v>EPA
United States Industrial Environmental Research EPA-600/7-80-047
Environmental Protection Laboratory March 1980
Agency Research Triangle Park NC 27711
Feasibility of Recovering
Useful Salts from
Irrigation Wastewater
Concentrates Produced
by Power Plant Cooling
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-600/7-80-047
March 1980
Feasibility of Recovering Useful
Salts from Irrigation Wastewater
Concentrates Produced by
Power Plant Cooling
by
Hugo H. Sephton
The University of California
Sea Water Conversion Laboratory
47th and Huffman Boulevard
Richmond, California 94804
Grant No. R804760
Program Element No. INE827
EPA Project Officer: Theodore G. Brna
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
U.S. Environmental Protection Agency
Region V, Library
230 South Dearborn Street
Chicago, Illinois 60604
-------
ACKNOWLEDGEMENTS
Mt-h TfCrIniCrCSt ^r su99estions Jn this work by Mr. Fred Roberts and Mr.
Irt !i S\ i ?T-' °u™er EPA Pr°jeCt Officers' have been appreciated and
were most helpful ,n this work. The technical assistance of Carl L Freel
and secretarial help by Judy A. Sindicic of this Laboratory have been
roml^r i6' Tnis.reP°^t was ^proved by incorporation of many helpful
comments and suggestions from Mr. Theodore G. Brna, the final Project
ii
UtS. Environmental Protection
-------
Contents
Acknowledgements **
Figures ™
Tables 1V
1 Introduction L
2 Conclusions 6
3 Recommendations
4 Introduction: Feasibility of Sodium Sulfate
Recovery from Cooling Tower Blowdown 5
Process and pilot plant design for field test
operation ]~
Field test operation at Firebaugh 12
Comparative cost estimates 15
Conclusions from the pilot plant tests 16
5 Sodium Sulfate Recovery by Bench-scale Pro-
cedures 18
(a) Recovery of sodium sulfate and its deca-
hydrate from final regenerant effluent 18
(b) Recovery of Na2SO4- 10H2O from cooling
tower blowdown 19
(c) Recovery of calcium sulfate 20
6 Sodium Sulfate Recovery by Pilot Plant Procedures 25
(a) Design, construction and testing of an
experimental evaporator-crystallizer with
internal product separation 25
(b) Adaptation of a vertical tube evaporator
operated by vapor compression for sodium
sulfate recovery, with interface enhance-
ment 28
(c) Vapor compression, VTFE pilot plant used
in this study 29
(d) Crystallization of sodium sulfate by conven-
tional VC-VTE and by the foamy-flow
enhanced VC-VTFE procedure 31
References 36
iii
-------
Figures
1 Flow diagram: proposed power plant coolant cycle 7
2 Pilot plant assembly for field tests on use of agricul-
tural wastewater for power plant cooling 11
3 Field operation of UC-DWR pilot plant, 11/77-2/78 14
4 Sodium sulfate decahydrate crystals on bottom of beaker 21
5 Sodium sulfate decahydrate crystals on paper 22
6 Sodium sulfate decahydrate powder form of crystals
after efflorescence 23
7 Evaporator-crystallizer for sodium sulfate recovery 26
8 Vapor compression vertical tube foam evaporator
(downflow) 30
9 Evaporation-crystallization of sodium sulfate 35
Tables
1 Agricultural Wastewater Components During Process Cycle 6
2 Solids Recovered from 1000 ML of Regenerant Effluent 19
3 Crystallization of Na2SO4 by VC-VTE with Slurry-feed
Recycle 33
4 Crystallization of Na2SO4 by VC-VTFE Slurry-feed Recycle 34
iv
-------
SECTION 1
INTRODUCTION
The objective of this work was to determine the feasibility of recover-
ing useful sodium sulfate from two types of irrigation drainage water con-
centrates. These concentrates were obtained during a co-incident field test
study conducted by the author-and others in the San Joaquin Valley, sponsored
by the California State Department of Water Resources (DWR) and power utility
entities. That study utilized an experimental cooling tower cycle to assess
the feasibility of using irrigation drainage water as coolant; it was coupled
with an ion exchange (IX) softening pretreatment of the drainage water feed;
the spent ion exchange resin was regenerated with cooling tower blowdown
concentrates only. This experimental facility, field test operated in the San
Joaquin Valley, provided the concentrated brines used in the study reported
here. The results of that associated study has been published elsewhere (1,
2,3). The concentrates from previously softened agricultural wastewater (AW)
brines used in this study were the cooling tower blowdown and the final
effluent from IX regeneration.
The motivation for this work derived from the promising new procedure
permitting the use of irrigation drainage water as a power plant coolant. In
that procedure, this wastewater was softened by ion exchange removal of
calcium and magnesium rather than by conventional lime and soda addition, to
provide brines relatively rich in sodium sulfate. The spent resin was
subsequently regenerated solely with the concentrated, softened blowdown^brine
obtained from the cooling tower during the regeneration step; solid calcium
sulfate was precipitated from the concentrated brine. This procedure was
demonstrated in a field test series with a pilot plant that provided an
opportunity for recovery of sodium sulfate from the two alternative concen-
trated brines produced in the coolant cycle. Since this wastewater had an
initial stoichiometric sulfate content much higher than its chloride content
that was also in excess of the combined calcium and magnesium ions, the
preferred softening procedure was by cation exchange, reducing both Ca and Mg
to about 10 parts per million. The softened wastewater concentrate therefore
afforded an opportunity for crystallizing sodium sulfate without concurrently
crystallizing magnesium sulfate or sodium chloride. In fact, the removal of
some of the sulfate anion appeared advisable to minimize the precipitation of
calcium sulfate during the subsequent regeneration of the spent resin with
the cooling tower blowdown concentrate. The alternative concentrate from
which recovery of sodium sulfate also appeared feasible was the final effluent
obtained after regeneration of the spent resin, this effluent would be satur-
ated with calcium sulfate, high in magnesium content and an apparently more
difficult source for sodium sulfate recovery.
-------
r,,r,Jr ^lu • ^ro* the possFb1e commercial value of the sodium sulfate
stnoircost fl^ SPr°fed StUdY alS° re'ated t0 the most significant
single cost factor in the proposed use of irrigation drainage water as a
power plant coolant in the San Joaquin Valley, I.e., the cos? o? d sposal of
the fmal concentrated effluent, for instance, by ponding. By removing part
of its sod, urn sulfate as crystalline salt, the final effluent could be
sign f, cant ly reduced in both its volume and its final solids residue
the ?ze9andn cos? of" S^T'0" T*' AS * Conse"uence> this would reduce
the size and cost of the solar pond proportionately. In fact the cost of
comblneYc:; I f^ C°n«ntrated blowd°- was shown to be higher tn^ the
combined cost of ,on exchange softening the original drainage water feed the
the're l^erT^n l^Z"""9 "?' b]°^m ^ Vertical ^ evaporation 'ad
tne regeneration of the ion exchange resin (3).
This study focused primarily upon the feasibility of sodium sulfate
recovery by both conventional evaporation-crystallization and by the
interface-enhanced, vertical tube foam evaporation technique (4), a novel
iZf f Y C9Pita' and energy costs may be reduced? This procedure
ncr^L TY> tW°-phaSe (vapor-liquid) flow upon the evaporating liquid to
increase , ts evaporat.on rate (5,6) or the evaporative heat transfer
tTo iro^S' Calclumksulfa,te Precipitated during the ion exchange regenera-
tion procedure was a by-product from the above cooling cycle, and could be
Wh"e ?tS
allevi^e disposal requirements
r recover ' "9 "seful sodium sulfate and calcium sulfate from
™IJM|9e "^ had ™* previously been reported by a similar
hnn < .f.su"essfu ' this procedure could reduce by about one third
(about $2 million for a 1000 MWe power plant), the cost of ponding the
final cooling tower blowdown concentrate after its use for ion exchange
regenerat.on In add, t, on the potential economic value of the recovered
products could provide sufficient incentive for adopting this technology.
-------
SECTION 2
CONCLUSIONS
Bench scale and pilot plant studies conducted on two brine concentrates
obtained as the process effluents during field tests on agricultural waste-
water for power plant cooling have shown that useful sodium sulfate and
calcium sulfate can be recovered from these effluents prior to their disposal
by ponding. Recovery of these salts for exportation from a power plant site
would reduce the cost of ponding significantly.
