PB-224 583
RECOMMENDED METHODS OF REDUCTION, NEUTRALIZATION,
RECOVERY, OR DISPOSAL OF HAZARDOUS WASTE
VOLUME IV, DISPOSAL PROCESS DESCRIPTIONS,
BIOLOGICAL AND MISCELLANEOUS WASTE TREATMENT
PROCESSES
TRW SYSTEMS GROUP
PREPARED FOR
ENVIRONMENTAL PROTECTION AGENCY
AUGUST 1973
DISTRIBUTED BY:
National Technical Information Service
U. S. DEPARTMENT OF COMMERCE
-------
4. Tuic and Subt.tic Recommended Methods of Reduction, Neutralization,
Recovery, or Disposal of Hazardous Waste. Volume IV, Disposal
Process Descriptions - Biological and Miscellaneous Waste
Treatment Processes
BIBLIOGRAPHIC DATA
SHEET
1. Report No.
EPA-670/2-73-053-d
PB 224 583
5- Report Date
issuing date - Aug. 1973
R. s. Ottinger, J. L. Blumenthal, D. F. Dal Porto,
6.
7.
G. I. Sruber. M. J. Santy. and C. C. Shih
&• Performing Organix,uion Kept.
N°" 21485-6013-RU-QO
(.-Performing Organization Name and Address
TRW Systems Group, One Space Park
Redondo Beach, California 90278
10. Ptoject/Task/Wotk Unn No.
11. Contract/Grant No.
68-03-0089
12. Sponsoring Organization Name and Address
National Environmental Research Center
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
13. Type of Report & Period
Coveted
Final
14.
15. Supplementary Notes
Volume IV of 16 volumes,
16. Absttacts
This volume provides descriptions of selected processes currently utilized for the
treatment or disposal of hazardous wastes. These descriptions detail the important
features of each process and discuss their applicability to the various classes of
waste materials. This volume contains descriptions of four biological processes
including activated sludge, aerated lagoons, oxidation ponds, and trickling filter.
There are also five miscellaneous processes described including dialysis, electro-
dialysis, ion exchange, reverse osmosis and radioactive waste solidification.
17.' Key Words and Document Analysis. 17n. Descriptors
Waste Treatment
Activated Sludge
Aerated Lagoons
Oxidation Ponds
Trickling Filter
Dialysis
Electrodialysis
Ion Exchange
Reverse Osmosis
Radioactive Waste Solidification
17b. Idcntifiers/Open-Endcd Terms
17c. COSATI Field/Group
. 06T. Q7B . Q7C. 07E; 13B; ^ 3R . lgA; 1gB
18. Availability Statement
Release to public.
LIBRARY
Environ Prot ftg«
BSsaii, hi* ltf*ey
,wga
19. Security Class (This
Report)
UNCLASSIFIED
IF"'"
is Ohi
20. Si-Lurity Class ('I his
UNCI.A^II IT!)
21. No. of
-------
EPA-670/2-73-053-d
August 1973
RECOMMENDED METHODS OF
REDUCTION, NEUTRALIZATION, RECOVERY
OR DISPOSAL OF HAZARDOUS WASTE
Volume IV. Disposal Process Descriptions -
Biological and Miscellaneous
Waste Treatment Processes
By
R. S. Ottinger, J. L. Blumenthal, D. F. Dal Porto,
G. I. Gruber, M. J. Santy, and C. C. Shih
TRW Systems Group
One Space Park
Redondo Beach, California 90278
Contract No. 68-03-0089
Program Element No. 1D2311
Project Officers
Norbert B. Schomaker
Henry Johnson
Solid and Hazardous Waste Research Laboratory
National Environmental Research Center
Cincinnati, Ohio 45268
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
-------
ERRATA - VOLUME IV
Pages 20, 34, 44:
Reference No. 0534 is referred to as Reference No. 0314
in Volume XVI.
ii
-------
FOREWORD
Man and his environment must be protected from the adverse
effects of pesticides, radiation, noise and other forms of pollu-
tion, and the unwise management of solid waste. Efforts to protect
the environment require a focus that recognizes the interplay between
the components of our physical environment—air, water, and land.
The National Environmental Research Centers provide this multidisci-
plinary focus through programs engaged in:
• studies on the effects of environmental
contaminants on man and the biosphere, and
« a search for ways to prevent contamination
and to recycle valuable resources.
Under Section 212 of Public Law 91-512, the Resource Recovery
Act of 1970, the U.S. Environmental Protection Agency is charged
with preparing a comprehensive report and plan for the creation of
a system of National Disposal Sites for the storage and disposal of
hazardous wastes. The overall program is being directed jointly by
the Solid and Hazardous Waste Research Laboratory, Office of Research
and Development, National Environmental Research Center, Cincinnati,
and the Office of Solid Waste Management Programs, Office of Hazard-
ous Materials Control. Section 212 mandates, in part, that recom-
mended methods of reduction, neutralization, recovery, or disposal
of the materials be determined. This determination effort has been
completed and prepared into this 16-volume study. The 16 volumes
consist of profile reports summarizing the definition of adequate
waste management and evaluation of waste management practices for
over 500 hazardous materials. In addition to summarizing the defini-
tion and evaluation efforts, these reports also serve to designate a
material as a candidate for a National Disposal Site, if the material
meets criteria based on quantity, degree of hazard, and difficulty of
disposal. Those materials which are hazardous but not designated as
candidates for National Disposal Sites, are then designated as candi-
dates for the industrial or municipal disposal sites.
A. W. Breidenbach, Ph.D., Director
National Environmental Research Center
Cincinnati, Ohio
-------
REVIEW NOTICE
The Solid Waste Research Laboratory of the National Environmental
Research Center - Cincinnati, U.S. Environmental Protection Agency has
reviewed this report and approved its publication. Approval does not
signify that the contents necessarily reflect the views and policies of
this Laboratory or of the U.S. Environmental Protection Agency, nor does
mention of trade names of commercial products constitute endorsement or
recommendation for use.
The text of this report is reproduced by the National Environmental
Research Center - Cincinnati in the form received from the Grantee; new
preliminary pages and new page numbers have been supplied.
-------
TABLE OF CONTENTS
VOLUME IV
DISPOSAL PROCESS DESCRIPTIONS
BIOLOGICAL AND MISCELLANEOUS WASTE TREATMENT PROCESSES
Page
Biological Processes
Activated Sludge 1
Aerated Lagoons 21
Oxidation Ponds 35
Trickling Filter 45
Miscellaneous Processes
Dialysis 57
Electrodialysis 75
Ion Exchange 91
Reverse Osmosis 103
Radioactive Waste Solidification 117
-------
ACTIVATED SLUDGE PROCESS
1. INTRODUCTION
The activated sludge process is a continuous process in which incomina
waste water is mixed with biological growths recirculated from the final
clarifier to form a mixed liquor, followed by aeration, separation of the
sludge from the effluent, and subsequent return of a portion of the settled
sludge to be mixed with additional waste. The activated sludge for the
process is formed by the growth of micro-organisms inbedded in gelatinous
matrices feeding on organic material in the waste water under aerobic
conditions. These flocculent suspensions that are developed usually posses
excellent settling and oxidative properties.
The conventional plug flow activated sludge process (Figure 1) is normally
designed for the treatment of domestic wastes, and yields BOD* reductions UD to
95 percent. The conventional system, however, has several disadvantages when
industrial wastes are to be treated - including uneven distribution of
biological solids and contaminants in the aeration basin and the inability
to cope with shock loading. Complete mixing aeration designs* (Figure 2)
are therefore generally considered for industrial waste water treatment.
Concentration gradients are minimized and load variations are dampened in
complete mixing areators, and the resultant effluent quality is usually
better than that achieved by the conventional process. Numerous modifications
to the activated sludge process exist and the more important of those are
described briefly in a later section.
M
Biochemical oxygen demand (BOD) is the amount of oxygen required for
the biological oxidation of the organic matter in waste water.
t The complete mixing activated sludge process is one in which the untreated
wastes are mixed almost instantaneously throughout the entire aeration
basin. Mixing is normally effected by mechanical aerators.
-------
Primary
Clarifier
Final
Clarifier
Industrial
Waste
»
1
Plug-Flow
Aeration
Basin
Return Sludge
Treated
Effluent
Settleable
Solids to
Disposal
Waste
Sludge to
Disposal
Figure 1. Conventional Activated Sludge Process
Primary
Clarifier
Final
Clarifier
Industrial
Waste
Complete
Mixing
Aerator
Treated
Effluent
Return Sludge
Settleable
Solids to
Disposal
Waste
Sludge to
Disposal
Figure 2. Complete Mixing Activated Sludge Process
-------
2. OPERATION PRINCIPLE
The basic mechanism for removal of organic material from waste water
by the activated sludge process can be represented by the three chemical
reactions occurring simultaneously - energy, synthesis, and endogenous
respiration. Energy and synthesis are often referred together as oxidative
assimilation, and involve the consumption as food of organic material
present in the waste by micro-organisms. A portion of the food is used as
fuel to supply energy for metabolism while the remaining food provides
building components resulting in the formation of new cellular material.
Endogenous resolration Is generally considered as a maintenance mechanism
of the micro-organisms, with the production of carbon dioxide, water, and
ammonia. In the initial growth phases when organic food is present, endogenous
respiration occurs simultaneously with the oxidative assimilation process.
As the food supply diminishes, however, organic matter within the sludge is
utilized and results in the endogenous growth phase.
The reactions of energy, synthesis, and endogenous respiration can be
summarized in the following way:
Organic material + bacteria + 02 - NH3 ^C02 + H20 + NH3 + AH
where AH is the heat liberated by the reaction and 1s related to the
energy made available for metabolic purposes.
The bacteria necessary to consume the organic material in the waste
are present in the activated sludge floes. The bacteria type depends
on the characteristics of the organic matter in the waste water, although
protozoa and metazoa are usually present on the surface of the floes.
-------
3. PROCESS DESIGN
The principle components of the activated sludge process include the
following units:
(1) Pretreatment units for screening, grit removal, and primary
sedimentation of raw waste;
(2) Diffuser or mechanical aerator;
(3) Aeration basin;
(4) Secondary sedimentation tank;
(5) Sludge return pump;
(6) Excess sludge disposal facility.
Screens are generally present in waste water treatment systems to remove
large objects but they may not have any value with industrial wastes. Aerated
grit removal units, however, are particularly suitable for the separation of
grit from highly organic industrial wastes. Primary sedimentation tanks are
used to reduce the amount of settleable solids in the waste and could lead to
as much as 60 percent removal of the suspended solids and up to 35 percent
BOD reduction. Because of the cost associated with sludge processing and
disposal, however, primary sedimentation tanks have been eliminated in some
activated sludge system.
Basically, there are two types of devices utilized to introduce air into
the waste water in activated sludge systems. Diffusers generally distribute
compressed air supplied by blowers to points of use in the aeration basin,
and entrain the air in waste water through the use of diffusers which divide
the air into fine bubbles to provide better oxygen transfer. Mechanical
aerations, on the other hand, transfer atmospheric oxygen by moving the
*
"mixed liquor" and bringing it into contact with air, and are more widely
used in treating industrial wastes because of easier maintenance. Both types
The activated sludge when combined with the influent waste water is known
as mixed liquor.
-------
of aeration systems are designed to insure sufficient dissolved oxygen
dispersed throughout the aeration basin and maintain the biological floes
in suspension, thus providing maximum contact among bacteria, waste water,
and oxygen. 'Combinations of diffusers and mechanical aerators are sometimes
used to meet high rates of oxygen demand.
The primary variables influencing the design of the activated sludge
process are the influent BOD and specified effluent BOD, the rate of waste
water flow, the BOD removal kinetics of the waste water, the influent
volume suspended solids, the oxygen requirements, the excess sludge yield,
and the settleability of the activated sludge. Other factors that enter
indirectly into the design include temperature and pH levels, and the nutrient
requirements, which all have a significant effect on the biological kinetics
of the system.
The proper design and operation o,f activated sludge systems involves
the balance of the relationship between the oxidation and synthesis mechanisms
for removing the organic matter from the mixed liquor. For this reason, the
Water Pollution Control Federation recommends the "sludge loading ratio" (SLR)
as the loading parameter to be used for activated sludge. The sludge loading
ratio is a measure of the food-to-micro-organisms ratio and is normally ex-
*
pressed in pounds of BOD per day per pound of MLVSS. Good activated sludge
operations are normally attained at organic loadings of between 0.3 and 0.7
Ib BOD/day/lb MLVSS. The equation for the calculation of the SLR is:
SLR = 24 Sa (1)
V
MLSS or "mixed liquor suspended solids" is the amount of sludge solids in
the mixed liquor and MLVSS or "mixed liquor volatile suspended solids" is
the volatile portion of the MLSS. Both are usually expressed in mg/1 and
related by MLVSS = 0.8MLSS. MLVSS is supposedly a better representation
of the active biological mass in the system. The presence of higher than
nominal level of inert solids, however, could lead to MLVSS considerably
less than MLSS.
-------
where Sfl is the influent BOD in mg/1 , Xy is the average MLVSS concentration
in the aeration basin in mg/1, and t is the aeration time in hr. Both the
MLSS and the MLVSS concentrations in the aeration basin depend partly on
the recycle ratio of the settled sludge. To produce good flocculation and
good liquid-solid separation, minimum MLSS of 1000 mg/1 are generally needed.
