized depends upon these input parameters
that are left as "floating" variables. These variables are given an
arbitrary value that will be changed as the program iterates and seeks
the lowest possible cost. For the purpose of this program, four
variables are designated as floating variables, and they form a four-
dimensional space within which the cost functions will be optimized.
The four variables are the system recovery (REG), which is the ratio
of product flow to primary flow; the recirculation ratio (RERAT),
which is the ratio or recirculation flow to primary flow; pump outlet
pressure (POUT), which is the pressure directly upstream of the tube
module; and VELMIN, which is the velocity as near upstream of each
tube as possible.
The main program controls the optimization process and assigns
starting values to REC, RERAT, POUT, VELMIN. The main program
also determines the incremental step size for each variable that
will be taken in a positive and then a negative direction from the
initial operating point. Only one variable has a constant step size:
VELMIN, at 0.1 meters per second. The other three step sizes are
165
-------
calculated at each step and are not necessarily constant. The technique
employed in the optimization code is a straight-forward hill-climber.
An operating point in four-variable space is assigned, and, with three
variables held constant, the fourth is perturbed slightly in a positive
direction and then in the negative by the amount of the step size.
The three values of the perturbed variable are sent to the subprogram
SOLUTE one at a time. SOLUTE establishes a separate module design
according to the specifications demanded by each value of the perturbed
variable. A total cost is evaluated for each of the three module
designs and sent back to the main program for analysis. The main
program examines the three costs, ascertains the lesser of these, and
adopts the value of the variable that produced the lower cost design
as the new operating point. The main program then cycles to the next
of the four variables and repeats the procedure. An optimum is
determined when none of the perturbations around the operating point
in any of the four floating variables produces a module design of lower
cost. The main program then signals the user that only higher costs exist
in the vicinity of that operating point, which is "therefore a minimum.
An important aspect of the optimization process is the method by
which the module design is established. The tube module is developed
from inlet to outlet by calculating the number of tubes that are needed
to handle the flow at each section. The flow that presents itself at
the inlet of the module is a sum of the primary flow and the recirculated
flow. The number of tubes required in the first section is calculated
by the equation:
tubes in a section = flow at inlet
cross sectional area per tube x minimum velocity
This calculation is not based on the actual velocity in that
section but on the minimum velocity (VELMIN) input from the main
program. The actual velocity is calculated and used in the mass
transfer equations to arrive at the amount of product flow produced
by the first section of tubes. This product flow is subtracted from
the inlet flow to arrive at the amount of flow going into the second
section. The number of tubes for the second section is calculated
by the above equation using the same value of minimum velocity. This
process is iterated, and the result is a number of tube bundles in
series that form the module. A design is finally arrived at when
enough sections of tube bundles exist to produce the amount of product
water demanded by the recovery ratio.
Of course, there have been some problems in the development and
application of this program, it was early discovered that the cost
function was not smooth but contained many peaks, cliffs, and holes,
mostly because sharp cutoff values were used to regulate the peripatetic
wanderings of some variables and functions. This problem was solved by
introducing smoothing functions with gentle incentives not to wander
too far in certain directions. Another continuing problem is the
discretization of the number of tubes in a section; that is, a section
contains only an integral number of tubes and cannot accommodate 1/2,
166
-------
3/7, or any fraction of tubes. This leads to small steps in the cost
function but is estimated not to deter the hill-climber optimization
technique from finding a minimum. It does, however, eliminate the
use of a faster, more efficient optimization technique based on the
directional derivative. A third problem is that the program may
seek a local "false" minimum and not the ultimate minimum located
within the search domain. As yet no means exists to tell a false
minimum from the ultimate minimum except to try many starting places
and compare the discovered minima.
The computer program has been exercised only a modicum due to
two factors. First, the program was constructed assuming the input
of an acceptable permeate quality. All permeate water has been decreed
reusable by plant personnel based on their experience and laboratory
tests described herein. Thus the limit has not been reached. Because
most of the intricate details of the optimization procedure are related
to water quality the lack of a quality limit specification causes the
bulk of the exercise to have no meaning.