Recovery of sodium sulfate, the main objective of this work, was shown to
be feasible by evaporation-crystallization utilizing a novel, energy-
conserving method. Using this method, foamy vapor-liquid flow imposed upon
the evaporating brine-crystal slurry provided enhancement of heat transfer
performance up to ^0 percent and concurrent energy reductions ranging to 28
percent, compared with conventional, non-foamy evaporation-crystallization of
sodium sulfate, in a 5000 gpd vapor compression vertical tube evaporation
pilot plant.
The overall conclusions drawn from this and related work are that the use
of agricultural wastewater for power plant cooling in the San Joaquin Valley,
utilizing ion exchange for pre-softening and its regeneration solely with
concentrated brine generated in the cooling process, is feasible and
economically attractive. Further, the recovery of sodium sulfate and calcium
sulfate from the cooling system effluents prior to their disposal by solar
ponding is feasible and provides significant potential advantages. These
advantages include a reduction in the size of the solar pond and the market
value of the recovered salts.
-------
SECTION 3
RECOMMENDATIONS
Larger scale pilot plant field tests of the recovery of sodium sulfate
from the spent ion exchange regenerant obtained as the final effluent in a
process system utilizing agricultural wastewater as a power plant coolant
should be made. Such tests should provide a sufficient data base for a
realistic assessment of both the technical feasibility and the environmental
and economic benefits to be anticipated.
The recovery of calcium sulfate, precipitated in an apparently homo-
geneous form during the ion exchange regeneration and its removal by flotation
by the fluidized bed procedure, should be field tested on a larger scale to
provide a data base for assessing its commercial potential.
The apparently available markets for sodium sulfate, for instance in
paper making and for calcium sulfate, for instance as an agricultural soil
conditioner and for wallboard construction should be assessed with reference
to the above recommended larger scale tests.
-------
SECTION
INTRODUCTION: FEASIBILITY OF SODIUM SULFATE RECOVERY
FROM COOLING TOWER SLOWDOWN
The recovery of sodium sulfate from agricultural wastewater after its
concentration in a cooling tower was to be examined by two procedures: con-
ventional evaporation-crystallization and vertical tube foam evaporation
(VTFE)* Two different brine concentrates were to be examined: the tmai
effluent after ion exchange regeneration^ the cooling tower concentrate
before its use in ion exchange regeneration.
The total dissolved solids (IDS) of the agricultural wastewater (AW)
collected at the DWR test facility at the Firebaugh site varied seasonally,
typically from about 2,000 ppm during the summer to about 6500 ppm m the
winter (3). Thiis variation was primarily responsive to rainfall, which
peaks in summer, and irrigation practices which depended on the needs of
crops being cultivated. During the period of this work Cal i forma
experienced one of its worst droughts. The result of this was that the IDS
of the wastewater remained high (above 3500 ppm), and the si .ca content was
also higher than anticipated; this caused delays in the supply of brine
concentrates, occasioned by the need to develop and implement a silica
control system for the field tests of the companion study (1).
Table 1 represents the ionic composition of the wastewater and coolant
brines during the summer of 1977. Most of the present work was done during
the following winter when the wastewater was of approximately twice the
concentration (at 7000 ppm of TDS) with about the same solute proportions.
The composition of two brine sources used for recovery of sodium sulfate
Eas;cz\^^
DOWN" and "REGENERANT EFFLUENT", respectively. Of these sources the former
suggested assured sodium sulfate recovery, while recovery from the latter
was questionable on account of its substantial magnesium and calcium
contents as well as a lower sulfate-to-chloride ratio which mcreases the
menhood of contamination by co-crystallization of magnesium sulfate and
odium chloride. However, the concentrated blowdown would not be ava, able
without reservation as a source for sodium sulfate recovery since it also
had to serve as the regenerant for the ion exchange (IX) resin after further
concentration by vertical tube evaporation (VTE) as shown ,n column 5 of
* U.S. Patent Number 3,8^6,25^, November 5, 1971*.
5
-------
TABLE 1. AGRICULTURAL WASTEWATER COMPONENTS DURING PROCESS CYCLE
MAJOR ION CONCENTRATION, rng/1
COMPONENT
Ca+*
Mg**
Na+
S°4
Cl~
HCO "
sio2
TDS
AGRICULTURAL
WASTEWATER
346
156
480
1,663
307
528
50
3,530
SOFTENED
7
6
.1,190
1,663
307
528
50
3,751
COOLING TOWER
SLOWDOWN
105
90
17,850
24,945
4,605
7,920
50*
55,565
VTE
CONCENTRATE
150
120
23,800
33,260
6,140
10,560
50
74 , 080
REGENERANT
EFFLUENT
150*
3,120
9,600
19,560*
6,140
10,560
50
49 , 1 80
*Ca2S04 precipitates during regeneration; *Si02 content was controlled.
Table 1. The sodium sulfate content of the blowdown was nearly twice the
mimmal or stoichiometric requirement as indicated by comparing the Na+ and
SOii ionic levels in columns 5 and 6 before and after IX regeneration. Part
of this excess in sodium sulfate could therefore be recovered, provided that
it did not reduce the sodium content to a level that would seriously deplete
the potential for displacing calcium and magnesium from the spent IX resin
during the regeneration procedure (1.2). This procedure would require the
recycle of about five volumes of concentrated coolant (column 5) followed by
one fresh volume of this coolant concentrate to achieve full IX regeneration.
Removal of some of this sodium sulfate would undoubtedly affect the regenera-
tion procedure used in the field test study and would require future adjust-
ments in that procedure.
A flow diagram showing the procedure proposed for the field test study
for power plant cooling with irrigation drainage water and also indicating
the two proposed withdrawal points of brine for the sodium sulfate recovery
of this study is shown in Figure 1.
The second or alternative source for sodium sulfate recovery would be
the final IX regenerant effluent having the composition shown in column 6 of
Table 1. This effluent would be saturated with calcium sulfate at the tem-
perature used in the regeneration cycle. Its calcium, magnesium, sodium,
chloride and bicarbonate contents would also be substantial, providing
sources of possible contaminants during the recovery of sodium sulfate by
crystallization. This more challenging of the two sources would be more
-------
COOLANT TO
CONDENSERS
ION EXCHANGERS
COOLANT MAKEUP
20 MGD
0.5% TOS
(14,000 gP"0
VERTICAL TUBE EVAPORATOR
UPFLOW OR OOWNFLOW
WITH INTERFACE ENHANCEMENT
ION EXCHANGERS
REGENERATION
FOAMING AGENT
RECYCLED TO VTE
RECOVERY OF
CALCIUM
AND SODIUM
SULFATES
FORCED
CIRCULATION
EVAPORATOR-
CRYSTALLIZER
FINAL BLOWOOWN
SLURRY OR SOLIDS
SODIUM SULPHATE
RECOVERY
I ______ 1
1
FINAL SLOWDOWN
FIGURE 1. FLOW DIAGRAM: PROPOSED POWER PLANT COOLANT CYCLE
-------
thP nr« 5 because (t wou d not adversely affect the IX regeneration step of
the procedure for power plant cooling by the field tested processes. If it
could be proven technically and economically feasible, the removal of useful
sod,urn sulfate from the final regenerant effluent would also reduce the
disposal cost of this effluent (i.e., the capital cost of lined sotar evapor-
for ?h!°r? f ^ntaining
-------
pilot plant scale in the San Joaquin Valley in the joint study with the
California Department of Water Resources (DWR), which was co-sponsored by
electric power utilities (3)•
Conventional methods of softening and silica control will suffice as pre-
treatment for irrigation drainage water, but they are not cheap, require
importation of large quantities of chemicals which add to the salt burden of
the final effluent, and are not necessarily the best way to pretreat this
wastewater. The use of lime and soda ash addition has been demonstrated to
be adequate for Wei 1 ton-Mohawk Canal water prior to its desalting by reverse
osmosis (RO) and for Palos Verde Drain water used in a pilot plant simulating
the proposed Sundesert power plant cooling system (9). The cost of this
conventional approach is substantial at about 45 cents per 1000 gallons, and
the imported chemicals add to the volume and cost of subsequent ponding of
blowdown and softening sludges. In the case of Sundesert, the anticipated
daily cost of pretreatment chemicals amounted to close to $5,000 per day,
while the anticipated capital cost of the solar evaporation ponds for the
effluents was about $17 million for the 2,000 MWe plant proposed (10).
The alternative method of softening by ion exchange (IX) coupled with
IX-regeneration solely with concentrated blowdown brine plus silica control
by alum dosing of a coolant side-stream provides both capital and operating
costs savings. It also permits the recovery of useful salt and reduces the
size and cost of ponds for effluents. This pretreatment approach may also be
applicable to desalting of AW with RO, and it could be used in seawater
desalination in conjunction with the vertical tube foam evaporation (VTFE)
technique to permit much higher performance ratios than obtainable before.