The nominal MLSS level is between 2000 and 5000 mg/1.
Experimental programs are almost always required to develop the necessary
design data for activated sludge systems. The design parameters requiring
determination will become apparent in the process design equations proposed
by Eckenfelder.126"
Aeration time, t (hr)
- 24 Sa (2)
~
from which aeration basin volume is readily computed knowing the rate of
waste water flow.
Effluent BOD, Sg (mg/1)
kXvt
where k is the biological reaction rate constant and is a strong function
of the waste water characteristics and temperature. The value of k ranges
from 0.00006 to 0.002 (1/mg - hr).
Excess sludge yield, A Xy (Ib/day)
AXy = 8.34 Q (SQ + a (S, - Sg) - b -^ • Xy) (4)
where S is the influent volatile suspended solids in mg/1, a is the sludge
yield coefficient in Ib VSS per Ib of BOD removed, Q is the rate of waste
water in MGD, b is the rate of sludge auto-oxidation fraction per day. The
parameter a reflects the synthesis of biological solids for soluble wastes
and has values ranging from 0.2 to 0.4 Ib VSS per Ib BOD ultimately removed
from the waste. The parameter b will usually have a value of 0.20 to 0.26
per day.
-------
Oxygen requirements, 02 (Ib/day)
02 = 8.34 Q (a1 ($a - Sfi) + b' -^J— Xv) (5)
where a1 1s the oxygen utilized for growth per unit substrate removed in
Ib 0- per Ib BOD removed and b1 represents the endogenous respiration rate
in Ib 02 Ib VSS per day.
In addition, knowledge of the sludge volume Index (SVI) of the mixed
liquor from the aeration basin is necessary for good process operation.
Since the ability to separate out the activated sludge depends on the
rate at which the sol Ids settle to the bottom of the secondary sedimentation
tank. The SVI is volume in milliliters occupied by 1 gram (dry weight) of
sludge after 30 m1n of settling. A good SVI is approximately 100. A
sludge with a SVI or 50 1s dense and compact and gives a somewhat turbid
effluent. A SVI of 200 is indicative of a light fluffy sludge with poor
settling characteristics that could lead to sludge bulking. To avoid
bulking, the return sludge rate must be Increased to maintain the solids
concentration in the aeration basin.
To produce a clarified effluent with minimum solids concentration,
mixed liquor from the aeration basin flows to a secondary sedimentation
tank where the sludge floes are separated from the liquid. The important
parameters 1n the design of secondary sedimentation tanks for good sludge
separation include surface overflow rate, retention time, and tank depth.
Modern sedimentation systems for muncipal or industrial wastes are normally
designed with surface overflow rates of 800 to l,000_ga;/day/sq ft,
retention time of about 2 hours, and a minimum tank depth of 10 ft.
Return sludge pumps usually have capacity up to 50 percent of the waste
water flow and are normally operated at 20 to 30 percent of the waste water
flow and 8,000 to 10,000 mg/1 suspended solids.
-------
Since only a portion of the settled sludge is returned to the aeration
basin, the excess sludge produced in the process must be disposed of through
anaerobic or aerobic digestion, dewatering, and return to the land. Vacuum
filtration to remove the excess activated sludge from secondary sedimentation
is usually not practiced because of the poor filtration properties of
activated sludge.*
To illustrate the design and operation concepts discussed here, the
detailed flow diagram of a municipal conventional activated sludge facility
is shown (Figure 3). The waste water treatment is located at San
Antonio, Texas and some of its design characteristics are:
Rated capacity = 24 MGD
Two primary clarifiers
Four 2-pass aeration tanks
Total volume of aeration = 7,750,000 gal
No reaeration
Detention time = 6.0 hr
Air diffusers - ceramic tubes, 24 in.long, 3 in.diameter,
equally spaced along one side of each tank
Two final clarifiers
4. PROCESS ECONOMICS
An effective approach in estimating the capital cost of an activated
sludge system is to calculate the unit costs of the major process components
and add to the total direct material and field labor costs and indirect
construction overhead, engineering, and contingencies. The design and
cost bases of principal process components are summarized on the next
page.
* A complete discussion of the sludge handling and disposal practices is
obtainable from the Federal Water Pollution Control Administration.0283
8
-------
Raw Wastewater
4 Aeration Tanks
Air Diffusers
0.119 MGF
Raw Sludge
To Digesters
Screens &
Grinder
Grit
Collector
No. 2
Primary
larifier
No. 1
Primary
Clarifier
No. 1
Final
Clarifier
No. 2
Final
Clarifier
Compressed Air
107 cu ft./gal of
Raw Wastewater
Waste Activated
Sludge To
Mitchell Lake
Stream No.
Flow, MGD:
BOD, mg/1
Suspended Solids, mg/1:
pH:
Treated Effluent To
Chiorination and River
TYPICAL FLOWS AND QUALITIES
123456
22.4 22.3 9.6 31.9 0.85 21.5
221 177 27.8
230 122 3080 1130 3080 12.6
7.3 7.9
Figure 3. Detailed Flow Diagram* of Rilling Wastewater Treatment Plant at San Antonio,
Texas
Figure Ic of Reference 1297
-------
Process Component
Design Basis
Cost Basis
Aeration equipment
Aeration basin
Secondary Sed-
imentation
Retur sludge pump 'i
Excess sludge dis-
posal
Sludge digestion i
Thickening
Centrifugation
Oxygen requirement
Waste flow rate-organic
loading
Overflow rate
Return sludge rate
Excess sludge rate-
digestion kinetics
Mass loading
Waste flow-solids
loading
Cost/HP
Cost/volume
Cost/surface area
Cost/HP
Cost/volume
Cost/volume
Cost/waste flow
The critical design parameters that indirectly determine the capital
cost of activated sludge systems include the waste water flow rate, the
influent and effluent BOD, the average MLVSS, and the biological reaction
rate constant k. A capital cost model for the treatment of industrial
wastes by the activated sludge process has been developed by Eckenfelder,
^B3 and indicates the weigliu of each of the deslqn parameters
in determining the total capital cost of the treatment plant:
Cost(1n $1000)=Q
(0.69 + 0.00019SJ
Q
0.76
(17 (Sa/Se) + 215)x(1.05 +0.44/(kXv))
Operating costs for activated sludge systems are usually determined by
waste water and return sludge pumping capacity and the power requirement-
of the aeration equipment and clarifiers. Operating costs could range from
$10 to $2,000/MG of waste water treated.
-------
5. PROCESS MODIFICATIONS
High Rate Aeration
The high rate or modified aeration process is similar to the conventional
process except that higher loading factors of 2 to 5 Ib BOD per day per Ib
MLVSS are used. Both the construction and power costs are lower, but the
degree of purification is also reduced.
Tapered Aeration
This is an operational improvement of the conventional process. Since
oxygen demand is relatively high at the head end of the basin due to the
impact of high BOD of the influent and then decreases along the length of
the basin, decreasing amounts of air are provided to match the oxygen demand.
Lower power costs usually result.
Step Aeration
Step aeration involves the introduction of waste flow at intervals
throughout the length of the tank (Figure 4) and should be more appropriately
named "step loading". The intent is to minimize the effects of shock loading
and the oxygen demand gradient.
Extended Aeration
Extended aeration processes are characterized by long aeration periods
and low loading factors. The objective of the process is to reduce the
quantity of excess sludge produced and to achieve uniform BOD and suspended
solids concentration throughout the aeration basin. Because of the high
suspended solids level in the mixed liquor and the extended length of aeration
time, endogenous respiration becomes a determining factor in defining the
sludge quality in the process. Effluents from this process usually contain
relatively less BOD but higher suspended solids.
11
-------
Primary
Clarifier
Final
Clarifier
f
Industrial/' \
Waste "*i J
Vx
Settleable
Solids to
Disposal
Aeration
Basin
Waste
Sludge to
Disposal
Treated
Effluent
Figure 4. Step Aeration Activated Sludge Process
Primary
Clarifier
Industrial
Waste
Aeration
Basin
Settleable
Solids to
Disposal
Reaeratlon
Basin
Final
Clarifier
Return
TTudge < -
Waste
Sludge to
Disposal
Treated
Effluent
Figure 5. Contact Stabilization Activated Sludge Process
12
-------
Contact Stabilization Aeration
In the contact stabilization process, the return sludge is aerated
separately (Figure 5) until hydrolysis and conversion to cells is complete.
The sludge floes tnus produced have a higher capability of removing sub-
strate BOD when finally discharged into the main aeration basin. The
process has an actual contact between waste water and mixed liquor of 30
min to 1 hr, and is completely inapplicable to wastes containing a
high fraction of soluble organics.
A summary (Table 1) of some of the differences in loading parameters
aniong variations of the activated sludge processes has been presented by
Genetelli.1266
6. PROCESS APPLICABILITY
The complete mixing activated sludge process has been applied very
extensively in the treatment of refinery, petrochemical, and biodegradable
organic waste waters. In order to avoid adverse effects on the system,
pretreatment to remove certain materials from the waste water are often
necessary. Some materials present could interfere with the treatment
process or be toxic to the bacteria in the sludge, and some limits must
be placed on the amount permissible in the waste water. Maximum limits
for selected waste water characteristics (Table 2) are given by Jones
and Klein.0669
In addition, strongly acidic or alkaline wastes must be neutralized
to pH limits between 5 and 9 before discharge to the activated sludge
system. Temperature should not exceed liO F for good operation because
of lower oxygen solubility, and essential nutrients of ammonia and phosphorus
must be supplemented if these are not originally present in the industrial
waste water.
13
-------
TABLE 1
LOADING PARAMETERS FOR ACTIVATED SLUDGE PROCESSES
Activated Sludge Sludge Loading Ratio
Process -lb BOD/Day/lb MLVSS
Conventional
Complete Mixing
High-Rate Aeration
Tapered Aeration
Step Aeration
Extended Aeration
Contact Stabilization
0.2 -
0.2 -
2 -
0.2 -
0.2 -
0.05-
0.2 -
0.7
0.7
5
0.7
0.7
0.2
0.7
Volumetric Loading Ib BOD/
Day/ 1000 ft3 of Aeration
Tank Capacity
35
50
100
35
50
20
70
14
-------
TABLE 2
MAXIMUM PERMISSIBLE LEVEL OF WASTE WATER CHARACTERISTICS
FOR BIOLOGICAL TREATMENT0534.0669
Characteristics Maximum Permissible Level
Oil and Grease 75 mg/1
Nonbiodegradable Suspended Solids 125 mg/1
Sulfides 250 mg/1
Cya>nides 1 ppm
Arsenic 0.2 ppm
Antimony 0.2 ppm
Cadmium 2.0 ppm
Chromium 0.5 ppm
Copper 0.2 ppm
Lead 1 ppm
Mercury 0.07ppm
Nickel 1 ppm
Silver 0.03 ppm
Zinc 5 ppm
15
-------
The activated sludge process Is generally applied where a high degree
of BOD removal is required and where suitable land is not available. Organic
wastes that have been successfully handled by the process include the
following: acetaldehyde, acetic acid, acetone, acrolein, high boiling alcohols,
ethyl alcohol, formaldehyde, formic acid, glycerol, glycols, isobutyl alcohol,
isopropyl alcohol, methanol, and phenolics. A summary by Jones (Table 3)*
of activated sludge plants treating petrochemical and organic wastes indi-
cates the representative types of mixed wastes that could be treated by the
activated sludge process along with the effluent quality obtained. The
typical BOD removal efficiencies are in the 90 to 95 percent range.
It has been claimed that toxic metals in waste water could be largely
removed by sedimentation and biological filtration in the activated sludge
process. f However, biological treatment by activated sludge does not
produce a metal free final effluent and in fact, toxic quantities of the
metals may still be present in the effluents. The approximate amounts re-
moved by the process are : Cr, 70 percent; Ni, 30 percent; Zn, 60 percent;
Pb,-90 percent; Cu, 80 percent.0669
Although the activated sludge process is probably not applicable to
the direct treatment of hazardous wastes in National Disposal Sites,
installation of the process with a proper analytical monitoring system
is recommended for treating all the process waste water (spent cleaning
solutions, incinerator scrubber waste liquor, and cooling water) generated
within the National Disposal Sites to ensure no release of pollutants to
the environment. The activated sludge process is most adaptable for treat-
ing biodegradable organic wastes with influent BODg less than 3000 mg/1.
Due to the sensitivity of the process to surges in waste loads » however,
it is recommended that the process waste water be partially pretreated in
trickling filters to stabilize the reaction of the activated sludge
process to surces in loading.
Table 40 °534
The influent level of these metals must be below the maximum limits
indicated in Table 2.
16
-------
TABLE 3
ACTIVATED SLUDGE TREATMENT OF INDUSTRIAL
ORGANIC WASTES
800
Product and/or Flow In Out Real
1'ioce.a
-------
TABLE 3
ACTIVATED SLUDGE TREATMENT OF INDUSTRIAL
ORGANIC WASTES
(Continued)
I'rn.tnd .mdi or
I'mn *
tliylme •iid
ir-|tylrn») Owldci,
luiil*. Hor-
liol mi «j. llliy-
«.m> lilcmlnti.
tin rt
1 per •tin*)
. b-U
(Acid Huh U»t*>«>
Ci Hi -tliK.