Second, the main design problem of optimizing costs to achieve
a known water production has not been modeled. Since the flux
dependence on concentration velocity, pressure, and temperature is
pronounced a much more profitable exercise could be conceived. Such
a program would have been beneficial in the design section of this
report.
Despite these shortcomings, the program indicated some general
trends. The effect of recirculation on water quality is disasterous,
even in small amounts. Only when perfect (>99%) rejection is achieved
or when permeate contamination is unimportant will recirculation
be tolerable. The recovery of material in concentrated form will be
diminished also by recirculation.
High pressure operation is indicated to be attractive. Partly
this is due to the short (300 day) evaluating period which tends to
decrease the relative operating cost penalty compared with the capital
outlay.
A factor not evaluated but pertinent is the estimation of
temperature effects. The use of high recovery recycles a major part
of any energy increment. Processes may benefit by higher temperature
as well and the cost is nearly exclusively tied to increased evaporation.
The appropriate question is, does the decreased cost of membrane
installation pay for the increased energy cost of evaporation? An
appropriate evaluation must include the flow versus temperature for
the process itself.
167
-------
APPENDIX G
TEXTILE WASHING STUDY
The evaluation of washer technology is a logical complement to an
evaluation of hyperfiltration for wastewater reclamation. Primarily
this is true because improved washer operation will reduce water
consumption without reducing washed off solids. The cost of membrane
filtration is estimated to be roughly proportional to the water flow
and the reclamation benefits are proportional to the concentration of
reclaimed material. Even without the implementation of membranes,
process efficiencies can benefit immediately by enhanced washer
technology.
A survey of washing equipment manufacturers concerning their
design information and field experience has been completed. Informa-
tion herein is primarily resulting from the following companies:
Gaston County Dyeing Machine Corporation, B-K Textile Machinery,
Morrison Machine Company, and Menoyama Kiko Company Limited.
Copies of various technical papers pertinent to the present
investigation have also been accumulated. Olson and Lyons8 showed
the effect of heat setting on the polyvinyl alcohol size removal
fraction. The study indicated an outstanding effect of wash water
temperature. Parrish9 also has studied the effect of wash water
condition on the removal of NaOH. The tendency of industrial plants
toward counterflow washing also suggests that mathematical modeling
could be beneficial.
Washing involves removal of sizing materials, scouring, bleaching,
and rinsing the cloth. Washing effectiveness is determined by the
amount of material removed from the cloth in the washing process.
Dry cloth enters a desize washer to remove sizing material, wax,
and other lubricants added to the fabric strands during weaving. The
water in the desize washer contains surfactant to reduce surface
tension and allow penetration of water into the cloth. If the cloth
8Olson, E. S. and D. W. Lyons, "Effect of Drying and Heat Setting
Temperatures on the Removal Characteristics of Polyvinyl Alcohol Size,"
Textile Research Journal 42, 4, pp 199-202, 1972.
9Parrish, G. J., "Continuous Rinsing of Impurities from Textile
Fabrics," American Dyestuff Reporter, Vol. 54, No. 5., pp 33-39, May 1965.
168
-------
is a polyester blend that uses either PVA or CMC as the lubricant
then desizing is not necessary and scouring would be the first cloth
operation. If the cloth is cotton or a cotton/polyester blend, then
starch is generally used as a lubricant and either enzymes or HCl are
added to the wash water in the desize operation to chemically break
down the starch in preparation for scouring. PVA and CMC are at least
soluble in water while starch is not soluble. Presumably the sugars
such as fructose produced from the starch are soluble in the water.
The effectiveness of the washer can be measured in terms of the
concentration of sizing on the fabric before and after the wash period.
This is expressed as
C2.
where
C2 = concentration of sizing (i) on fabric as it leaves the tub
C = concentration of sizing (i) on fabric as it enters the tub
The effectiveness varies from 0 when C0 = C, (the water has
M 1i
not changed the concentration of sizing on the fabric from what it
was initially) to 1 when
-------
positions has its own advantage, though the laboratory approach
probably is the only way to achieve truly different results from
those being observed at present.