This approach would conserve energy and reduce the cost of pure water
produced.
Several methods of pretreatment and desalination developed at the
University of California Seawater Conversion Laboratory (SWCL) provided a
new and promising combintion for a power plant coolant process that could use
irrigation drainage water. These processes were: (l) the ion exchange
softening coupled with fluidized-bed regeneration and (2) the energy-
conserving foamy-flow evaporation of brine to be used as regenerant andjts
clarification by foam flotation. When combined with and serving a cooling
tower, this process combination appeared feasible and economically competitive
with the conventional approach. The proposal (11) incorporating these
combined processes was submitted to the DWR and favorably reviewed by a group
of prospective co-sponsors who were subsequently requested to join in the
study. Its scope of work was improved by suggestions made during several
discussions with representatives of the prospective sponsors, the DWR^and
the SWCL. The final project plan providing for the design, construction and
shakedown testing of a pilot plant facility at the SWCL followed by its field
test operation at the DWR site near Firebaugh in the San Joaquin Valley over
a total project period of 30 months and at a projected cost of about $600,000
was approved by all co-sponsors for start-up in July of 1975- The project
sponsors were the Pacific Gas and Electric Company, the Southern California
Edison Company, the Los Angeles Department of Water and Power, and the
Electric Power Research Institute in addition to the DWR, and to some extent
the University of California and the Water Resources Center.
-------
fact A favorable environment for this project was also contributed to by the
dpLnH M PJOJ"tlons of California's future fresh water demands exceeded
f!r«f 1Q7L 6Sh.water suPP^es (12) and by the California Waste Water Reuse
AUn nf ctr/eqU'n-?. wastewater be reused to the maximum extent (13).
Also of strong significance was an active interest and co-sponsorship bv the
Environmental Protection Agency (EPA) in developing these and other relatea
processes. Another factor was the constraints imposed by the EPA guidelines
^ooUng ^ersr9e FndUStr?al effluents that ^"courage "he increasing use of
of a]Th?n^rJT/S Planned uprov!ded for a <*eaningful and active co-operation
- «^.sir^^ SLns-Js.ir-sf^-
The work reported here was a separate but co-incident and related
"? ^ *•? !earbmtV of ^covering useful salt from the effluent of the
test pi ot plant operation and sponsored by the EPA, responsive to the
UU 15°*) Th ^A^ " C0nt'nuatj- °f t-o earlier EPA-spon'sored projects
i!; +A ' u ! EPA-sPonsored work reported here was closely related and
•ntegrated with the larger field test project but had separate objectives
Process and Pilot Plant Design for Field Test Operation
*t™, H facl1ity 'ncorporating the following process elements and
steps was designed constructed and shakedown tested at the SWCL, transported
H mn%han ?Tn, ^Iley ^ fleld teSted °n AW J"oint1y with ^e DWR over a
13-month period ending February 1978 (See Figure 2):
(a) Ion exchange (IX) softening of AW collected from an irrigated field
to remove ca cium and magnesium to residues of about 10 ppm each and replacing
these ions with sodium exchanged from the resin.
th* rib) EfTratinn °f ^ IX softened AW by a Permissible factor, within
the range of 5 to 20, m an experimental cooling tower under process
^!t f'0"^ T ar !? th°Se °f 3 POWer Plant' {S!1lca Sca1e developed as the
mam field test problem; an effective and economical method for silica
control by aluminum sulfate treatment of a small coolant side stream was
also developed and tested during the course of the project).
(c) Further concentration of the 5- to 20-fold concentrated coolant
blowdown by the energy-saving vertical tube foam evaporation (VTFE) process
W to a salt concentration level of 70,000 ppm, sufficient to serve as the
sole regenerant for the IX resins used in step (a) above. Clarification of
this brine concentrate by a foam flotation procedure (17) to remove both
particulates and the foaming agent additive used for the VTFE step noted in
{c) was a beneficial and effective procedure.
(d) Regeneration of the IX resin with the brine concentrate using a
novel fluid.zed-bed procedure that simultaneously removes the calcium sulfate
10
-------
•;,./'1:.;':-;:•.;•%: 1 i. SON EXCHANGE COLUMN
COOLING TOWER
3. VERTICAL TUBE EVAPORATOR
FIGURE 2. PILOT PLANT ASSEMBLY FOR FIELD TESTS ON USE OF AGRICULTURAL WASTEWATER FOR POWER PLANT COOLING.
-------
precipitated during the regeneration process by hydraulic classification (1,2).
(e) Evaporation-crystallization of the concentrated blowdown from this
vnh^r fTeT SV Pr°V?de S°dlUm SUlfate as a bYPr°d^t and reduce the
volume of final effluent to be disposed by ponding.
The integration of process steps (a) to (e) are shown in the flow diagram
nllnJ9nE 1 w'th.th"VndivIdual steps enclosed in broken lined blocks. The
pilot plant, designed for continuous operation at the field test site is
shown m Figure 2 as assembled at the UC-SWCL during Its shakedown testing
hFah^J ^ ^V cy''"drical 'X softening column of UHn. diameter, 8-ft
high made of Plexiglass and containing a 3-cubic foot layer of polystyrene-
diviny benzene sulfonate resin in the sodium form (regenerated form), and
several large vessels for AW feed storage, regenerant storage and recycle.
of sof^n H Iu " °i theTUrl9h1: in Fi9ure 2' could P^duce about 1000 gallons
of softened AW per cycle. The custom-made cooling tower, about 30-feet high
is shown on the left of Figure 2. It had a packed section of 20-ft. height and
2-square foot in cross section, similar to a vertical section from the packed
core of a large cooling tower. Induced upward air flow and downward flow of
coolant through this core were similar to that of a large power plant cooling
tower. 3
Provision was made for controlling the salinity of the coolant within the
range of 55,000 and 70,000 ppm of TDS and within a ±1000 ppm fluctuation by
"I!,-!.'"9 * flow"tnrou9h salinity sensor controlling the periodic, automatic
addition of fresh coolant to the cycle. Blowdown from this cycle was by over-
How from the cooling tower basin controlled at either 55,000 or 70,000 ppm of
Jhe heat input and rejection rate of the cooling tower was 40,000 watts
Heating of the recycled coolant by 20° to 30°F was by passing it at a velocity
of about_7 feet per second through 80-foot long copper-nickel condenser tubes
heated with recirculated hot water. A 0.8% sidestream of this heated coolant
now of 10 gallons per minute was diverted through the silica control system
where it was dosed with 150 ppm of aluminum sulfate, and the precipitated floe
containing silica and magnesium was removed by sedimentation and filtration
mmedlately to the right of the cooling tower in Figure 2, are the boiler
1100 horsepower) and vertical tube evaporator (5,000 gpd) used to further
concentrate cooling tower blowdown to the 70,000 ppm of TDS, as needed for
regenerating the IX resin bed and to close the cycle. This pilot plant desiqn
was detailed elsewhere (3).
The crucial link in this process cycle was the IX resin regeneration with
blowdown brine concentrates only. This procedure eliminates the need of
softening chemicals and also provides for the possible recovery and sale of
potentially useful gypsum and sodium sulfate. Sale of these products would
reduce the cost of disposal of effluents, for instance by reducing the size
of the pond required.
Field Test Operation at Firebaugh
Field operation for a 13-month period of the pilot plant in the San
12
-------
Joaquin Valley by DWR staff was started in January 1977 with assistance^ rom
the staff of the SWCL. These tests were divided into five phases. During
The final phase of 3 months duration, all systems were operated in their
typical or optimized modes as determined or developed during the preceding
four phases. Tests and observations made durig the final (fifth) phase are
shown in Figure 3-
The main objectives of this project were achieved during its final phase.