(•.o-ort -at Inn of
•int. ni> «ntl Naph-
Ui, p. Alt.vlatl.in.
llciiii ic. Toluene,
AUo'ioll. Kctuiift.
iirs\ . ic Acid*
lilixU.u.
Avutylcne
VvKm M»nu( .-
A.UplC ALUI
Alk Ur«*nUi
BOD COO
Mow In Out Hem In Oul Rem lb ""S *'*y »ni§
(KD1 (»R/I) (mh/l) (7.) (mt/l) (mg/l) (.) Ih »ILSb R«n«I B.m-rk.
O.I) 1.9)0 10 99 7.970- 3,1 JO- 2)- 0,il Nont Lab Scale, cvtcndtd
a, MO 5.9)0 40 Mr.i(.o.i. hl,;h nan-
blodrfir«il«blc Tract lun
follawd by ttab pond*
01 i *vin in ait 9 VIA Uin HA n IB Nil 1 1 mlvturr of ac Id
1 I.D/0 111 9Ze3 2(91111 3W BU u to nn. i i miwiurr 01 ijciu
(MLVJS) vaih novLiI. Leb Scale
20 0 n- ro Elll phenol 0 1
0 JJ * m«/l
Lid oil 1 pp.
9) BJ l.U- P0f Nil Oil u-.eit ms nu-
3.0 m' trlent end neutiall-
1 rlnR •Hunt waete
diluted 2 1
18
-------
Nomenclature
a = sludge yield coefficient
a' = oxygen utilized for growth per unit substrate removed
b = rate of sludge a to-oxidation fraction per day, day'
b' = endogenous respiration rate, day"
k = biological reaction rate constant, 1/mg-hr
Q = rate of waste water flow, MGO
S - influent volatile suspended solids, mg/1
S = influent BOD, mg/1
a
S = soluble effluent BOD, mg/1
t = aeration time, hour
X = average mixed liquor volatile suspended solids, mg/1
AX = excess sludge yield, Ib/day
19
-------
7. REFERENCES
0283. Burd, R. S., comp. A study of sludge-handling and disposal.
Department of Interior Grant No. PH 86-66-32, Federal Water
Pollution Control Administration, May 1968. 369 p.
0534. Jones, H. R. Environmental control in the organic and petrochemical
industries. Park Ridge, New Jersey, Noyes Data Corporation, 1971.
264 p.
0669. Klein, L. River pollution control, v. 3. London, England,
Butter-worths and Company, 1966. Chapters 3 and 4. 484 p.
1266. Genetelli, E. J. Biological waste treatment. Ii± Water and water
pollution handbook. Ed. by Ciaccio. New York, Marcel Dekker,
Inc., 1971.
1267. McKinney, R. E., and W. J. O'Brien. Activated sludge-basic design
concepts. Journal of Water Pollution Control Federation, 40(11):
1,831-1,843, Nov. 1968.
1268. Eckenfelder, W. W. Comparative biological waste treatment design.
Journal of Sanitary Engineering Division, Proc. ASCE, 93(SA6):
157-170, Dec. 1967.
1283. Eckenfelder, W. W., and J. L. Barnard. Treatment-cost relationship
for industrial wastes. Chemical Engineering Progress, 67(9):76-85,
Sept. 1971.
1297. Vacker, D., C. H. Connel, and W. N. Wells. Phosphate removal by
activated sludge at San Antonio, Texas. J^ Water reuse, v. 63.
Series No. 78. Ed. by L. K. Cecil, Chemical Engineering Progress
Symposium Series. 1967. 275 p.
-------
AERATED LAGOONS
1. INTRODUCTION
An aerated lagoon is generally defined as a basin of significant depth
(usually 6 to 17 ft), in which organic waste stabilization is accomplished
by a dispersed biological growth system, and where oxygenation is provided
by mechanical or diffused aeration equipment.
There are two distinctly different types of aerated lagoons—the
aerobic lagoon and the aerobic-anaerobic or facultative lagoon. The
aerobic lagoon is designed with sufficient power input to create a
turbulence level high enough to maintain the solids in suspension.
Aerators must be selected not only to deliver the required oxygenation
capacity but also to induce eddies near the bottom of the basin to
prevent sludge deposition.
The aerobic-anaerobic or facultative lagoon is designed with only
sufficient power input to maintain turbulence levels adaquate to insure
uniform dissolved oxygen distribution throughout the basin. The major
portion of the solids are not in suspension but settle to the bottom of
the basin where they undergo anaerobic decomposition. A separate
sedimentation area or clarifier is somestimes included to produce a more
clarified effluent.
Aerated lagoons are a relatively new innovation in the waste water
treatment field. The aerated lagoon was originally developed as an
attempt to increase the BOD removal capacity of existing waste stabili-
zation ponds (oxidation ponds) by artificial aeration. More recently,
aerated lagoons are being used successfully to meet the needs of higher
degrees of treatment without the high construction costs associated
with an activated sludge plant.
21
-------
2. OPERATION PRINCIPLE
Biologically, the aerated lagoons resemble the activated sludge
processes In that the typical activated sludge floe ;(e.g., protozoa) is
invariably present and responsible for the biological removal of organic
matter present in the waste. In effect, with aerobic lagoon is analogus
to the extended activated sludge process with no sludge return. Due to the
dispersed nature of the culture, however, aerated lagoon systems do not
require close control over variabions in temperature, pH, and organic
loading.
Facultative lagoons employ aerobic biochemical oxidation near the
surface and anaerobic decomposition of organic matter in bottom of the
basin. In the absence of dissolved elemental oxygen at the bottom of the
pond, breakdown of organic molecules is normally considered as a two-
stage process. The complex organic molecules are first biologically
converted to low molecular weight organic fatty acids by the action of
acid forming facultative and anaerobic bacteria. The amounts of energy
released for growth during the initial conversions are relatively small
so that only a small portion of the waste is converted to new bacteria
cells. The small quantities of bacterial growth minimizes the problems
of biological sludge disposal.
Real waste stabilization occurs in the second stage of an anaerobic
process, where organic acids are converted by specific strictly anaerobic
micro-organisms into carbon dioxide and methane. The most important of
the methane forming bacteria live on acetic and propionic acids, and
require retention times of four days or longer for their growth.
When nitrogenous materials are also present in the waste water,
these may be converted to ammonia at the same time complex organic
compounds are converted into volatile acids, so that the volatile acids
would be neutralized upon formation. Under these circumstances, waste
stabilization occurs without methane formation.
22
-------
3. PROCESS DESIGN
An aerated lagoon is usually an earthen basin with some protection
near the bank for wave action caused by the aeration device. The aeration
equipment used to provide uniform dissolved oxygen dispersion and mixing
can either be a mechanical or diffused aerator. The effluent normally is
removed without sludge separation, although clarification can be provided
after the aeration basin.
The basis of design for the aerobic lagoon is quite different from
that of the facultative lagoon. For this reason, the process designs of
the two types of aerated lagoons will be discussed separately here.
Aerobic Lagoon
The design of the aerobic lagoon depends on the following variables:
rate of waste water flow, influent and specified soluble effluent BOD, influent
volatile suspended solids, the lagoon temperature and other experimentally
determined parameters characterizing the waste. The following process design
equations for the aerobic lagoon have been proposed by Eckenfelder.
(i) Aeration time or sludge age, t (hr)
"< s.ESo * •
-------
Volume of the aerobic lagoon is readily computed knowing the rate of
waste water flow. The effect of lagoon temperature is taken into account as:
T ?n
kT = k2Q 1.035 '^u (2)
where k, is the biological reaction rate constant at T°C.
(ii) Volatile suspended solids, X , (mg/1)
Sn + a(Sa - SJ
X - ° a e
X
.
V : * b t
~24
The total suspended solids in the lagoon, Xfl in (mg/1), will be
approximately equal to 1.25 Xv . Xfl is also the total suspended solids in
the lagoon effluent.
(iii) Total effluent BOD, ST (mg/1)
The BOD of the effluent from lagoon systems with a high solids carry-over
will be contributed by both the soluble organics remaining and the oxygen demand
exerted by biological solids in the effluent. The BOD due to the solids depend
on the sludge age and on the total suspended solids present in the effluent.
The fraction of solids which are actually biological f is related to the sludge
age t obtained from equation (1), (Figure 1) and the total effluent BOD ST is
given by:
ST = Se 4 f Xa «)
Facultative Lagoon
The effluent solid levels in facultative lagoons depend on the influent
BOD and suspended solids, the aeration power level, and the mixing intensity.
Field experience has indicated that the normal range of volatile solids
level X 1s from 50 to 100 mg/1. The design of the facultative lagoon is
based on the influent BOD and specified soluble effluent BOD. the biological
reaction rate constant k, and a knowledge of X .
24
-------
I/I
>>
(O
a>
o>
Biological Sludge from
Soluble Substrate
I I I | I I IX I I
0.5
f = mg BOD/mg SS
1.0
Figure 1. BOD Characteristics of Biological Sludges
-------
(1) Aeration time or sludge age, t (hr.)
The effect of the lower volatile suspended solids 1n the facultative
lagoon 1s that the sludge age 1s usually longer and a large aeration basin Is
required. On the other other hand, the lower volatile suspended solids level
also leads to higher degree of BOD removal and a lesser problem of solids
disposal. The effect of lagoon temperature 1s taken Into account by:
kT " k2Q 1.085 T"20 (6)
The large value of the temperature coefficient (1.085) indicates the sensitivity
of the facultative lagoon to temperature changes, and designs based on summer
lagoon temperatures cannot be depended on to produce high quality effluents during
the winter months.
(1i) Total Effluent BOO, ST
The total effluent BOD 1s again composed of both the soluble BOD
remaining and the BOD contributed by the suspended solids, and Is obtained
by using equation (4) and the sludge age—biological suspended solids
relationship (Figure 1).
The oxygen requirements of both the aerobic and the facultative lagoon
1s obtained from the following equation:
02 * 8.34 Q (a1 (Sft -Se) + b1 ^ Xv ) (7)
where 02 1s given in Ib/day, a' 1s the oxygen utilized for growth per unit
substrate removed 1n Ib Op per Ib BOD removed, and b1 is the endogenous
respiration rate in Ib 02 perlb VSS per day.
-------
In addition to providing sufficient oxygen and uniform oxygen dispersion,
minimum power levels to maintain solids in suspension are required for aerobic
lagoon systems. This means minimum basin bottom velocities of 0.4 to 0.5 ft
per second to prevent sludge deposition and a power level in the order of 30
to 100 HP per million gallons of basin volume for surface aerators. For
facultative lagoons, the power levels needed are much lower and only in the
order of 15 to 20 HP per million gallons of basin volume.
4. PROCESS ECONOMICS
The capital investment costs of an aerated lagoon system are determined
by the volume of the basin and the aeration equipment required, and range from
$0.2MM for a 5 million gallon aeration basin to $2 MM for a 100 million gallon
0534
aeration basin, when land values are not included. Approximate land
requirements for aerated lagoons are 8 to 16 acres per million gallon per day.
Operating costs for aerated lagoons are usually determined by the waste
water pumping cost and the power requirement for the aeration equipment.
Operating costs could range from $30 to $1,200 per million gallon of waste
treated.
As to be expected, the capital and operating costs of an aerated lagoon
are intermediate between those of an oxidation pond and an activated sludge
unit.
5. PROCESS MODIFICATIONS
An aerated lagoon is quite simple consisting of only an aeration device
in a basin. The two principal types of aerated lagoons--the aerobic lagoon
and the facultative lagoon have already been described in the previous sections.
Physical shapes of aerated lagoons are usually dictated by the size and shape
of the land available.
-------
6. PROCESS APPLICABILITY
Aerated lagoons have been used successfully as an economical means to
treat industrial wastes where high quality effluents are not required. In
most cases, aerated lagoons are being considered as an addition or modification
of existing waste stabilization ponds or as an interim treatment process
which can later be converted to an activated sludge system.
The main advantages of aerated lagoon systems include equalization of
shock loads due to the dispersed nature of the culture, no floe dispersion or
bulking problems and relatively low capital and operating costs. The main
disadvantages are maximum of 80 percent BOD removal and a turbid effluent,
and inactivity at low temperatures during winter months.
The removal of intiividual organic pollutants from waste water by aerated
lagoons have been studied, and the typical range of reduction (after 4 hr
of secondary settling) have been computed for some specific chemicals (Table '.).
It must be pointed out, however, that the system operated under conditions of
either oxygen starvation or cold-temperature inhibitation during most of the
period of data gathering, and thus the reduction indicated is influenced to
some degree by competition. It is conceivable that the biological reduction
for some of these chemicals could be considerably enhanced under more favorable
conditions.
Aerated lagoons have been shown to be useful in the treatment of
textile wastes, pulp and paper mill wastes, and cannery wastes, where BOD
reductions of 60 to 80 percent were obtained. More recently, aerated lagoon
systems have proven to be quite successful in treating synthetic organic wastes
and refinery-petrochemical wastes. A layout of the aerated stabilization
facility at the Institute Plant of Union Carbide is reproduced here, along
with some typical operating data (Figure 2). The facility treats a waste
flow of 5.3 MGD of fairly concentrated waste water (average BOD of 2100 mg/1)
from a plant manufacturing several hundred organic products and intermediates,
Including many hazardous waste materials under consideration in this program.