It appears that mathematical modeling of a group of washers could
be used to profitably evaluate the effect of counterflow. The model
would use overall washer performance data from the field as a basis.
The modeling of a single washer is not as simple and would require
either additional laboratory work or field data to substantiate it.
It is considered beyond the scope of the present work to attempt
such a model.
170
-------
REFERENCES
1. Brandon, C. A. and J. J. Porter, "Hyperfiltration for Renovation
of Textile Finishing Plant Wastewater," EPA-600-2-76-060, March 1976.
2. Brandon, C. A., J. J. Porter, and D. K. Todd, "Hyperfiltration for
Renovation of Composite Wastewater at Eight Textile Finishing Plants,"
Final Report, Grant S802973.
3. Brandon, C. A. and Max Samfield, "Application of High Temperature
Hyperfiltration to Unit Textile Process for Direct Recycle," (Membranes:
Desalination and Wastewater Treatment Conference) Jerusalem, 1978.
4. Dahlheimer, J. A., D. G. Thomas, K. A. Kraus, and J. R. Love,
"Applications of Hyperfiltration of Treatment of Municipal Sewage
Effluent," FWQA Report ORD-17030EOH01/70, 1970.
5. From General Electric Survey, approximately 1972 information.
6. Woodall, L. C. and E. F. Godshall, "Energy Economics in a Dyehouse,"
presented at Clemson University, Energy Conservation in Textile Industry,
January 12-13, 1977.
7. Holman, J. P., Heat Transfer, McGraw Hill, 1976 (4th Edition).
8. Olson, E. S. and D. W. Lyons, "Effect of Drying and Heat Setting
Temperatures on the Removal Characteristics of Polyvinyl Alcohol Size,"
Textile Research Journal 42, 4, pp. 199-202, 1972.
9. Parrish, G. J., "Continuous Rinsing of Impurities from Textile
Fabrics," American Dyestuff Reporter, Vol. 54 No. 5, pp. 33-39, May 1965.
171
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-79-131
2.
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Energy Conservation Through Point Source Recycle
with High Temperature Hyperfiltration
5. REPORT DATE
June 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
J.L.Gaddis, C.A.Brandon, and J.J.Porter
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Clems on University
Department of Mechanical Engineering
Clemson, South Carolina 29631
10. PROGRAM ELEMENT NO.
1BB610
11. CONTRACT/GRANT NO.
Grant R803875
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 7/75 - 12/78
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES T£RL-RTP project officer is Max Samfield, Mail Drop 62, 919/
541-2547.
is. ABSTRACT
results of a. study of energy conservation effects of point
source recycle with high-temperature hyper filtration (HF) in the textile industry.
(HF and ultrafiltration (UF) are pressure-driven membrane processes which have
potential for recycle of water, energy, and chemicals in wet finishing operations.)
The reuse of water , energy , and chemicals can be best achieved if separations are
applied to individual point-source streams rather than to total-plant mixed effluents .
Five wet processes comprise a large fraction of total textile operations and require
over half of the total energy used: preparation in rope and open- width ranges, and
dyeing in continuous thermosol ranges , in pressure becks , and in atmospheric becks
Plant sites were visited and data taken on operations on which to base estimates of
potential energy and materials to be saved. Each process effluent was sampled and
analyzed to determine which membrane (HF or UF) should be used. Two small equip-
ment skids allowed membrane operation at the plant sites. The permeate water in
each case was reusable. Estimates of energy recoverable per mass of cloth proces-
sed (kJ/kg) for each operation are: rope preparation, 2646; open-width preparation,
5766; continuous dyeing, 2449; atmospheric beck dyeing, 20,115; high-pressure beck
dyeing, 3910; and low-pressure beck dyeing, 1964.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Pollution
Textile Finishing
Circulation
Pressure Filtration
Fluid Filters; Membranes
Water Conservation
Pollution Control
Stationary Sources
Energy Conservation
Hyperfiltration
Ultrafiltration
Chemical Conservation
13B
13H
14B
07A
13K;11G
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport}
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
172
------- |