These objectives comprised: (l) softening of irrigation drainage water (AW)
by conventional ion exchange, (2) using the softened water for cooling^.n a
process that closely simulated power plant cooling, and (3) concentrating
this coolant by VTE to a sufficient level for it to be used as the sole
regenerant of the ion exchange softening resin. In addition, a process was
developed to solve silica scaling, the main unanticipated problem that
surfaced during the project. The sidestream silica control system (SCS) _
proved satisfactory on the basis of each of the following criteria: sustained
heat transfer performance, reduction of silica to a level well below its
scaling or fouling limits, maintainance of low silica levels at an acceptable
reagent cost, and maintainance of alum reagent residual to a level well below
its fouling or scaling potential. Plot A in Figure 3 shows that the heat
transfer performance of the condenser-simulator did not deviate from its
"clean-tube" performance level throughout this period (except for the normal
experimental fluctuation of about ±3-5*). In this regard, the best perfor-
mance data obtained during two earlier periods within the field development
phase of the SCS were sustained on an apparently indefinite basis. All other
data confirmed this observation; for instance, the silica concentration in
the main coolant system remained generally below 50 ppm as shown in plot B,
while the scaling/fouling limit had earlier been established as being about
three times higher. The Si02 level in the SCS outflow was readily maintained
below 30 ppm and was as low as 5 ppm (see Plot C). The concentration of _
aluminum in both the main coolant and the SCS sidestream was in the vicinity
of 0.4 ppm, well below the 1 ppm limit generally considered as acceptable.
Confirmation of the effectiveness and adequacy of this SCS was sought by
the project coordinating committee through a detailed on-site study by an
expert in the field. The consultant supported the above findings for the SCS.
He concluded that this method was preferable to the alternative one of using
ferric sulfate and his final report of January 27, 1978 also supported the
early unit cost estimates for the SCS operation presented to the Committee.
This report also supported our findings that floccultion and its settling
were favored by the PH of 9 and the TDS of 55,000 ppm of this coolant as
field tested. Similarly, he mentioned that the SCS operated well since
November 23, 1977, that dissolved aluminum was low, and that our use of an
annular filter with adequate intermittent cleaning should suffice to pre-
vent occasional leakage of floe. He also acknowledged that our high TDS
brine system was unique in his experience and that our SCS should operate
more satisfactorily in a large-scale application than in the pilot plant
field tests.
During the final test phase, a series of samples were collected for
monitoring dissolved aluminum and silica in appropriate process streams. The
data and analytical results obtained by DWR staff at Firebaugh, by the
13
-------
ISO-i 500
££
es~
5 It • 40
1st
sil
£ J to,ooo
§-
£ ^ 40,000-
a
5 M.OOO-
HON-TTPICAL OPERATION
NON-TYPICAL OPERftTlON WITH LOW TDS
COOLANT, ftND WITHOUT SILICA CONTROL
SYSTEM. THE TDS LEVEL WAS CONTROLLED
TO MAINTAIN SiO? AT 130 MG-L
PERIOD FOR REPLACING DISCARDED
BRINE BY GRADUAL ACCUMULATION
OF CONCENTRATE IN BASIN I. SILICA
CONTROL SYSTEM IN NON-TYPICAL USE
DUE TO LOW TDS BRINE AND INADEQUATE
BUFFERING CAPACITY
TYplcftL OPERAT10H
TYPICAL OPERATION
N0 TIJBE CLEANING
ID 20 JO 10 20 30
SEPTEMBER OCTOBER
20 JO
30 JO
500 riso
D
80,000
60,000
40.000 oL
-zo.ooo
20 JO
PLOT A: CONDENSER HEAT TRANSFER PERFORMANCE
PLOT B: COOLANT Si02 CONTENT
PLOT C: Si02 CONTENT OF SCS OUTFLOW
PLOT D: COOLANT SALINITY, TDS
FIGURE 3. FIELD OPERATION OF UC-DWR PILOT PLANT
NOVEMBER 1977 - FEBRUARY 1978
-------
DWR-Bryte Laboratory, and by the consultant's laboratory showed sufficient
agreement to support the following conclusions:
(1) The silica control system (SCS) was adequate in maintaining the
silica concentrations at levels well below the scaling threshold (1 ppm) .
(2) Levels of dissolved and total aluminum in the effluent from the
silica control system, as well as in the recycled coolant, were maintained at
levels well below the suggested maximum level of 1 ppm.
(3) The sidestream SCS was adequate for maintaining high heat transfer
performance over about 1 k weeks with a recycled softened agricultural waste-
water coolant concentration ranging from 58,000 to 70,000 ppm and without any
cleaning of the heat exchanger tubes.
Of promising significance was the finding that the cooling tower could
be operated satisfactorily on a 15- to 20-fold concentrated softened AW
confirming the early data obtained during the pilot p ant development tests at
the SWCL (1). As shown in Figure 3 the coolant TDS (Plot D) was maintained
at about 58,000 ppm until about December 28, 1977, and then increased to_
70 000 ppm and maintained at this level until the end of the project period.
Consequently, the VTE could be bypassed, and the cooling twer blowdown was
henceforth used directly as ion exchanger regenerant The capital, operation
and maintenance costs of the VTE step could thus be el immated .from the
essential process sequence. Alternatively, the VTE. fac.l i ty could be
designed for the additional service of recovering useful, saleable salt,
such as sodium sulfate from the regenerant blowdown, and used for the latter
purpose when not required for cooling tower blowdown concentration.
The alum and brine feed flow rates through the SCS were reduced for the
last month of operation, February 1978. The heat transfer performance, and
the silica and aluminum levels remained entirely acceptable, md.catmg that
0 83 percent sidestream should be adequate. The cost of alum reagent was
reduced to a level of about $200 per day for a 1000 MWe plant. This cost_
might be further reduced when the SCS is optimized with a Si02 concentration
of about 100 ppm (still below the scaling threshold) in the coolant and in
the cooling tower blowdown rather than the 50 ppm limit of these tests.
The VTE was operated satisfactorily, both with the VTFE mode (15 ppm of
surfactant added) and the VTE mode (without surfactant addition to the feedj ,
and the heat transfer performance data confirmed earlier data obtained at
the SWCL. Overall heat transfer coefficients of about 1 800 and 1200 Btu per
hr-ft2_°F were obtained with the VTFE and VTE modes (at 160 F) , respectively.
The surfactant additive was successfully stripped from the concentrated VTFE
brine by foam f ractionation, and the clarified concentrate (70,000 ppm of TDSj
was used for ion exchange regeneration (1,2,3).
Comparative Cost Estimates
Cost estimates were obtained as part of this study by an independent
engineering firm during the first phase of the f:rld test series indicated
15
a
-------
e
amounted to $3,8,000 in 1976 dollars.
s-s: ss-
be deductab fromCrenKIOna S?tenin9 sVstem ("me and soda ash) would
system 6 C9pital C°St °f $3'85/t million for this cooling
Conclusions Drawn From The Pilot Plant Tests
The following conclusions and recommendations derived from the field
tests conducted in the San Joaquin Valley:
1. Softening of agricultural wastewater by ion exchange and its
ra°n 6ntirely with Concentrated softened AW of 70,000 ppm of TDS
2. The_ necessary concentration to provide regenerant brine (70,000 ppm)
can be done .n either a vertical tube evaporator (VTE) or in the cooling
tower, thus eliminating the need for blowdown concentration by VTE if this
16
-------
high a coolant IDS is acceptable from the standpoint of corrosion and cooling
tower drift emission.
3. The silica content of the AW supply at Firebaugh was higher than
anticipated (up to 50 ppm) during the test period. The silica control
system, based on a 0.8% sidestream treated with aluminum sulfate, developed
in this project for this purpose was satisfactory and effective; it main-
tained silica and aluminum levels well below their scaling limits and enabled
a continuous 3-month period of operation without any significant degradation
in the performance of the heat exchanger, which maintained an overallQheat
transfer coefficient of 3378 W per m2-°C (595, ± 3-53 Btu per hr-ft2- F).
The cost of chemicals for silica control appear reasonable at about $200
per day for a 1000-MWe plant. This cost could be reduced by lower AW silica
levels and by permitting a higher Si02 content in the recycled coolant and
b1owdown.
k. The 5000-gpd vertical tube evaporator performed satisfactorily both
in terms of heat transfer capability and as a tool for blowdown concentration,
and for distilled water production. The cost of this operation would be
substantially reduced if waste heat is utilized (i.e., if turbine exhaust
steam is used for brine evaporation rather than prime steam) an alternative
that is presently being developed under EPA-cosponsorship by the author under
a separate project.
5. The foam fractionator performed satisfactorily for stripping
surfactant additives and particulate matter from the VTE concentrate prior
to its use for ion exchange regeneration.
6. It was shown that the capital cost of solar evaporation ponds
required for the disposal of the final effluent from this process sequence
was almost twice the capital cost of the ion exchange softening -and silica
control systems.
7. The capital cost of the effluent disposal ponds could be reduced by
about one third if sodium sulfate is recovered from this effluent before
ponding it.
8. Further field tests to be conducted on a larger scale, and improve-
ments to permit automated operation of the test facilities, would be
advisable before this process is applied for cooling a commercial power plant.