The key unit operations of the treatment facility are 4-hr preliminary
28
-------
TABLE 1
CHROMATOGRAPHICALLY MEASURED CHEMICAL REDUCTIONS AT UNION
CARBIDE INSTITUTE PLANT'S AERATED STABILIZATION FACILITY
Percent Reduction
of Compound Compounds in Category
0-10 Ammonia *
Acetophenone *
Ethylene Cyanohydrin
10-30 Acetone
Methyl ethyl pyridine
30-50 Ethylbutanol
2-Ethylhexanol
Methanol
Paraldehyde
50-70 Acrylic and butyric acids
Dioctylphthalate
Phenyl methyl carbinol
70-90 Acetaldehyde
Acetic Acid
Acrylonitrile
Butanol
Ethanol
Hexanol
Isopropanol
Isopropyl ether
Napthalene
Styrene
Toluene
90-100 Benzene
Benzoic Acid
Crotonaldehyde
Ethyl acetate
Ethyl acrylate
Ethyl benzene
2-Ethylhexyl acrylate
Heptane
i;
*Possible biological intermediates;
Note: Removals were limited seasonally be either oxygen availability
or cold temperature.
29
-------
INI I UI.NI-
*-
(JLI)IN(j
•\ hr)
L'TII.IM, .1 ).,)
j
-<
>\[.'<»IIUM | [il. til )
t
M'0« *
0
0
\
0
(0
1
1
t
0wi-)
(•)(:) (•)
\
1
1
1
C.O 0 0
000
1
t
1.1 UUU. blOHAU"
Average Operating Data
Variable
Temperature ( C)
DO (mg/1)
Power (bhp/1 ,000 gal )
Detention time (days)
BOD load (lb/day/1,000
cu ft.)
BOD removal eff.(%)
BOD removal (Ib/day)
COD load (lb/day/1,000
cu ft)
COD removal eff.(%)
COD removal (Ib/day)
Specific chemical load
(lb/day/1,000 cu ft.)
Specific chemical removal
eff. (%)
Specific chemical removal
(Ib/day)
Summer
35
nil
0.077
2.9
46
54
48,000
97
41
78,000
42
62
51 ,000
Winter
15
20
0.062
2.3
56
48
52,000
113
39
86,000
44
59
51 ,000
Figure 2. Aerated Stabilization System for Treating Synthetic
Organics at Union Carbide's Institute Plant
30
-------
holding, 66-hr aeration, and 4-hr secondary settling with provisions for the
parallel or series operation of the three aeration basins. Each aeration
basin has a 17-ft water depth, a 1.3-acre surface, and a holding capacity
of 4.8 million gallons.
A summary of reported data from the aerated lagoon treatment of
petrochemical and refinery waste water has been compiled (Table 2). Additional
treatment of effluents from aerated lagoons In waste stabilization ponds prior
to release have become quite common practice.
The limitations on the BOO removal efficiency will probably circumvent
the use of aerated lagoons as a single waste treatment unit where high quality
effluents are specified. At a National Disposal Site for handling hazardous
wastes, It is recommended that the Installation of aerated lagoons be considered
only under the following circumstances:
(1) as an interim treatment process that will be later converted to an
activated sludge unit or;
(2) as an "equalization tank" preceding other treatment units in a
multistage biological treatment facility or;
(3) as a "polishing pond" following other treatment units in a multi
stage biological treatment facility.
31
-------
TABLE 2
AERATED LAGOON TREATMENT OF PETROCHEMICAL WASTES*
w
2S3
Product and/or
Process
Refinery
Butadiene,
Butyl Rubber
Refinery,
Detergent
Alky late
BOD
Flow In Out Rem
(MGO) (mg/1) (rag/1 ) %
19.1 225 100 55
2.45 345 50- 71-
100 85
Organic
rnn Loading Nutri-
l-uu ... IK Rnn ents
In Out Rem 5 Reqd.
(mg/1) (mg/1) % Acre day
610 350 43 4,630 P04
855 150- 77- 6,300 PO.
200 83
Remarks
Followed by stab.
pond
temp - 32 C
30% COD is non-bio-
degradable
Lab Scale
Influent phenols
160 mg/1
Influent sul fides
150 mg/1
Lab Scale
Cyclohexane,
p-Xylene,
Benzene, Para-
ffinic Naphtha,
o-Xylene
Gasoline
Nylon Fibers
Chemicals for
Lubricating oils
0.51 100 25
75
400
Surface aeration,
Waste is extensively
pretreated.
Followed by pond
0.2 465 180
61 1,050
600
43
Table 42 of Reference 0534
-------
NOMENCLATURE
a = sludge yield coefficient
a1 = oxygen utilized for growth per unit substrate removed
b = rate of sludge auto oxidation fraction per day, day"
b1 = endogenous respiration rate, day"
f = fraction of suspended solids which are actually biological
k = biological reaction rate constant, 1/mg-hr
Q = rate of waste water flow, MGD
Sa = influent BOD, mg/1
a
Sg = soluble effluent BOD, mg/1
S = influent volatile suspended solids, mg/1
ST = total effluent BOD, mg/1
t = aeration time or sludge age, hr.
T = lagoon tempeature,°C
X = total suspended solids in lagoon, mg/1
a
X = volatile suspended solids in lagoon, mg/1
33
-------
7. REFERENCES
0534. Jones, H. R. Environmental control in the organic and petrochemical
industries. New Jersey, Noyes Data Corporation, Review No. 3.
1971. 264 p.
1268. Eckenfelder, W. W. Jr. Comparative biological waste treatment design.
Journal of the Sanitary Engineering Division. Proc. ASCE, 93 (SA6):
157-170, Dec. 1967.
1424. Bess, F. D., and R. A. Conway. Aerated stabilization of synthetic
organic chemical wastes. Journal of Water Pollution Control
Association, 38(6):939-956, Jun. 1966.
34
-------
OXIDATION PONDS
1. INTRODUCTION
Oxidation ponds are large shallow basins which depend on algal photo-
synthesis and natural surface reaeration for the oxygen used in treating
biodegradable wastes. In the treatment of industrial wastes, oxidation
ponds are best used to polish the effluents from other biological waste
treatment processes, although in some instances they have also been used
to treat entire plant wastes. Oxidation ponds have been widely used where
land is plentiful and climate conditions are favorable.
The optimum depth of oxidation ponds is about 4 ft, for aerobic
conditions to prevail. In heavily loaded ponds and in the bottom muds of
relatively deep ponds, a certain amount of waste stabilization will occur
anaerobically. Anaerobic treatment for industrial wastes, however, has
been generally unsatisfactory and usually limited to meat industry wastes.
Retention times in oxidation ponds range from a few days up to 3 months,
depending on the nature and composition of the waste, and the environmental
conditions present during treatment. Oxidation ponds require a minimum of
maintenance, but some control is desirable to ensure proper algae growth
throughout the depth of the basin. Multiple ponds in series are sometimes
used to minimize short-circuiting of flow.
2. OPERATION PRINCIPLE
Fundamentally, the oxidation pond utilizes bacteria to aerobically
stabilize the organic material present in the waste water, resulting in
the oxidation of carbon to carbon dioxide, hydrogen to water, and nitrogen
to ammonia. The oxygen for the bacteria is supplied by both air surface
transfer and the metabolism of algae in the pond. The cycle is completed
when the algae use the waste stabilization products carbon dioxide and
ammonia to synthesize new cells in the presence of sunlight, and thereby
-------
liberate oxygen as an end product to be used by the bacteria for oxidizing
the organic waste and synthesizing bacterial protoplasm.H.The important re-
lationship responsible for ogranic waste stabilization in the oxidation
pond is the relatively slow rate of degradation of cellular protoplasm as
compared to the rate of degradation of organic matter for the synthesis of
cellular protoplasm.
In addition to biological stabilization, other processes taking place
in oxidation ponds may include balancing of the acids, coagulation and sedi-
mentation, neutralization of any alkalinity by carbon dioxide from the air
and produced by oxidation of organic matter. The net effect of these is a
possible reduction in color and toxicity of the effluent.
3. PROCESS DESIGN
The performance of oxidation ponds is affected by the prevailing
environmental conditions such as air and waste temperature, wind velocity,
solar radiation, and types of micro-organisms responsible for waste
stabilization. The proper design for ponds must therefore involve the use
of established design equations supplemented with results from laboratory
model investigations simulating the local climate conditions as well as the
vshape of the treatment pond and the hydraulic flow patterns.
Thirumurthi 1273 suggests that the design of oxidation ponds should be
based on the Wehner-wi 1 kern equation, which is derived for a chemical reactor
with first order kinetics and a diffusivity constant:
1/2d (1)
(Ha)V/2d -
-------
in which d is the dimensionless dispersion number or diffusivity constant,
2
D is the axial dispersion coefficient in ft /hr, U is the fluid velocity in
ft/hr, L is the characteristic length of travel path of a typical particle
in the tank in ft, K is the first order BOD removal coefficient in day ,
and t is the detention time in day.
The dispersion number d incorporates the hydraulic mixing characteristics
of the system and has to be either evaluated experiment!ally or obtained
from previous studies under identical conditions. For most oxidation ponds,
the value of d seldom exceeds 1.0 and typically ranges from 0.1 to 0.3. For
the two extreme cases of plug flow and complete mixing systems, the values
of d are zero and infinity respectively.
The BOD removal coefficient k takes into account the organic loading,
the influent waste qualities and other biological factors, and is the other
parameter that has to be obtained from laboratory studies. The effect of
temperature can be corrected by the correlation:
KT= K2Q 1.0727'20 (2)
where KT is the BOD removal coefficient at temperature T( C).
For oxidation ponds with dispersion number d less than one, the
design equation (1) can be simplified by neglecting the second term in the
denominator to:
Se= 4a e(1"a//fcu (3)
Sa (1 + a)2
37
-------
The use of equation (1) can also be simplified with the help of Figure 1,
which was prepared by Thirumurthi for commonly expected BOD removal efficiencies
and the corresponding values of K and d.
The design equation presented here predicts that long rectangular ponds,
because of the lower dispersion number caused by the longer length of
travel path, will have a higher efficiency of BOD removal. However, excessively
long lengths are impractical because of maintenance problems and lack of
suitable land space.
In designing an oxidation pond, the procedure is to first obtain values
for the parameters K and d and then determine the detention period t from
either Figure 1 or equation (3) for th'j specified BOD removal efficiency.
The volume of the basin is then readily computed knowing the waste water
flow rate.
A minimum of 2 ft and maximum of 6 ft liquid depth is recommended for
oxidation ponds. Minimum depths are necessary to discourage weed growth,
whereas deep ponds do not permit aerobic conditions.
The actual organic loading for the oxidation pond can be obtained from
the waste water flow rate and influent BOD and the calculated basin volume.
Current design standards for oxidation ponds normally specify an organic
loading range from 20 to 100 Ib BODg per acre per day for handling raw wastes
and 10 Ib BOD per acre per day for polishing waste effluents from other
treatment units. The medium organic loading values in Ib BOD per acre per day
for industries employing extensive use of oxidation ponds are: meat and
poultry, 72; petroleum, 18; dairy, 22.
Detention times in oxidation ponds treating industrial wastes snow
extreme variations ranging from 1 day up to 90 days. The average retention
time is approximately 30 days.
38
-------
* 5-
-------
4. PROCESS ECONOMICS
The capital Investment costs of oxidation ponds are relatively low,
and range from $20,000 for a one acre pond to about $280,000 for a 30 acre
pond, °534 when land values are not included. Approximate land requirements
for oxidation ponds are 3-21 acreas/MGD* of waste water flow.
Operating costs for oxidation ponds are determined by the waste water
pumping and manpower costs, and are usually negligible contributions in a
multistage waste treatment facility. Annual operating costs are of the order
of 1,000 including analytical control.
5. PROCESS MODIFICATIONS
Oxidation ponds are simply shallow basins or lagoons where waste
stabilization occurs under aeroblic conditions when properly operated.
Physical shapes of the ponds are usually either rectangular or square,
but may also be dictated by the shape of the land available.
6. PROCESS APPLICABILITY
Oxidation ponds have been widely used in treating industrial waste
waters when a high degree of purification is not required. In a 1962
survey 1272 it was reported that more than 827 industrial premises in the
United States employed oxidation ponds in waste treatment. Amongst
those receiving treatment are wastes from meat and poultry packing stations,
canneries, dairies, chemical plants, oil refineries and petrochemical works,
paper and pulp mills, and textile mills. The typical BOD removal efficiencies
reported were from 60 to 90 percent.
Since temperature changes have significant effects on algal activity
and the effluent quality, designs of oxidation ponds should be based on
the average temperature during the winter months.
* MGD = million gallons per day
40
-------
More recently, oxidation ponds have proven to be quite successful in
treating steel mill wastes. The Granite City Steel Company uses a 2-cell
lagoon covering about 350 acres in an area fashioned out from an adjacent
lake for treating a total of approximately 38 MGD of waste water. The pond
system has a total capacity of 634 million gal. Detention times in the
oxidation ponds are about 10 days for waste water from the blast furnace
(for chemical contamination to disappear) and about 3-1/2 days for waste
water from the rest of the steel plant. The reductions in various kinds of
pollutants from the steel plant waste water by the oxidation pond system
varies (Table 1) but generally indicates that the water quality discharged
from the ponds is better in nearly every characteristic than that being
withdrawn from the Mississippi River for the cooling and cleaning needs of
the plant. The significant reductions in phenol and cyanide concentrations
are particularly notable.
At a National Disposal Site for treating hazardous wastes, it is
recommended that oxidation ponds be considered for installation as a
polishing stage for effluents from other biological waste treatment pro-
cesses. Because of the complexity and possible toxicity of the incinerator
scrubber waste liquors, spent cleaning solution wastes and spill control
system wastes which possibly will be handled at a National Disposal Site-,
a multistage biological treatment system involving trickling filter -
activated sludge - oxidation ponds will, probably be required on effluent
water streams. Such biological treatment would follow the removal of
toxic inorganic components from the streams.
41
-------
TABLE 1
OXIDATION POND INFLUENT AND EFFLUENT
QUALITY AT GRANITE CITY STEEL PLANT
Oxidation Pond
Influent and
Effluent
Settleable
Solids
mg/1
Suspended
Solids
mg/1
COD
ppm
BOD
ppm
Grease
ppm
NH,
ppm
Phenol
ppb
Cyanides
ppb
pH
Mater Withdrawn
from Mississippi
River
0.60
62
24
10
7.84
Steelworks Effluent 0.10
41
68 8
0.4 12
7.10
Blast Furnace
Effluent
0.10
98
93 10
14 1300
1800
7.58
Oxidation Pond
Effluent
0.07
18
50 5
8.01
-------
NOMENCLATURE
d = dispersion number or diffusivity coefficient
-1
2
D = axial dispersion coefficient, ft /hr
K = first order BOD removal coefficient, day
L * characteristic length of travel path of a typical particle
in the pond, ft
S = influent BOD, mg/J
a
Sg = effluent BOD, mg/l
T = temperature, °C
t = detention time, day
U = fluid velocity, ft/hr
43
-------
REFERENCES
0534. Jones. H. R. Environmental control in the organic and petrochemical
industries. Review No. 3. Park Ridge, New Jersey, Noyes Data
Corporation, 1971. 264 p.
1272. Forges, R. Industrial waste stabilization ponds in the United States.
Journal of the Water Pollution Control Federation, 35(4):456-468,
Apr. 1963.
1273. Thirumurthi, D. Design principles of waste stabilization ponds.
Proceedings of the American Society of Civil Engineers. Journal
of the Sanitary Engineering Division. 95(SAL):311-330. Apr. 1969.
1426. Cairns, D. F. Stabilization lagoons successfully treat steel mill
wastes. Journal of the Water Pollution Control Federation,
38(10).1,645-1,655, Oct. 1966.
44
-------
TRICKLING FILTERS
1. INTRODUCTION
The trickling filter consists essentially.of a rectangular or
circular bed of rocks or synthetic media,*usually from 3 to 15 ft deep,
over which waste water is distributed. The medium serves as a habital
for bacteria growth, and organic material present in the waste is removed
by the biological slime as the waste water flows through the filter.
Proper hydraulic loading is maintained by recirculation of a portion of the
filtered-effluent.
Because of the short residence time, BOD removal by trickling filters
is usually far from complete. For. the same reason, however, toxic substances
discharged into the system are less likely to cause complete disruption, and
shock loadings have a less pronounced effect.
Trickling filters for industrial wastes are primarily used to level
out and reduce organic loads in a multistage biological treatment system.
A schematic flow diagram of a typical trickling filter system is provided
(Figure 1).
2. OPERATION PRINCIPLE
The trickling filter process utilizes bacterial growths adhered to
the bed packing for the biochemical oxidation of organic matter present
in the waste water. A portion of the remaining organic material, however,
is converted to new cells after being absorbed through a thin film into
the biological slime on the surface of the filter medium. With the high
rate type filter used for the treatment of industrial wastes, the hydraulic
loading is such that a part of the attached growth is continuously sloughed
off from the filter and a bacterial population equilibrium is often
automatically maintained. The operation of the trickling filters is thus
45
-------
Recirculation
Industrial
&
CT5
Waste
Primary
Sedimentation
Trickling
Filter
Secondary
Sedimentation
To Other
•*•
Biotreatment
Systems
Sludge
Sludge
Figure 1. Trickling Filter Flow Diagram
-------
very similar to that of the packed tower in chemical engineering unit
operations, especially because different degrees of waste purification
are obtained at different levels within the filter bed.
Proper functioning of trickling filters requires aerobic conditions in
the bed and sufficient void space in the packing must be provided. Under
normal conditions in properly designed filters, adequate air flow is usually
produced by the temperature difference between the air and waste water.
3. PROCESS DESIGN
Trickling fiters are generally 3 to 15 ft in depth and consist of a
filter media retained by a wall over which waste water is evenly applied
by either rotary distribution arms or fixed spray nozzles. An underdrain
system is used to remove effluent from the filter and to provide air
passage for the maintenance of aerobic conditions.
The filter medium provides surfaces for biological slime growth. The
material constituting the medium should be hard, free from dust, resistant
to the action of the waste treated, and a high surface area/volume ratio.
The filter should have sufficient void space to permit air circulation.
For the treatment of industrial waste, rock and synthetic plastic have
generally been used as the filter material. The synthetic media usually
consist of flat or corrugated sheets or rigid materials assembled in units
resembling a honeycomb, and are primarily used to treat wastes with higher
hydraulic and organic loadings. Because of the controlled uniformity and
void space, and a greater surface area/volume ratio, the plastic filter
media provide greater BOD removal efficiency and are free from plugging
and ventilation problems. The lightness of the plastic media also enables
stacking to a greater height and trickling filter designs of depth up to
40 ft.
47
-------
Trickling filters are normally designed according to the applied
hydraulic and organic loadings. The hydraulic load is the total volume
of waste water applied to the filter per day per unit area of surface
2
area, and ranges from 25 to 100 gpd/ft for high rate filters. The organic
load is the pounds of 5 day BOD applied to the filter per day per unit
volume of filter medium, and ranges from 5 to 25 Ib BOD/day/1000 ft for
standard rate filters and 25 to 300 Ib BOD/day/1000 ft3 for high rate filters.
Recirculation is often utilized for high rate trickling filters to
increase the filter efficiency and to dilute the initial high organic
concentration in the waste. Recycle ratios as high as 40 are sometimes
required.
The design of the trickling filter depends on the following variables:
hydraulic rate of flow, filter depth, recirculation ratio, influent and
1276
specified effluent BOD, and also waste water temperature. Eckenfelder
suggests the following formula for the removal of BOD from industrial
wastes:
Le= 1 (1)
L , + 2.5 D°'67
t-f)
0.5
where Lg is the BOD of filter effluent in mg/1
L is the BOD actually applied to the filter (including recirculation)
in mg/1
D is the depth of filter media in feet
Q is the plant influent rate in MGD
A is the filter surface area in acres
48
-------
Eckenf elder also states that the effect of reclrculation is to dilute
the Influent BOD (not including recirculation), L., by the formula:
L * R L
L
L
o 1 + R
where R is the recirculation ratio
(2)
The effect of waste water temperature is taken into account by
Eckenf elder as:
ET = E2Q 1.0351'20 (3)
where ET is the filter BOD removal efficiency at temperature T( C) and
E20 is the efficiency at 20 C.
The volume of the filter media is responsible for the major portion
of the construction cost of the trickling filter systems. Using equations
(1) and (2), the volume V in units of 1000 ft is:
v = 7-°D^ (i +V (TIT)
(4)
It is seen from the Eckenfelder formula that the filter media volume varies
2 ?
directly with the flow rate Q, is a function of efficiency as E /(1-E)
0 33
(influent BOD is not directly included as a parameter) and depth as 1/D .
2
It is also a function of recirculation as -1/(1 + R) , which causes the
volume to decrease with greater recirculation ratios. For practical purposes,
however, further decrease in volume should not be considered in the de-
sign at recycle ratios about 4 or 5. Furthermore, designing for minimum
filter medium volume by increasing bed depth and recirculation would also
significantly increase the associated cost of pumping. The use of the
Eckenfelder formulas described here for trickling filter process design
usually gives better predicted values in filter performance than other
design formulas..
49
-------
4. PROCESS ECONOMICS
The primary factors determining the capital cost of trickling filter
systems include the waste flow rate and the organic loading. Depending
on the treatment capacity (volume of the filter media), construction costs
for trickling filters usually range from $0.2 MM to $3 MM.
Operating costs for trickling filter systems are normally due to
pumping and distribution of waste water. For the treatment of chemical
wastes, operating costs have ranged from $700 to $2000 per MG waste or
$50 to $4000 ner 1000 Ib DOU removed.
5. PROCESS MODIFICATIONS
Modifications of the trickling filter systems generally involve
1 Pfifi
differences in recirculation or distribution schemes, and are described
very briefly here:
Biofilter
The biofilter utilizes a shallow trickling filter with recirculation
and high rate of application. As indicated in the flow scheme (Figure 1),
the effluent from either the filter or secondary sedimentation is recircu-.
lated back to the primary sedimentation tank.
Accel ofilter
In the accelofilter system (Figure 2), a portion of the filter effluent
is directly recirculated back to the trickling filter, and places no addi-
tional burden on the sedimentation tanks.
50
-------
Aerofliter
In the aerofilter system (Figure 3), a portion of the effluent from
secondary sedimentation is recirculated back to the filter, where the
waste is applied continuously by high speed disc distributors. Normal
p
hydraulic loadings are greater than 300 gpd/ft of surface area.
6. PROCESS APPLICABILITY
Trickling filters have been extensively used in the treatment of in-
dustrial wastes due to their ability to cope with shock loading by main-
taining proper hydraulic loading with recirculation. Industrial wastes
that have been successfully handled by the process include the following:
acetaldehyde, acetic acid, acetone, acrolein, alcohols, benzene, butadiene,
chlorinated hydrocarbons, cyanides, epichlorohydrin, formaldehyde, formic
acid, ketones, monoethanolamine, phenolics, propylene dichloride, terpenes,
ammonia, ammonium nitrate, nylon and nylon chemical intermediates, resins,
^and rocket fuels. A summary of the types of petrochemical and organic
wastes treated by the trickling filter process along with the effluent
qualities obtained, has been presented by Jones (Table 1). The BOD removal
efficiencies range from a low of 30 percent to a high of 98 percent.
In general, the materials that are toxic to the bacteria in the
activated sludge have the same effects on the biological slimes in the
trickling filter. The maximum limits for selected waste water char-
acteristics could be obtained from Table 2 of the Activated Sludge Process
Evaluation.
Below threshold toxic limits of the bacteria, metal content in the
waste water could be reduced by the trickling filter process. The approximate
amounts removed by the process are: Cr, 30 percent; Ni. 40 percent; Zn, 30
percent; Pb, 30 percent; Cu, 20 percent, . These metals are concentrated
in the biological sludges obtained from the secondary sedimentation tank
which must be disposed of later.
51
-------
Recirculation
Industrial
Waste
Primary
Sedimentation
Trickling
Filter
Secondary
Sedimentation
Sludge
^To Other
Biotreatment
Systems
Sludge
Figure 2. Accelofliter
Recirculation
Industrial
Waste
Primary
Sedimentation
Trickling
Filter
Secondary
Sedimentation
To Other
Biotreatment
System
Sludge
SIudge
Fic-ure 3. Aerofliter
52
-------
TABLE 1.
TRICKLING FILTER TREATMENT,. OF
INDUSTRIAL ORGANIC WASTES1"
BOO
Product and/or Flow In Out Ken
Process (K.D) («s/l> (Bg/1) (I)
Phenol, Salicylic 2.19 190 58 69.9
Acid, Rubber Chen.,
Aspirin, Phenecetln, 2.59 58 M 41.5
Phthalle Anhydride
Plastics, Amines, 1.06 1,960 37 98.1
fcnzyaea
Cthylene, Propy- 0.63 170 85 50
lene. Butadiene.
Benzene, Poly-
ethylene, Fuel Oil
Aliphatic Acids, 0.57- 1,100- 23- 37.
Esters, Alcohols. 0/86 2,300 470 99
Aromstlcl. Anlnes,
Inorganic Salta
Cthylene. Propy- 0.43 1.300
Icne, Butadiene,
Benzene, Naph-
thalene, Phenol,
Acrylonltrlle,
Soft Detergent
Bases. Resins
Pentseryihrltol 0.118 5,080- 225- 95-
Vaste contain) 5,800 232* 96*
Fornsldehyde,
Sodium Fonsste.
Nethinol, Pent-
aerthrlcol
Realns-Foraaum, 0.17 81.6
Amlnoplasts,
Phenol -Forael, 0.03 89.3
Epoay Reslne,
lextlle Aus.
Acrylic Fibers 0 32 13 30-
10
Synthetic Resins- 93-
Plienol , lormslde- 98
hyde. Fatty Acids,
Rilhsllc Acid, Halelc
Arid, Llyrerol.
Penlaerythrltol,
Uame i onlelne
Airylales .
Inlill-ltor oils.
AUulioli, Lmers,
Hib°'i
Organic Acids
COO Orgsnlc
la Out Rem "* M»i '«•»
(Bg/1) (Bg/l) (t) V 1,000 ft1
40.9
11.8
1,660 130 91.1
400 200 30 89
42.1-
82
ents
None
None
None
None
(Both (liters
combined)
1,300 **0 *°- 140
70
lit stage 69
11.7
14.6
49 30- 50
70 84
1st suite 85
(as Phenol)
2ml si SRC II
18 J «pd/ll'
51-
79
mi
]
PO
NH.