17
-------
SECTION 5
SODIUM SULFATE RECOVERY BY BENCH-SCALE PROCEDURES
(a) Recovery of Sodium Sulfate and Its Decahydrate From Final Regenerant
Effluent.
The clear supernatant brine remaining after regeneration of the cation
exchange resin and in equilibrium with precipitated calcium sulfate after
stirring this suspension for 1 hour to promote the growth and settling out of
larger gypsum crystals at the expense of smaller ones (1,2,3) was used in
these experiments. A series of experiments was conducted in which the clear
regenerant effluent having 50,000 ppm of TDS (see Table 1) was further con-
centrated to residual brine concentrations of about 100,000, 150,000 and
200,000 ppm (in 1-liter batches) on a rotary evaporator (Bu'chi, Rotavapor)
under high vacuum, heated with a water bath 45°C (110°F). The first two of
these residues remained clear, while considerable cloudiness was noticeable at
the highest concentration. This was due to precipitation of a white powdery
substance that would not entirely re-dissolve readily upon reheating to 95°C.
These residual brines were subsequently cooled in open glass beakers
placed in the refrigerator. Clear chunky needle-shape crystals started
growing from the bottoms of the beakers containing these concentrates, usually
within about two hours and when the brine temperatures were approaching about
4 C (39 F). Crystallization was permitted to gc overnight, when the brine
temperature was reduced to slightly below 0°C. The brine of the two-fold
concentrate remained clear except for crystal and ice formation. The three-
fold concentrate deposited a few clouds of fine white, gelatinous-1ike
material in suspension in addition to a preponderance of the large chunky
crystalline bars, while the four-fold concentrate had a considerable amount of
the suspended white material in addition to the crystalline bars.
The crystals formed overnight upon refrigeration were removed by filtra-
tion on a Buchner funnel, followed immediately by washing three times in rapid
succession with ice-cold disti1 led water, and the excess of water removed by
suction through the funnel. The residues were partially dried (to a damp
condition) by tumbling over filter paper and weighed. The crystalline
products were then left exposed to the atmosphere to remove water of crystal-
lization and collapsed into fine white powdery residues which were weighed.
Table 2 shows typical results obtained.
18
-------
TABLE 2. SOLIDS RECOVERED FROM 1000 ML OF REGENERANT EFFLUENT
Concentration Factor Hvdrated Crystals Anhydride Powder
(TDS in ppm)Weight, g. Weight, g.
2 (100,000)
3 (150,000)
k (200,000)
5-7
67.0
69-0
2.5
27.5
28.0
It was apparent that the weights of the main hydrated crystall me product
obtained and its final powdery anhydride were consistent w, ^ the expected
formation of crystalline sodium sulfate-decahydrate (Na2SO,. 10H20 "Jich
readily loses 10 moles of water on exposure to the atmosphere (efflorescence)
and collapses into the powdery, white anhydride (Ha2^k> •
Further confirmation of the identity of the main ^stalllne product was
obtained by heating these hydrated crystals in an erlenmeyer flask on a steam
bath and thus recrystal 1 izing the anhydrous form (Na2SO,) from the water of
hydration of the decahydrate (Na.SO,- 10H20) at temperatures above 83 C ( 81 F) .
In other experiments, the cooled, three-fold concentrated regenerant effluent
when seeded with Na2SO,- 10H20 would rapidly produce fine needles of the
crystalline decahydrate.
Analytical data showed that the composition of the powdery white product
was 99-2 percent Na2S04.
Based on these experiments, approximately 40 percent of the total salt
content present in the original irrigation wastewater could be recovered as
either Na2S0^1oi20 or as Its anhydride from the final IX regenerant effluent
ob at ed from this experimental power plant coolant system Th, * recovery
assumes an approximately three-fold concentrat.on of th.s effluent, its
temperature reduction to about 0°C to form the crystalline ^"hydrate
(mirabilite or Glauber's salt), or production of the anhvdr.de (thenard.te)
by further evaporation-crystallization above 83 C (181 F) .
(b) Recovery of Na9SOu-10HoO From Cooling Tower Slowdown
The cooling tower operated at the Firebaugh site was initally operated
°
s
concentration with vertical tube evaporation (VTE) .
Sodium sulfatewas readily recoverable from this sof Jened concentrate
This is not surprising, considering that its composition (about as shown ,n
Table 1 column 5) was relatively low in calcium, magnesium, and chlor.de
dium and sulfate. It was decided that recovery of no
with reference to sodium and sulfate.
19
-------
-• »
temperature of 45 C. The clear concentrate was poured into a 1-liter
28°F) " hu' L nlndfherf r?96rat?r °Vernl9ht ?°r Sl°W cooling toearbo
U» FJ. Chunky, needle-like crystals measuring about i-inch x 2 inches
--
e crystas were drained of water, removed from the beake
tn thenae r°and th"' °' Kf?iter R3Pr by tUmbUn9' Th; Product' a'^ ^dr
Thlfm ^ • ? then we.ghed on a sheet of paper, is shown in Figure 5 (83 0 g)
lowd ?'!a ,Wa- al1uWe2 t0 S'°Wly efflor^ce (to the open air) to a whi e
powder still having the form of the above crystals (see Figure 6) After
oercln Vr?r' r^^ ^^ ^^^ 3^3 9 W9S anal^ed as 100.1?
percent NagSO^, These data were consistent with the crystallization of
' a°^10H2?) t»«t effloresced to the anhydride, th^ardi^e with
qulfsli «h«, S k from these bench-scale experiments that useful sodium
sulfate should be recoverable from either the final wastewater effluent, after
IX regeneration in a yield representing about 40 percent of the total salt
content of the cooling tower blowdown, or directly from this blowdown before
using it for IX regeneration, in a maximal 50 percent'yield.
(c) Recovery of Calcium Sulfate
The recovery of calcium sulfate was evaluated during this EPA-sponsored
part of the overall program for using softened irrigation drainage water as a
power plant coolant. Its formation during the regeneration of the IX resin
and its separation by hydraulic classification by a fluidized bed regeneration
procedure was part of the DWR field test series conducted with the pilot plant
in the San Joaqum Valley (1,2,3). A comparison of the ionic contents of
columns 5 and 6 of Table 1 indicates that approximately 14,200 mg/1 of sodium
20
-------
FIGURE 4. SODIUM SULFATE DECAHYDRATE CRYSTALS ON BOTTOM OF BEAKER
21
-------
.•.-" - >,.-
715
N)
NJ
FIGURE 5- SODIUM SULFATE DECAHYDRATE CRYSTALS ON PAPER
-------
- *JvV.
fc^^-llfe^j^af.
',;--«:;V- -''"',-:' v"-v;"-,';.,;:;V'--';>'^!;" • | '
FIGURE 6. SODIUM SULFATE ANHYDRIDE POWDER FORM OF CRYSTALS AFTER EFFLORESCENCE
-------
thIT
geneous
usnq
useu as
durlna IX
tion on a
"°
mil,
com0
composi
'
'™ *
°f ^^ Pr«''P'tated fro^oc'esTbrl e during
G^sum was Pr°d^ed in an apparently homo- "
*V °f SYPsum to be anticipated from a 1000-MWe power
9pd of wastewater coolant with the cooling system
'^be a^0^^ V 77 tons per day when lie
shown in Table. 1. The potential for using this
in this work' but gypsum might conceivably be
°" ^ "*" bOard -"anufacture. Gypsum produced
h H
percent'
*U?Pended In
Thls resid^
hot taP water and removed by flltra-
was washed well with hot water
-------
SECTION 6
SODIUM SULFATE RECOVERY BY PILOT PLANT PROCEDURES
Two evaporator-crystallizers were constructed and tested in this work:
(1) a small and versatile pilot plant utilizing forced re-circulation of
brine through a heater followed by flashdown and separation of the crystalline
sodium sulfate produced into a quiet zone and (2) adaptation of a 20-cubic
meter per day vapor compression vertical tube evaporator for a slurry-feed
recycle of brine and crystalline product and crystallization of anhydrous
sodium sulfate into the slurry and its subsequent separation by centrifugation.
(a) Design. Coptrurtion and Testing of an Experimental Evaporator-Crystal 1izer
with Internal Product Separation
The pilot plant facility, shown in Figure 7, was modeled after the Oslo
type of evaporator-crystallizer. It could be operated at low or high tempera-
tures to crystallize sodium sulfate as its decahydrate, to simulate the use of
waste heat for this purpose (by evaporation at temperatures below 50 C (122 F).
It also allowed recrystal1ization of the decahydrate by cooling alone. It
was further used to crystallize the anhydride at evaporation temperatures above
83°C (l8l.if°F).