P04
Yes
None
None
NHj
PO,
6-
None
None
Rcoarhs
Rock Media, reclrc.
ratio 2.84 1
Rock Media, treats
effluent from above
filler, effluent
to act sludge
2 filters, followed
Plastic filler media
lol lowed by act
sludge
phenol removal • 951
Influent diluted 2 1
M/coollng water
pll Adjusted prior to
treatment. 2 filters In
series. Recyile on 1st
state Is 16-21 I
Sour Waters,
Rock Media
2 filters In scrlei
followed by act.
sludge, recycle
1,0-1 on prim filter,
11-1 second
Both filters treat
act. sludge effluent.
Blast furnace slag
Waste contains
phenol, lormaldehyde.
netltanol
Waste contains
ucrylonltrlli! and
ilnt plastic filter
mrdls
Plaltlc media, 2-
Inf lucnt
I'lienol . 4.1UO mg/l
Fomalilrhyue •
2 000 mg/l
Fatty •clJi-300
Phthalle andulelc
eclds- I.OOOmg/l
Iff phenol. 1.5 eaj/l
Origins! loedlng wee
lover vslue, loading
Increased w/o any
•Jvcm i fleets
15, QUO Ib noil, re-
moval pir day
Entire treatment system
Table 41°534
53
-------
The complexity of Industrial organic wastes and the stringent spec-
ification of effluent qualities have at times circumvented the use of trick-
ling systems as a Dingle stage treatment unit. At a National Disposal Site
for the disposal of hazardous wastes, it 1s recommended that trickling filters
be Installed as a roughing device and the first stage In a multistage bio-
logical treatment facility. For example, systems can be designed with high
rate trickling filters in series with the activated sludge process, to take
advantage of the trickling fiter's ability to handle shock loads and the
ability of the activated sludge process to produce an effluent of high
quality, and thus eliminate some of the shortcomings of each.
One distinct disadvantage of the Installation of trickling filters at
National Disposal Sites 1s their unsuitabllity for treating waste waters
that are discharged Intermittently or infrequently as might be the cases
with the effluents from batch operations. Under these conditions, the
filters would not function efficiently as they are designed to run
continuously. A possible solution to this problem is the parallel use
of trickling filters and aerated lagoons as the first stage in a multistage
biological treatment unit, so that occasional overflows and irregular dis-
charges could be directed to the lagoons, which also serve as equilization
tanks for the second stage biological treatment process.
-------
NOMENCLATURE
A = filter surface area, acres
D = depth of filter media, ft
ET = filter BOD removal efficiency at T C
Lg = BOD of filter effluent, mg/1
Li = BOD of filter influent (not including recirculation), mg/1
L = BOD actually applied to filter (including recirculation), mg/1
Q = rate of waste water flow, MGD
R = reciruclation ratio
T = temperature, C
V = volume of the filter media, 1000 ft3
55
-------
REFERENCES
0314. Jones. H. R. Environmental control in the organic and petrochemical
industries. Park Ridge, New Jersey, Noyes Data Corporation, 1971.
264 p.
0669. Klein, L. River pollution control, v. 3. London, England,
Butterworths and Company, 1966. Chapters 3 and 4.
1270. Hanumanulu, V. Performance of deep trickling filters by five methods.
Journal of the Water Pollution Control Federation. Part 3, 42(8):
1,446-1,457, Aug. 1970.
1276. Eckenfelder, Jr., W. W. Trickling filtration design and performance.
v. 128. Part 3. Transactions, American Society of Chemical Engineers',
1963. p. 371-398.
56
-------
HAZARDOUS WASTE DISPOSAL PROCESS DESCRIPTOR-DIALYSIS
1 . INTRODUCTION
Dialysis Is a process by which various substances In solution having
widely different molecular weights may be separated by solute diffusion
1 A R^
through semi -permeable membranes. The driving force is the differ-
ence in chemical activity of the transferred species on the two sides of
the membrane. Until recently, dialysis was of limited applicability be-
cause the parchment membranes used had very poor acid resistance, but the
development of microporous polyvinyl chloride membranes has extended its
range of application from the recovery of caustic soda in viscose rayon
production to the recovery of mineral acids from aqueous solutions and the
1453
recovery of colloidal organics in the manufacture of Pharmaceuticals.
The process is basically passive with feed solution moving into a two-
dimensional cellular array and then separating into waste and product
streams, and is used for caustic recovery from a caustic-hemi cellulose feed
in the rayon industry (Figure 1)
2. OPERATING PRINCIPLES
Dialysis may be described in terms of a diffusion process. For steady-
state mass transfer through a membrane with countercurrent flow of the solu-
tions,™2
H = V0AtClm 0)
where W = mass transferred in g/min, V. = overall dialysis coefficient in
2
cm/mi n, A = membrane area in cm, and A C-jm = logarithmic mean concentra-
tion gradient across the membrane, i.e.,
57
-------
WEAK CAUSTIC+ HEMICELLULOSE
WASTE SOLUTION
MEMBRANE
WATER
MEMBRANE
STRONG CAUSTIC + HEMICELLULOSE
FEED SOLUTION
FRAME
CAUSTIC SOLUTION
PRODUCT
Figure 1. Plate-and-Frame Dlalyzer
1454
58
-------
where AC1 and AC2 are the respective inlet and outlet concentration differ-
ences in g/cc.
1 Jl Q O
This relationship is only approximately correct, although rea-
sonably accurate estimates can be made by adjustment of the coefficient V .
Complications are introduced by the two-dimensional nature of the concen-
tration gradient; not only is there a gradient in the direction of diffusion
through the membrane, but there is also a gradient in the direction parallel
to the membrane as the feed solution is progressively depleted of solute in
its traverse from one corner of the cell to the other. The passage of sol-
vent through the membrane, particularly when concentrated solutions are
involved, decreases the efficiency of the process and entails a modification
of Eq. 1 to reduce the area A to an effective A'.1452'1454
As dialysis has gained wider attention and investigative interest, the
simple concept of a porous membrane which admits or rejects species on the
basis of their molecular dimensions has given way to complex, poorly under-
stood ideas involving membrane-species interactions, electrostatic effects,
and microelastic effects in the neighborhoods of the individual pores. The
science is still relatively primitive, but these principles have been uti-
lized in the recent art.
3. PROCESS DESIGN
As with reverse osmosis, the characteristics of the dialysis membrane
are the determining factors in the application of dialysis to any parti-
cular operation. The membrane must of course be chemically resistant to
the materials in the feed stream. Parchment has long been used success-
fully for neutral and basic streams, and new polyvinyl chloride membranes
have been used successfully for acid streams.
The chemical and physical properties desired for a dialysis membrane
1433
material are summarized succintly by Kirk-Othmer:
"Selection of the membrane polymer is a key factor in deter-
mining diffusion properties. Ability to swell in the dialysis
solvent is important if high rate and low selectivity among
59
-------
small molecules are sought, since swelling generally increases
greatly the mobilities of the diffusing species. Solubility
and chemical behavior in the dialysis solvent of the units
of the polymer chain are strong factors affecting swelling.
For the selective, rapid transport of a particular species,
the membrane units should be chosen to be good and unique
solvents for that species. All practical membranes must
not only possess good diffusional properties but must also
remain insoluble in the dialysis media, and must maintain
sufficient mechanical strength to withstand handling and
momentary or sustained pressure differences, depending on
the process. The compromise between strength and diffu-
sional properties may be achieved in many ways. Membranes
may be cast on a fabric of open weave, or thin, strong
fibers may be combined with the casting solution. Strength
compromises may be achieved in homogeneous membranes by
the use of polymer mixtures, copolymers of the random or
alternating type, or the less familiar block copolymers
(Figure 2). The latter system is made up of polymer
chains comprised of two monomer units with each monomer
unit attached to its own type many times.
Membranes developed to date generally have a maximum operating tem-
1453
perature of 120 F.
To maximize the amount of material transferred in a unit, the overall
dialysis coefficient, VQ must be as large as possible. For purposes of
analysis, VQ may be conveniently broken down into two terms,1433'145
_ ._ _
Vo Vl V2
(2)
where V1 = dialysis coefficient in the liquid film adjacent to the membrane
and is approximately equal to 1000 times the normal diffusion coefficient
for the species in the solvent medium, and Vg = membrane dialysis coefficient,
which may be approximated by the expression PS where D is the
60
-------
• B-B
-A-A-A
(a)
(b)
(c)
Figure 2. (a) Mixture of two homopolymers. (b) Random copolymer.
(c) Block copolymer.
.08
.06
MEMBRANE
DIALYSIS
COEFFICIENT,
CM
MIN.
.02
LiOH-
o^5 No OH
,Ba(OH)
CM-
DEXTROSE
SUCROSE
RAFFINOSE
NoCI
LiCI
BaCI2
CaCI,
.10
.20
.30
.40
DIFFUSION COEFFICIENT. ,n.D CM
WET THICKNESS OU z ' MIN.
Figure 3. Experimental Membrane Dialysis Rates 1454
.50
61
-------
diffusion coefficient, F is a drag factor which accounts for the interfer-
ence of the membrane with the normal diffusive flow, V is the volume
fraction of the membrane occupied by pores, h is the tortuosity (ratio of
capillary length to membrane wet thickness), and Z is the membrane wet
Y«rj>
thickness. Perry presents values for these parameters for a variety
of commercially available membranes. These will not be pursued further
here except to note that they are largely experimentally derived and
attempts have been made to predict V from a minimum number of para-
meters with fair success. Figure 3 presents experimentally-determined
values for V for a variety of bases, salts, and sugars for a typical
o 1453
dialysis membrane, viz., DuPont 450 P.O. cellophase. Table 1 presents
diffusion coefficients for several common acids, bases, and salts. It
should be noted that widely separate coefficients, e.g., H2S04 vs CuS04>
imply disparate values of VQ and the possibility of efficient separation
by dialysis.
This analysis indicates that, for maximum material transfer, the inter-
membrane spacing and membrane thickness are both minimized, so that, with
considerations of surface area and concentration gradients as presented in
Eq. 1, the ideal dialysis unit has a flow path of great width, small inter-
membrane spacing and membrane thickness, and short length. However, practi-
cal design considerations work against these extremes, with the result that
stream convection velocities parallel to the membranes may be as low as 0.1
cm/mi n.
The required membrane area for a given throughput and recovery is in-
creased as a function of the solute concentration in the feed stream and
the ration of DH2°/Dsolute' F19ure *1452 Presents a series of parametric
curves showing the approximate corrections required.
Several limitations of variable severity are encountered in the design
of any dialysis plant. Since dialysis is essentially a passive process
which draws on no outside energy source, separations can be painfully slow -
and required area-to-volume ratios may prove difficult to achieve technically
and difficult to justify economically. Furthermore, if the chemical poten-
tials of the diffusing species on either side of the membrane do not remain
62
-------
TABLE 1
REPRESENTATIVE DIFFUSION COEFFICIENTS
ACIDS
Hydrochlorie 2.6
Nitric 2.6
Sulfuric 1.7
Acetic 0.88
Phosphoric 0.9
ALKALIES
Sodium Hydroxide 1.5
Potassium Hydroxide 2.3
SALTS
Sodium Chloride 0.35
Sodium Sulfate 0.77
Copper Sulfate 0.58
Zinc Sulfate 0.59
Nickel Chloride 1.1
63
-------
a
•*
<
LU
<
LU
z
CO
UJ
40,000
20,000
10,000
6,000
4,000
2,000
1,000
600
400
200
ino
/
r
^, •
X
— ^
^m «•
^•••B
A
/
/
^^ ^
^
*~ •*"*
^-^
-A.
E)
•Q=4
^
-~'
•^^
>WOX\
:ACTA/
0
^--
^^
--**"
•s*^
MATE /
ETHOC
^^
^-^^"^
=^^=
• *
i/lETHC
1
Q-
10
Q=4
D _
0 20 40 60
SOLUTE CONCENTRATION, WT. %
0_ DIFFUSION COEFFICIENT OF WATER
" DIFFUSION COEFFICIENT OF SOLUTE
Figure 4. Calculated membrane area to recover 90% of
at a feed rate of 100 g/min. T45Z
solute
64
-------
disparate, the driving force for separation will be severely reduced. This
happens when there Is a low Initial concentration In the feed stream or a
required low final concentration in the waste or product stream. Low Is
generally under 0.1 percent. (Figure 5 relates the required membrane
area, dialysis coefficient, and feed stream flowrate to the desired recovery
for a series of feed stream/product stream ratios. For a desired recovery
of 75 percent and a feed stream/product stream ratio of 2/3, AV /q = 2.3,
so that for q = 10,000 gpd and VQ = 0.04 cm/min, A = 706 ft2.
In addition, if the dialysis coefficients of two solutes are close to-
gether, the efficiency of their separation will be very poor. This occurs
whenever molecular dimensions and molecular weights are close together.
1454
Figures 6 and 7 are parametric representations of experimental results
for attempted separations of a NapSO.-sucrose solution. Here both molecules
are under 10 A in diameter. Note that as the pore size decreases in the
direction of the size of the smaller Na«SO. molecule, the ratio
V0(Na2S04)
VQ(sucrose)
increases, as does the separation efficiency. But, for a given pore size,
however, good separation efficiency results in a lower total recovery of
the Na2S04.
4. PROCESS ECONOMICS
It is difficult to arrive at valid cost estimates for large-scale
dialysis systems because the problems of slow separation and large required
surface-to-volume ratios have hitherto hindered large-scale commercial
exploitation of dialysis. It is to be expected that capital investment
would be high for these reasons, but that operating requirements would be
1453
low, since no energy input is required save for low pumping speed. Dvorin
is optimistic that the newer membranes will prove to be essentially main-
1 A CO
tenance-free, but no actual field data was found. Perry reports a
selection of dialysis units available in 1960. These are continuous counter-
current units, consisting of a series of alternate water and solution cells
-------
10
8
6
AV,
q
i.o
0.8
0.6
0.4
0.2
0.1
RATIO OF FEED RATE TO WATER RATE
2/3
/1/2
-'V3
0 20 40 60 80 100
PRODUCT RECOVERED IN DIALYSIS , %
Figure 5. Membrane area required vs product recovery at various
values of feed-to-water flow rates; A = area sq. cm;
U0 » overall dialysis coefficient, cm/min; q = flow
rate of feed solution, cu. cm/min.145Z
66
-------
RECOVERED
uu
80
60
40
20
0
1
\
\
\
SODIM SUI
SUCROSE
\
X
50A 30
PORE SIZE
o
A
S
M.W.
.FATE 142
342
-
.0 1.2 1.4 1.6 1.
/Na,SO,i\ /Na0SOA
1*41 • 1241
SUCROSE/ REMOVED \SUCROSE/FEED
Figure 6. Dialytic Fractionation
1454
67
-------
No2SO4
Vo
• SUCROSE
£.3
2.1
1.9
1.7
1.5
1
t
V
SODIUM SULF
SUCROSE
^
:ATE
M .W . "
142
342
0 20 30 40 50 6(
MEMBRANE PORE DIAMETER A
Figure 7. Dialysis Rate vs Pore Diameter
1454
68
-------
connected in parallel and separated by the membranes. The Brosites Machine
2
Company supplies a unit containing 850 ft of membrane for $10,900 +
installation, and the Graver Water Conditioning Company, New York, supplies
2
a unit containing 1,000 ft of membrane for $9,000 + installation. Laboratory-
size units are also available.
5. PROCESS MODIFICATIONS
Process modifications are concerned primarily with attempts to increase
the speed of separation and the concentration differences between the feed
and product streams. Dialysis may be used in conjunction with energy-input
processes. Various processes such as boiling and condensing in distilla-
tion, solvent processing in extraction, and chemical regeneration in ion
exchange have been considered in attempts to increase energy differences
in the respective streams and thereby speed the process. Concentration
differences may be promoted by the recycling of the waste stream. Figure 8
shows a process flow sheet for an acid recovery unit incorporating recycling
1433
in a copper refinery.
6. PROCESS APPLICABILITY
The unique characteristic of dialysis is its passivity and lack of
reliance on any external energy input. This is of advantage when time is
not a limiting factor or when it is desired to separate species in solutions
of high concentration, e.g., 20 to 50 percent, where the energy require-
ments of other processes such as electrodialysis are very high.
The oldest continuing commercial use to dialysis is in the textile
industry where feed streams containing 17 to 20 percent NaOH contaminated
with hemicellulose are routinely recovered to 9 to 10 percent NaOH virtually
free of the hemicellulose. 3 An 800 ft2 unit with 0.5 inch intermembrane
spacing will typically handle 3600 gpd with a fluid residence time of 2 hr
and a total caustic recovery of 200 tons/yr.
69
-------
TANK ROOM LIQUOR BLEEDOFF
18% H SO
WATER *
Cu CRYSTALS
'ORATOR
COOLER
DIALYZER
WATER
I +DI
DIALYSATE <• 1 '—*DIFFUSATE
TO1 *
NICKEL TO
SULFATE JANK
PLANT R°OM
Figure 8. Process Flow sheet for acid recovery from
salt solution with recycling 1433
70
-------
Dialysis is particularly applicable when concentrations are high and
dialysis coefficients are widely disparate (see Section 3 Process Design).
Dvorin reports success In treating concentrated acia waste streams
from metal finishing shops and refineries. Good recoveries are claimed for
chromic, hydrochloric, nitric, hydrofluoric, and sulfuric acids. Results
are quoted in Table 2 for an acid recovery unit in a copper refinery which
produced essentially Ni-free H2S04. It is possible to improve efficiencies
by recycling the waste stream. A representative process flow sheet in-
14*53
corporating recycling is provided (Figure 8).
Dialysis is a suitable means of separation for any materials on the
hazardous materials list which form aqueous solutions, e.g., inorganic salts
such as ammonium chrornate, or acids and bases such as phosphoric acid and
sodium hydroxide. It is particularly suitable where high solute concentra-
tions are involved, since reverse osmosis is then inapplicable and electro-
dialysis requires large energy inputs and concomitant high cost. Its in-
herent passivity, however, makes it inefficient where concentrations of
feed or product solutions are much below 0.1 percent. With regard to acids
and bases, dialysis does not require neutralization prior to treatment, as
reverse osmosis does. But no dialysis membranes presently available are
suitable for both acids and bases.
With regard to National Disposal Sites, dialysis could most effectively
be used for the further concentration of concentrated waste streams of
extreme pH. The waste would then be stored or recycled to the supplier.
-------
TABLE 2
RECOVERY RESULTS IN A COPPER REFINERY1453
Gal/Hr
H2S04 g/1
CuS04-Cu g/1
NiS04-Ni g/1
Liquor
Feed
400
300
30
45
Water
Feed
400
0
0
0
Dialysate
Effluent
420
100
26
43
Diffusate
Effluent
380
200
2
Trace
-------
REFERENCES
1433. Kirk-Othmer. Encyclopedia of chemical technology. 2 d ed. 22 v.
and supplement. New York, Interscience Publishers, 1963-1971.
1452. Perry, J. H., at al. Chemical engineers handbook, 4th ed. 1963.
1453. Dvorin, R. Dialysis for solution treatment in the metal finishing
industry. Metal Finishing, 57(4), 52-54, 62, 1959.
1454. Lane, J. A., J. VI. Riggle. Dialysis. Chemicaj^ Engineering,
Progress Symposium Series 24, 55, 127-143, 1959.
73
-------
HAZARDOUS WASTE PROCESS DESCRIPTOR—ELECTRODIALYSIS
1. INTRODUCTION
Electrodialysis is similar to dialysis in that dissolved solids are sep-
arated from their solvent by passage through a semi-permeable membrane. It
differs from dialysis in its dependence on an electric field as the driving
force for the separation. Thus the passivity and concomitant low energy
requirements of the dialysis process are traded for the speed of separation
associated with an energy-intensive process.
The membranes employed are specific for the passage of either cations or
anions, and since the driving force is the electric field, only electrically-
charged species can be separated. The process is shown schematically
(Figure i)1M6 for the desalination of brackish water, the most popular applica-
tion to date. As seen in the figure, the membranes are alternately
cation-specific and anion-specific; so that, when the electric field is applied,
the ions will segregate from the solvent to form a concentrated waste solution
and fresh water. Sixty or more cells may be stacked in parallel and five or
six stacks are often placed in series to achieve the 95 percent removal
efficiencies obtained with reverse osmosis or dialysis.
2. OPERATING PRINCIPLES
Electrodialysis depends on the forced migration of ions selectively
through a membrane. The driving force is an electric field set between two
electrodes toward which the ions flow. The amount of material which can be
processed is governed by the voltage and manifold pressure which can be applied
without jeopardizing the structural integrity and materials properties of the
membranes and the plumbing.
Preceding page blank
75
-------
SALINE WATER IN
Q
1 1 "
/
:) 1 0
M +
Na
0 | | ((
X
r— Na+
> 1 1 (
Cl
„ +
P —I
CONCENTRATED
BRINE WASTE
FRESH PRODUCT WATER
Figure 1. Electrodialysis Process
-------
Electrodialysis membranes are insoluble polyelectrolytes, electrically-
charged polymers, which Ideally have high electrical conductivity and high
selectivity for transport of the desired ionic species. Other desirable
characteristics are mechanical strength, chemical stability, and resistance
to fouling.1433'1452 A popular membrane consists of polystyrene made insolu-
ble by cross-linking with divinyl benzene. The cation-permeable membrane is
then made by sulfonating with sulfuric acid to 5 meq/g dry weight. Similarly,
the ani on-permeable membrane is made by chemically bonding quaternary ammonium
groups to the polystyrene at a concentration of 1 to 3 meq/g dry weight.
Unlike reverse osmosis or dialysis, there have been no significant innova-
tions in electrodialysis in the past few years, so that the technology is
assumed to be relatively mature.
3. PROCESS DESIGN
It is advantageous to put as many cells as possible in series with the
electrodes1433 because: (1) the energy consumed at the electrodes is relatively
constant with respect to the number of cells, (2) a pair of electrodes costs
approximately ten times as much as a pair of membranes, and (3) membranes
adjacent to the electrodes are subject to a more severe environment and
degrade more easily. The total number of cells in a unit is limited by the
maximum voltage which can be safely applied and general plumbing considerations.
Many theoretical and empirical expressions for the energy and power
required for a given separation of a feed stream into a dilute product stream
1 All 1 fil\^
and a concentrated waste stream have been derived.0'1 ° A particularly
useful one for design considerations is:
<»
77
-------
where V = power consumption, kilowatt-hours per 1,000 gallons
Fp = total concentrated stream flowrate, gallons per hour
A = total cell-pair area, square feet
N:,,, = concentrated stream normality, equivalents per liter
av
Nf = feed stream normality, equivalents per liter
Nd = dilute stream normality, equivalents per liter
k = conversion factor = 25.4
R = resistance of 1 square centimeter of a cell pair (dilute solution
+ concentrated solution + cation - permeable membrane + anion
- permeable membrane + polarized layers adjacent to membranes),
ohm - square centimeters
E = current efficiency s
In
where
F = Faraday's constant, 96,500 coulombs per equivalent
f. = dilute stream flowrate, liters per second
I = current, amperes
n = total number of cell pairs
Power requirements may run 10 kwh/ljOOO gal. for 90 percent removal in an
1651 1652
880,000 gpd treatment plant with feedwater concentrations of 2000 ppm. '•
The resistances in equation 1 are usually empirically determined. They
are not particularly amenable to design variation. When feedstream concen-
trations are 1000 ppm or more, current efficiencies may drop drastically
with concomitant increases in power consumption. These decreases are usually
attributable to decreases in membrane selectivity with increasing concentra-
tions, back diffusion of the solute ions, diffusion of solvent through membranes,
and electrical short circuiting through the solvent and the plumbing. Also,
polarization of the membrane and its vicinity may become significant, and
further decrease efficiencies and membrane life.
78
-------
Back diffusion of the solute ions may be decreased by increasing the
current density. This also decreases the capital cost of the electrodialysis
plant, but increases the operating costs, since more power is required. Also,
if the current density is too high, the membrane and the solution in its
vicinity become polarized. Polarization is a serious and costly problem,
because it leads to hydrolysis of the membrane and precipitation. The
prospect of polarization inhibits wide variations in the operating parameters.
Most electrodialysis units operate successfully with current densities in
2
the range 0.05 - 0.08 amp/cm .
Electrical short circuiting is minimized by increasing the velocity in
the system to decrease stagnation and associated charge build-up. This is
most easily accomplished by restricting pipe diameters. The benefits must
be traded against the increased pumping costs involved.
Leakage of solute ions through the wrong membrane is a function of
concentration and ionic size and charge. For multivalent ions of large size
it is almost nil, while hydrogen ions pass quite easily. The transport number
which is defined as the ratio of ions which pass a plane with and without
the membrane present, varies for various ions under several different conditions
(Table I).1452 A perfect membrane gives a transport number of zero for the
wrong ion and a transport number of unity for the right one.
Operation is restricted to temperatures under 100 F to promote membrane
life. Removal efficiencies increase markedly with temperature and number of
stages- (Table2).1651-1652
With strong acid or base groups in the polyelectrolyte membrane, the
membranes are relatively insensitive to pH and operation over a range of pH
1 to 14 is common1 . If weak groups such as acetic acid are used, the
membranes are sensitive to hydrolysis and the pH of the solutions must be
kept within a narrow range. Therefore, strong groups are used almost
exclusively.
*each stage consisting of 60+ cells
79
-------
TABLE 1
SELECTIVITY OF IONIC-BARRIER MEMBRANES
Barrier Solutions IN Permeable
ion
An ion NaCl NaOH Na*
Na2S04 NaCl Na+
NaCl Na2S04 Na+
MgS04 MgCl2 Mg++
Cation NaCl HC1 Cl"
HC1 NaCl Cl "
NaCl MgCl2 Cl"
Current density 90 amp./sq. ft.
Leakage
1on
OH"
Cl"
so4"
Cl"
Mg**
Transport No.*
leakage 1on
0.3100
0.0500
0.0030
0.1200
0.5200
0.0090
0.0002
-
TABLE 2
TYPICAL MINERAL REMOVALS FOR ED UNITS
AT DIFFERENT WATER TEMPERATURES
50 F
Two Stage ED 51 - 65%
Four Stage ED 76 - 88%
Six Stage ED 88 - 96%
Temperature
75 F
62 - 79%
85 - 95%
94 - 99%
100° F
76 - 88%
94 - 98%
98 - 99+%
80
-------
Auxiliary equipment includes a manifold system to direct the respective
feed, dilute, and concentrated streams into and out of the system and the
assorted pumps, plumbing, valves, and instrumentation required in almost any
chemical engineering operation. A source of DC power is of course required.