This pilot evaporator-crystallizer was made of pyrex glass, brass pipe
sections and appropriate fittings. It utilized electrical immersion heating
with two stainless-steel-sheathed heaters of 3000 W each mounted in a 60-cm
(24-in.) long, 7-5 cm (3~in.) diameter Pyrex pipe through which the brine was
pumped upward for recycling with a centrifugal pump (stainless steel) while
being maintained under sufficient pressure to prevent boiling.^ The heated
brine was then released through a control valve for flashdown into a
horizontal Pyrex tube of 120-cm length and a 2.5-cm diameter which was
tangentially attached near the top of a separator vessel made of brass and
Pyrex pipe sections, 15-cm diameter and 120-cm long, mounted vertically. The
top of this separation vessel was attached to a condenser via a short pipe
elbow of 5-cm diameter. The body of this vessel was divided into an upper
cyclone separation section for disengaging the liquid phase from the flashed
vapor and a lower quiet zone for settling out the crystalline product formed
by evaporation of the recycled and re-heated brine. This was accomplished by
providing a funnel-shaped down-spout to deliver the brine to a central point
about 30 cm down into the quiet zone which was about 75 cm in vertical
dimension and separated from the cyclone zone by a brass plase. Recycled brine
was withdrawn from a level just below this plate. Provision was made for
withdrawing the solid phase product (crystals) through a large valve attached
25
-------
FIGURE 7. EVAPORATOR-CRYSTALLIZER FOR SODIUM SULFATE RECOVERY
26
-------
to the conical (domed) bottom section of the lower Pyrex vessel.
This pilot plant was operated under vacuum (oil pump, 29.5~in. HgJ , and
fresh feed liquid was continuously introduced through a valve connected to
the preheated brine line immediately upstream of the horizontal flashdown pipe.
The condenser was cooled with either tap water passed through a copper pipe
coil contained in the horizontal tube condenser or by recycling chilled water
through the coil from a refrigerated container. A commercial refrigeration
unit having a freon-evaporation coil made of stainless steel connected to an
automatic freon compressor and heat rejection unit, shown in Figure 7 (Van
Waters & Rogers) was used.
This evaporator-crystal 1izer unit was used for the recovery and
recrystall ization of Na2S(V 10H20 as well as Na2SOit. In one typical applica-
tion, the evaporator was evacuated of air and cooling tower blowdown trans-
ported from the San Joaquin Valley field test site in large polyethylene^
bottles (see column 5 of Table 1) was introduced under vacuum to fill this
apparatus to the level of the cyclone separator. The brine was then recir-
culated over the electrical immersion heaters, and cold tap water was passed
through the condenser. Flashdown of the recycled preheated feed brine usually
reached a steady-state evaporation rate at a temperature level of about kO C
(104°F). Evaporation was maintained at this level with the volume being
maintained by controlling the inflow rate of fresh brine feed under vacuum and
by collecting distillate at an approximately equal flow rate in an evacuated
reservoir connected to the condenser.
After about a three-fold concentration level of the recycled brine was
reached, crystallization would initiate in the cyclone and crystals would
settle in the conical base area of the quiet zone. Recycled brine was drawn
off from the top of this zone and returned to the heater by the centrifugal
pump. This process could proceed indefinitely with steady flow rates of
distillate, crystalline product and of mother liquor sufficient to maintain
about 50 percent of the original salts in the brine (about 11J,000 ppm of TDS
at the three-fold concentration level). In these tests, however, the needed
control instruments for continuous operation were not available, and a batch-
wise procedure was adopted.
The crystalline product was removed from the mother liquor by centrifu-
gation. A sample of this product was allowed to effloresce to the anhydride
and when analysed showed 97-3 percent Na2SOif.
Recrystal1ization was done by re-dissolving the decahydrate in a small
volume of water at about 40°C with continuous stirring. The apparatus used
was an open cylindrical vessel (made by silver-soldering a brass plate to one
end of a 50-cm long brass pipe of 10-cm diameter), placed in a similar vessel
(made from a 15-cm diameter' brass pipe) through which hot water was circulated.
The decahydrate was recrystal1ized by recirculating chilled brine through the
outer cylindrical vessel and the Plexiglass container in which the cooling coil
of the refrigeration unit was submerged as shown in Figure 7. The crystalline
Na2SOjt-10H20 was recovered by centrifuging and allowed to effloresce to the
anhydride by exposing it to the air on a sheet of filter paper. This product
(oven-dried) analysed to 100.2 percent Na2SOtt.
27
-------
In an alternative method of operating this pilot plant, the recrvstal1i7a-
tion of the decahydrate to produce the anhydride form Si rec^ly was as foHows:
0) The evaporator was evacuated of air and a concentrated solution of the
decahydrate ,n distilled water was drawn in under vacuum to the level of the
cyclone separator.
(2) The feed rec?rculation pump was started and the immersion heaters were
adjusted to heat the feed to about 85°C. nepers were
(3) Cold F; during the rest of the evaporation-crystallization procedure.
(4) The anhydride form of Na2SO,, settled in the base of the separation vessel
and was recovered through the drain valve and spun free of water in the basket
type centrifuge. This finely crystalline material was analyzed and showed
y^'^j percent ^3250^.
(b) Adaptation of^a Vertical Tube Fv-apo£ator_0peratedI by Vapor Compression
forjgduim Sulfate Re^wej^L._j1j[th^^ —*
The vapor compression vertical tube foam evaporator VC-VFFE, constructed
and tested under a previous EPA-sponsored project in this Laboratory and shown
in Mgure 8 d) , was modified for this phase of the work. The modifications
PTi? / ^'"e preheatir'S b> counterflow heat exchange with the distillate
and blowdown and improved the performance ratio of this equipment by varying
the belt drive ratio for a better match between the motor and the compressor
(Roots type) available for this evaporator. In earlier use (4), the compressor
p:°v±?,a'a!e:, :9* !™p:e!slon rat??(and AT> but.f- <*- aPPncation it
*aS SrV.TrE3*6? that 3 reduced compression ratio would improve the economy of
the VC-VTFE un.t. Two sets of V-belt pulleys, adapters, and belts were used
in these tests. One set provided a compressor speed of 300 rpm and the other
600 rpm compared with the 915 rpm of the original equipment (4).
The use of the interface-enhanced method of operating an evaporator also
known as vertical tube foam evaporation (VTFE), was described in detail else-
where (4,5) and has also been the subject of two earlier EPA-sponsored studies
in this Laboratory (4,14,16;. The substance of the VTFE improvement is to
cause the evaporating liquid to flow over a heat transfer surface in the form
of a foaming layer of the liquid and its vapor. This mode of two-phase flow
is facilitated by the addition of a few parts per million of a selected
surfactant or foaming agent. This foamy flow over the heat transfer surface
increases the evaporation-side coefficient several-fold and can approximately
double the usual overall heat transfer coefficient obtainable without foamy
flow in modern evaporators having double-fluted distillation tubes.
In the work discussed here, the VC-VTFE mode of operation was applied to
the concentration of softened irrigation drainage water (1,3), and to the
evaporatative-crystallization of sodium sulfate anhydride from such a
28
-------
concentrated brine.
(c) Vapor Compression. VTFE Pilot Plant Used in This Study
In this study, a vertical tube evaporator of 20 cu. m/day (5000 US
gallon/day) capacity was operated with a Roots type vapor compressor. The
evaporator was rotatably mounted, to permit comparative tests with upflow
and downflow modes of feed flow through the distillation tubes. The distil-
lation tubes were double-fluted aluminum-brass, 3-8-cm (1.5-in.) diameter by
1 83 m (6-ft.) heated length (supplied by Yorkshire Imperial Metals, Ltd.,
Leeds, England). Forty-nine tubes were assembled in a wedge-shaped bundle.
These tubes were "0"-ring sealed through naval bronze tube sheets, with the
tubes spaced 0.64 cm (0.25-in.) apart on equilateral triangles.
The evaporator, shown in Figure 8, was constructed of type 316 stainless
steel and mounted on an A-frame. Rotation of the evaporator unit around the
compressed vapor pipe permitted either upflow or downflow operation. For this
purpose, two mist eliminators of stainless steel wire mesh were installed at
appropriate elevations within the vapor-liquid disengagement vessel. The
vapor produced during passage of the liquid through the distillation tubes was
separated from the residual liquid in the disengagement vessel and passed
upward through the mist eliminator and through a 15-2-cm (6-in.) diameter
stainless steel pipe to the compressor. The latter was driven with a 25~np
electric motor at 300, 600 or 900 rpm. A digital, totalizing watt hour meter
was provided for the continuous monitoring of compressor energy consumption.
The vapor compressor increased the vapor pressure by up to about 3^.5 KPa (5^
psi); the compression ratio imposed upon the vapor was variable by the experi-
mental use of one or more of three by-pass lines of 5-1-cm (2-in.) diameter
with valves which diverted compressed vapor back to the inlet of the compressor.
Different feed distribution devices were preferred for each mode of
operation to provide the most effective interface enhancement effects. For
the downflow mode used, feed was directed through orifices onto the top tube
sheet at locations interstitial of the tubes and then passed into the tubes as
annular layers at about 4 liters per tube per minute.
Surfactant additives and their concentration levels were selected to ^
sustain foamy flow within the tubes but not beyond their outflow ends. This
limited the foaming agent additive to concentrations below the level where
foaming in the vapor-liquid disengagement vessel would present problems such
as carry-over and excessive pressure drop. Surfactant concentrations of 15
to 50 ppm were used.
Heat transfer performance data were obtained from temperature measurements
and from the rate of condensate flow which was measured directely with a
calibrated, in-line cylindrical collection vessel. From this vessel the
condensates were either pumped for removal from the system in measured ^
quantities, or for recycle to the feed. Temperature data were obtained with
platinum resistance probes having associated bridges and digital voltmeters
providing a continuous, direct digital display of the steam-side temperature
(Ts) in the tube bundle, the feed temperature (Tf), and the residual brine
29
-------
• I CURE 8. VAPOR COMPRESSION VERTICAL TUBE FOAM EVAPORATOR^DOWNFLOW)
30
-------
temperature (Tb) in the disengagement vessel. These data were checked by
means of mercury-filled thermometers also mounted in the individual streams.
The seawater feed was preheated, and the distillate and discharged brine
were cooled by counterflow heat exchange, utilizing a plate type heat^ex-
changer having 90-10 copper-nickel plates (supplied by Mechanical Equipment
Company (New Orleans, USA) and providing an approach temperature of about 5 C.
Further heating and deaeration of thefeed was done by letting ,t flow downwards
as a film over the inside wall of a 7.6 cm (3-!n.) diameter copper tube 3.05 M
(10-ft ) long, in counter-flow with steam blown upwards through the tube. The
degassed, preheated feed was subsequently transferred to the evaporator with a
pump at a controlled flow rate. (The evaporator vessel and the vapor com-
pressor were preheated by passing steam into them followed by draining all
condensates before introducing the preheated feed). Additional heating of the
feed before start-up was by a steam-heated coil of soft copper tubing mounted
into the feed distributor section above the orifice plate. During operation
of the VC-VTFE, compressed vapor vented from the tube bundle was passed
through this coil before its discharge. The objective of this operation was
to maintain the temperature of the recycled residual feed brine close to or
slightly above the evaporation temperature in the distillation tubes.
Brine was recycled from the bottom of the disengagement vessel through a
5 1-cm (2-in.) diameter copper tube with a centrifugal pump, and its flow
controlled with a valve. This feed was injected into the feed distributor on
a tangent, causing swirling flow over the coil heated with vented vapor, and
thence through the orifice plate into the tube bundle at a flow of about 3.8- _
1 (1-gal.) per minute per tube.
Several windows mounted on the inlet distributor section and on the
disengagement vessel permitted close observation of the vapor-liquid flows.
(d) Crystallization of Sodium Sulfate by Conventional VC-VTE and by the
Foamy-Flow Enhanced VC-VTFE Procedure.
The vapor compression vertical tube evaporation (VC-VTFE) facility (see
Figure 8) was preheated by passing steam into both the steam- and brine-side
and then draining all condensates. The preheated feed, containing 274 sodium
sulfate in water, was pumped into the vessel, and circulated through the tubes,
its temperature raised to about 103°C by passing steam through the copper tube
heating coil located in the brine distributor above the tube bundle. Next,
the vapor compressor was turned on and the compressed vapor by-pass valves were
closed in sequence, raising the steam-side pressure stepwise to its maximum
level The condensate collected from the steam-side of the tube bundle was
returned to the recirculated brine feed until steady-state was established, as
indicated by the steam- and brine-side temperatures, the condensate flowrate,
and the electrical energy drawn by the VC-motor. Deaeration of the feed by
passing of non-condensible gases through the tube bundle to the atmosphere
with a small flow of compressed vapor as a bleed-off was usually completed and
steady-state established within about 20 minutes. Gradual evaporation-
concentration of the brine was then initiated by diverting the distillate
(condensate) flow from the brine recycle stream and rejecting it into a
31
-------
f thep
eeoe f ^ ^7^ coeff!cient and the VC power consumption
andPB n Hour* P T^" ° *"? ^ COncentrat Ion f^tor, shown as plots
A and B ,n F.gure 9. These results show a continuous increase in the VC
on r6SP°nSlve tO the -creasing brine concentration due to
?«; J he P°Wer co^umption per 1000 liters of distillate dips
for the brine concentration factors of 1.2 to 1.3 and then rises
contmuously ,n nverse proportion to the VTE heat transfer coefficient U
increasSeSrseannf Y r"P°nS'^.to Cryfan'zat'on wa* initiated at a feed concentration
tactor of 1.5 and proceeded to a factor of 2.5 at which level the residual
. e e resua
t on J±me ^ t0° SmaH to maintain 'ts recirculation, and the evapora-
,
rur H , PPeCL The accumulated distillates (125 1) were then
leve? Tn°^-e ^ ^^f ?tUt?ng ?t t0 !ts °nslnal 21% concentration
^ "
99 act f °f Hnear 3lkyl benzene sulfonic acid
aS 3 "9 ant bef°re rePeatin9 the above evaporation-
crvstsH ~-
of hril A Sr°u !' th'S tFme imP°s!n9 foamY evaporative two-phase flow
of br.ne downward through the distillation tubes and repeating the data and
observations taken above.
The data from the VTFE mode of operation are recorded in Table k and
d ?H n 8- ^T C/nd I 11 Fl9Ure 9' Go°d foamy f1ow was observed from the
distillation tubes for the first data set (brine concentrations of 1 to 1.1)
but httle foammess emitted from the tubes thereafter until crystallization'
set m [at 1.5;. Good foamy flow persisted from here on throughout the
crystallization until the process was stopped as before (at 2.5) when the
residual brine volume became too small to sustain its recirculation. The
data plotted as curve C in Figure 9 shows the overall VC-VTFE heat transfer
performance during the gradual evaporation-concentration process; a gradual
reduction in performance was again observed responsive to the boiling point
elevation requiring gradually increasing AT (and Ts). However, this plot
shows _ that whenever foamy flow was maintained, the heat transfer performance
was significantly higher than in the previous process (plot A) without
surfactant addition and foamy flow. The VC-VTFE performance was typically
about 30/0 higher than the VC-VTE performance during the crystallization stage.
These data are consistent with the VC power consumption that was measured
directly and independently with a kilowathour meter as plotted in curve D.
The latter fell below the energy consumption of the first process series
(plot B) . During the most significant crystallization stage, VC-VTFE opera-
tion showed an energy saving of about 25%.
It was apparent from these data and from other similar experiments with
this VC-VTFE pilot plant that foamy flow of the evaporating brine-salt slurry
appears to be a prerequisite to enhanced heat transfer performance and VC
energy reductions. Such foamy flow was not always readily obtained with
brine-solid slurries. It appears that for the most effective use of VTFE for
different evaporation-crystal 1 izations, VTFE will need to be pilot plant
tested for each application prior to its use. Apparently, the salt
32
-------
TABLE 3. CRYSTALLIZATION OF Na2S0lt BY VC-VTE WITH SLURRY-FEED RECYCLE
OJ
Feed:
Time
10.30
10.35
10.40
10.45
10.50
10.53
10.57
11.00
11.03
11.06
11.09
N32S04
Ts,°C
108.4
108.8
109.2
109.4
109.6
109.6
109-9
109.7
109.6
109.3
109.2
. 56
Tb,°
103.
103.
103.
103.
103.
103.
103.
102.
102.
101.
100.
.7
c
1
2
4
7
6
5
2
7
1
3
7
kg
AT
5
5
5
5
5
6
6
7
7
8
8
in
.'C
.4
.5
.7
.8
.9
.1
.7
.1
.6
.0
.5
tap water,
Power Di
kWh/h
15-86
16.39
16.98
17.65
18.20
19.44
19.8
20.62
21.6
22.1
23.6
208 liters (27%)
stillate
1/h
478.1
495-5
545.0
592.4
579.8
567.7
567.7
534.5
469.9
446.8
432.6
U-Coefficient
W/m2-°C
5246
5326
5627
6047
5752
5485
5031
4474
3679
3310
3021
Dist
Li
11
22
34
45
56
68
79
90
102
113
125
5 Hate
ters
.36
.72
.09
.44
.8
.16
.52
.88
.24
.6
.0
Power/Con-
sumption Remarks
kWh/1000 1
33.18 No Crystall ization
33.08
31.15 " "
29.78
31.39
34.24 Crystallization
34.87
38.57
45.97
49.41
54.43
-------
TABLE 4. CRYSTALLIZATION OF Na2S0lt BY VC-VTFE SLURRY-FEED RECYCLE
Feed :
Time
12.08
12.17
12.20
12.23
12.26
12.29
12.32
12.35
12.38
12.42
12.45
NagSO^, 56.7 kg
Ts,
106
108
109
109
109
110
109
109
109
109
109
°C
.4
.9
.2
.3
.6
.0
.9
.6
.4
.3
.1
Tb,°C
102.1
102.9
103.2
103.4
103.7
103-9
104.0
103.6
103.0
102.4
101.8
AT
+ Linear Alkyl Benzene Sulfonic Acid,
°r
4.
5.
5.
5.
5.
6.
5.
5.
6.
6.
7.
3
9
9
9
8
1
9
9
4
8
2
Power
kWh/h
13.7
16.36
16.74
17.61
18.0
18.62
19.15
18.95
19.15
19.32
20.62
Disti 1 late
1/h
594.8
592.4
592.4
619.4
592.4
599.0
586.1
592.4
573.7
524.1
504.7
U-Coeff icient
W/m2-°C
6649
5899
5893
6217
5996
5786
5871
5888
5269
4524
4139
10 ml in tap water, 208 liters (27%)
Distillate
L i ters
11.36
22.71
34.07
45.42
56.78
68.13
79.49
90.84
102.20
H3.55
125.0
Power/Con-
sumption
kWh/1000 1
23.03
27.61
28.27
28.53
30.38
31.10
32.68
31.97
33.37
36.86
40.87
Remarks
No Crystal 1 ization
ii ii
ii ii
ii n
ii ii
Crystal 1 ization
1 1
n
ii
n
1 1
-------
EVAPORATION-CRYSTALLIZATION OF SODIUM SULFATE BY VC-VTE AND VC-VTFE
•z.
— PLOT C O VC-VTFE Heat Transfer Performance
< PLOT A • VC-VTE Heat Transfer Performance
— PLOT D D VC-VTF£ Power Consumption
-> PLOT B • VC-VTE Power Consumption
.6000
£ 5000
o
o
a:
1(000
3000
a
i_
o
•3
o
1.0
1.5 2.0
FEED CONCENTRATION FACTOR
2.5
Figure 9
FIGURE 9. EVAPORATION-CRYSTALLIZATION OF SODIUM SULFATE
35
-------
concentration as well as the surfactant concentration in addition to the AT
tion are fact°rs that affect foulness during VTFE crystalliza-
REFERENCES
Sephton, H H., and G. Klein. A Method of Using Irrigation Drainage
Water for Power Plant Cooling. ,n: Proceedings! Firs? Desalinai?on
Continent, Mexico City, 1976.
2. Klein, G and H. H. Sephton. Reclamation of High-Sulfate Irrigation
Pr.Plan C°°1In9- -n: Recent Deve Epie i in
and Harr[s
3. Lindholm, R., D. Walker, H. H. Sephton and G. Klein. Agricultural
water for Power Plant Cooling: Development and Testing of Treatment
anS SSen'« • Trt by the Ca1iforn^ Department of Water Resources
June i;78?IVerS'ty °f CaUf°rn?a Seawater Conversion Laboratory *06 pp!
No. 382en t0 °^°» ^ Liquids.
^
U.S. Environmental Protection Agency. 53 pp., 1977.
Notice of Intention of San Diego Gas and Electric Company to File Appli
"si?on°L.nLI iC;tF°n °f SAUndeSert Nuc1ear P1a^, Together With Tra'ns
mission Lines and Necessary Appurtenances. 1976.
Cr \'f P°5en!Ial Use of Agricultural Waste Water for Power Plant
California Department of Water Resources Report No.
9. Stone s Webster Engineering Corporation. Water Treatment Demonstration
Facility Report for Sundesert Nuclear Plant. Boston, Massachusetts,
10. Stone & Webster Engineering Corporation. Conceptual Engineering Cooling
System and Associated Water/Waste Treatment Systems for Sundesert Nuclear
Plant. Boston, Massachusetts, 1975.
36
-------
11. Sephton, H. H. The Use of Interface-Enhanced Vertical Tube Evaporation,
Foam Fractionation and Ion Exchange to Improve Power Plant Cooling With
Agricultural Wastewater. Proposal to the California Department of Water
Resources, UCB-Eng-384l , 1975-
12. The California Water Plan - Outlook in 197^. Bulletin Number 160-7^.
California Department of Water Resources. Summary Report, Table 27,
13. The California Water Code. Division 1, Chapter 6, Section
Ht. Valdes-Krieg, E., C. J. King, and H. H. Sephton. Foam and Bubble Frac-
tionation for Removal of Trace Metal Ions from Water. In: Environmental
Protection Agency Conference on Traces of Heavy Metals in Water, Removal
Processes and Monitoring, Center of Environmental Studies, Princeton
University. EPA Publication No. 902/9-7^-001, 1973.
15. Sephton, H. H. Recycle of Power Plant Cooling Tower Slowdown with Inter-
face Enhancement. In: Proceedings of the Second National Conference on
Complete WateReuse, Chicago, Illinois, 1975.
16. Sephton, H. H. Vertical Tube Evaporation with Fluted Tubes and Inter-
face Enhancement: Comparative Performance of Upflow Versus Downflow of
the Feed. ASME Paper No. 75-HT-^3, American Society of Mechanical
Engineers, New York, N.Y. In: ASME-AIChE Heat Transfer Conference, San
Francisco, CA, 1975.
17. Valdes-Krieg, E., C. J. King and H. H. Sephton. Removal of Surfactants
and Particulate Matter from Seawater Desalination Slowdown Brines by
Foam Fractionation. Desalination, 16:39~53, 1975-
37
-------
REPORT NO.
EPA-600/7-80-047
TECHNICAL REPORT DATA
(flease read Instructions on the reverse before completing)
2.
4. TITLE AND SUBTITLE
Feasibility of Recovering Useful Salts from Irrigation
Wastewater Concentrates Produced by Power Plant
Cooling
6. PERFORMING ORGANIZATION CODE
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
March 1980
Hugo H. Sephton
9. PERFORMING ORGANIZATION NAME AND ADDRESS
The University of California
Sea Water Conversion Laboratory
47th and Huffman Boulevard
Richmond. California 94804
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
INE827
11. CONTRACT/GRANT NO.
Grant R804760
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 PER
Final; 10/76 - 12/79
'ERIOD COVERED
14. SPONSORING AGENCY CODE
919/541-2683! """"
EPA/600/13
pr°^ect officer is Theod°re *• Brna, Mail Drop 61,
The report evaluates the feasibility of a novel energy-conserving way to
recover useful salts (sodium sulfate and calcium sulfate) from concentrated brines
by evaporation/crystallization. The concentrated brines examined were cooling
tower blowdown from agricultural wastewater and this blowdown after further con-
centration and use in ion exchange regeneration. Laboratory and pilot tests were
made, with both conventional evaporation/crystallization and interface-enhanced,
vertical-tube foam evaporation which increases evaporation. Sodium sulfate and
calcium sulfate recovery provides potential capital cost savings, about S2 million
for a 1000-MWe power plant. Sale of these products is an added incentive to their
recovery. The use of foamy vapor/liquid flow on the evaporating brine/crystal
SlUoo
-------
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Center for Environmental Research Information
Cincinnati, Ohio 45268
OFFICIAL BUSINESS
PENALTY FOR PRIVATE USE, S3OO
AN EQUAL OPPORTUNITY EMPLOYER
POSTAGE AND FEES PAID
US ENVIRONMENTAL PROTECTION AGENCY
EPA-335
If your address is incorrect, please change on the above label
tear off; and return to the above address.
If you do not desire to continue receiving these technical
reports, CHECK HERE\H\; tear off label, and return it to the
above address.
Publication No. EPA-600/7-80-047
------- |