This may run to 400 KVA for an 800,000 gpd plant.
4. PROCESS ECONOMICS
On a per stage basis, capital and operating costs for an electrodialysis
plant are low. However, each stage has only a 35 to 45 percent per pass
efficiency in removing dissolved solids, »lbbl,lbW SQ ^^ stages ^y j,e
required to obtain water of acceptable quality, with proportionate increases
in cost. The Sanitation Districts of Los Angeles County1648'1649 have
operated a 15,000 gpd electrodialysis pilot plant with the object of reducing
the IDS of a secondary sewage treatment effluent. The results indicate
that electrodialysis is not economically viable when compared to reverse
osmosis, as shown by the projected capital and operating costs for a 10 million
gpd system (Table 3). Note that the reduction in TDS was only 34 percent for
this single stage unit and additional 34 percent stepwise reductions could only
be obtained by increasing the number of stages with essentially no reductions
in per stage costs.
A cost breakdown for a 28,000 gpd facility installed at Coalinga,
California to make brackish water potable has been presented by Farrell
and Smith (Table 4).1647 No information with regard to concentrations is
available.1433 Parametric studies of cost as a function of required reduction
in TDS and plant capacity have been performed (Figures 2 and 3). Note
that the per gallon cost is relatively insensitive to capacity, but is very
sensitive to the amount of reduction required.
5. PROCESS MODIFICATIONS
If it is desired to remove soluble organic material, a pretreatment
with activated carbon is necessary. This will also remove insoluble
materials which might otherwise clog the membranes. To maintain a desir-
81
-------
TABLE 31648
CAPITAL AND OPERATING COSTS FOR
10MGD ELECTRODIALYSIS PLANT
Amortization of Capital 6/1000 Gallons
$1,830,000; 20 years at 5% 4.0
Operation and Maintenance
Chemicals 4.4
Membrane replacement 3.3
Maintenance materials 0.9
Power 1.9
Labor 2.5
13.0
TOTAL 17.0
Assumptions: 1. Influent TDS = 700 mg/1, effluent TDS = 450 mg/1
2. Membrane life = 5 years
3. Water recovery = 90%
82
-------
TABLE 41647
OPERATING COST FOR 28,000 GPD ELECTRODIALYSIS
UNIT AT COALINGA, CALIFORNIA*
Cents per Thousand Gallons
Labor 16.67
Chemicals 6.25
Parts and replacements membranes 23.25
Filters 18.61
Other parts 3.01
Power 21.46
Amortization of capital investment 54.00
Total cost 143.25
*These figures include an allocation of existing labor time—none was
added for this operation. Increased power cost and shorter filter life
contribute slightly to the apparent high figure and straight line
amortization on an investment of $105,000 was used. A decrease by as
much as 15 pents per thousand gallons is indicated by increased filter
and membrane life. Potable water costs $7.05 per thousand gallons before
installation of the electrodialysis unit.
83
-------
00
H
in
8
o
z
M
3.00
2.00
o
o
o
1.00
0.80
0.70
^ 0.60
£ 0.50
| 0.40
8 0.30
10"
FEEDWATER SALINITY, PPM SALT
10'
Figure 2. Water desalting cost vs feed water salinity for A, 100,000 gal/day capacity; B, 1,000,000
gal/day capacity. Product water assumed to contain 500 ppm salt.
-------
u
§
o
H
en
3.50
3.00
2.50
2.00
M 1.50 -
CO
UJ
Q
I
H
1.00
0.50 -
10 10 10" 10 v 10' 10"
PLANT CAPACITY, GAL/DAY OF PRODUCT
Figure 3. Water desalting cost as a function of plant capacity. Comparison of electrodialysis for
A, 35,000 500 ppm salt; B, 2000 500 (or 1000 150) ppm salt, the center line for that
band representing the statistical average; C, 1962 estimates for various distillation
processes; D, 1964 estimates of combined power generation and seawater distillation by
multistage, multieffect distillation.
-------
able pH of 3.5, the Los Angeles Sanitation Districts add sulfuric acid to
the feed stream.
Ionics, IncJ65°.l65l»1652 fl large manufacturer and instaiier of
electrodialysis plants, suggests electrodialysis for reduction from thousands
to hundreds of pptn followed by ion exchange to reduce the TOS to a final,
acceptable level. While this approach may be efficacious under some
circumstances, e.g., in making brackish water into high quality potable
water, it is generally more expensive than reverse osmosis.
6. PROCFSS APPLICABILITY
Electrodialysis is applicable when it is desired to separate out
a variety of ionized species from an unionized solvent such as water. In
this regard, it might prove advantageous over reverse osmosis, where
different species may interfere with one another, or dialysis, where the
relative diffusivities and activities of the species play an important role.
lonizable nitrates and phosphates (e.g., Pb(N03)2, Hg(N03)2» Na3 P04)are.
removed with varying degrees of efficiency. The Los Angeles County
Sanitation Districts remove as much as 50 percent of the nitrates (as
nitrogen) with a one-stage pilot plant (Figure 4 and Table 5). This
plant was recently shut down because its 34 percent TDS removal was non-
competitive on a cost basis with the higher removal efficiencies associated
with reverse osmosis and ion exchange. Also, ,electrodialysis was found to
lie ineffectual for the removal of organic materials.
Electrodialysis has been used by military bases and other small
communities for the production of marginally potable water (500 ppm TDS)
from brackish water.1651'1652'1433'1646
With regard to National Disposal Sites, electrodialysis is applicable for
the treatment of waste streams where it is desirable to reduce the concentra-
tions of ionizable species in the intermediate range (10,000 ppm to 500 ppm)
over a broad range of pH (e.g., pH 1 to 14). Such streams may derive
86
-------
FEED
12 gpm
DILUTE
STREAM
11 gpm
CONCENTRATE
STREAM 1 gpm
I
I pH
'3.5
L.
ELECTRODIALYSIS
STACK
ACID INJECTION
BRINE RECIRCULATION
11 gpm
-^- PRODUCT
WASTE BRINE
Figure 4. Schematic Flow Diagram of Electrodialysis Pilot Plant
-------
TABLE 5
WATER QUALITY CHARACTERISTICS OF ELECTRODIALYSIS PROCESS
TEST (mg/1) INFLUENT EFFLUENT % REMOVAL
COD
NH^-N
N03-N
P04-P
TDS
9.4
9.0
6.2
10.1
705.0
8.0
5.1
3.1
7.8
465.0
14.9
43.4
50.0
22.8
34.0
NOTES
1. 15,000 GPD Pilot Plant
2. Influent is Carbon Effluent
3. Averages Based on 10 Samples
4. Analyses Run on 24-Hour Composite Samples
5. Average Water Recovery - 83 Percent
88
-------
directly as combustor scrubber liquors, for example, or they may be the
effluent from another treatment process which handled a stream of much higher
concentration (e.g., dialysis). If an effluent of concentration lower than
500 ppm is desired, the electrodialysis effluent could be fed into another
treatment process such as ion exchange. Its applicability to unionized
organic species is effectively nil.
89
-------
7. REFERENCES
1433. Kirk-Othmer encyclopedia of chemical technology. 2d ed. New York,
Interscience Publishers, 1963.
1452. Perry, J. H., et al. Chemical engineering handbook. 4th ed.
J. H. Perry, et al, 1963.
1646. Ruebsamen, W. B. Electrodialysis water treatments. The Military
Engineer. (411): 34-36, Jan-Feb 1971.
1647. Farrell, J. B., R. N. Smith. Industrial and Engineering Chemistry.
54 (6): 29-35, June 1962.
1648. Dryden, F. D. Mineral Removal by Ion Exchange; In. Proceedings:
Workshop of Waste and Reuse, University of California, Berkeley,
California, South Lake Tahoe, California, June 25-26, 1970. 19 p.
1649. L. A. County Sanitation Districts. An introduction to the sanitation
districts of Los Angeles County. 4 p.
1650. Ionics, Incorporated. Fresh water from saline sources by electro-
dialysis. (6): Ionics, Incorporated, Watertown, Massachusetts.
4 p.
1651. Katz, W. E. Electrodialysis preparation of boiler feed and other
denrineralized waters. Ionics, Incorporated, Watertown, Massachusetts,
1971. 24 p.
1652. Katz, W. E. The application of electrodialysis equipment for the
treatment of waters below 1000 ppm total dissolved solids. Ionics,
Incorporated, Watertown, Massachusetts, 1970, 15 p.
1653. Tuwiner, S. B. Diffusion and membrane technology. New York,
Reinhold Publishing Corporation, 1962. 421 p.
90
-------
ION EXCHANGE
1 . INTRODUCTION
Ion exchange may be defined as the reversible Interchange of Ions
between a solid and a liquid phase In which there Is no permanent change
In the structure of the solid. It provides a method of separation useful
In many chemical processes. Ion exchange is a method of collecting and
concentrating undesirable materials from waste streams.
The mechanism of ion exchange is chemical, utilizing resins that react
with either cations or an ions. A waste treatment system may use a cation-
exchange resin, an anion-exchange resin, or both types of resins, depending
on the ions to be removed from the waste stream.
There are basically two types of ion-exchange operations - fixed bed
and continuous countercurrent exchangers. In recent years the continuous
ion exchanger has been used to reduce dilution, increase operating efficiency,
and reduce operating cost.
In recognition of this current industrial trend and since the design
considerations and underlying operating principles are essentially identical
for both the countercurrent and fixed-bed systems, this report will emphasize
the continuous operation (Figure 1).
2. OPERATION PRINCIPLE
The great utility of ion exchange rests with the ability to use and
reuse the ion-exchange materials. For example, in water softening:
2 R Na+ + Ca+2 ^ R"2 Ca+2 + 2Na+
The exchanger I in the sodium ion form is able to exchange calcium with an
91
-------
MIXING HOPPER
OUTLET u 1 ||
TREATED WATER
1
1
I
INCOMING WATER
IMPURITIES
TREATMENT
ION
* EXCHANGED
I
TRANSFER
IMMHi
^^•M
SEPARATION
COLUMN
r"
1
1
rn '
•^^H
•4^
i
--
—
•q^
I I
l_—_ J ^_.
ANION
REGENERATION
COLUMN
J
CATION
REGENERATION
COLUMN
I
ALKALI
ACID
Figure 1. Continuous Ion Exchange Process - Mixed Bed
-------
equivalent quanity of sodium. Subsequently, the calcium-loaded resin may
be treated with a sodium chloride solution, bringing it back to the sodium
form, ready for another cycle of operation. This conversion step is
commonly called regeneration. The reaction is reversible and the resin is
not permanently changed.
To be generally useful, an ion-exchange material must meet a certain
number of basic requirements:
(1) The material should have a high total capacity (a high degree
of ionic substitution).
(2) The resin should be structured chemically to exhibit good
equilibrium characteristics when handling the expected solutes.
It should be designed for operation in the proper pH range and
with a degree of selectivity sufficient for the purpose and
capable of regeneration.
(3) Kinetic properties (determined by the nature of functional
groups and backbone structure) should be such that a high
operating or equilibrium capacity may be attained.
(4) Extremely low solubility. Not only is the low solubility
important for reuse but, inasmuch as the usual application
of ion exchange is to remove impurities, the leaching of
materials from the resin would have an adverse effect on
the quality of the product.
(5) Good chemical stability. This includes resistance to attack
by an acid or base, oxidation or reduction, and radiation.1433
The speed with which ion exchange takes place determines the portion
of the equilibrium capacity which may be utilized in a practical dynamic
situation. In column operations, rates determine the extent of deviation
from the theoretical elution curve.
The rate of ion exchange depends, like ordinary adsorption, upon rates
of the following individual processes: (1) diffusion of ions from the bulk
liquid to the external surface of an exchanger particle; (2) inward diffus-
ion of ions through the solid to the site of exchange; (3) exchange of the
ions; (4) outward diffusion of the released ions to the surface of the solid;
(5) diffusion of the released ions from the surface of the solid to the
93
-------
bulk of the liquid. In some instances, the kinetics of the exchange re-
action (3) may be controlling, but in others the rate of reaction is
apparently very rapid in comparison with the rate of diffusion. The
diffusion rates can be described by appropriate mass-transfer coefficients.
For those cases where the exchange reactions are rapid in comparison
with the rates of mass transfer, the methods of design for conventional
adsorbers may be applied to ion exchange operations directly.
3. PROCESS DESIGN
The "continuous countercurrent ion exchange system"' is a closed system
in which ion exchange resin moves from process section to process section
in a batchwise manner. In this way a fresh quantity of resin is available
for exchange with the waste stream almost all of the time, not being
available only during the times the "slugs" of resin are moved from
section to section. The three activities required for use and reuse
of the resin, direct exchange, regeneration, and cleaning are accomplished
in five process sections: 1) the service or loading section; 2) the service
rinse section; 3) regenerant section; 4) regeneration rinse section and 5)
backwash section (Figure 2).
The service section of a continuous ion exchange system is the primary
working section of the apparatus. It is here that the resin is contacted
with the waste stream and the specific ions exchanged. The material to
be processed is fed into the service section and flows countercurrent to
the direction of resin movement allowing complete utilization of the resin.
One of the important considerations in the design of the service section
is resin selection. The ion exchange operating capacity, ionic selectivity,
particle size, and hydraulic characteristics of the resin are factors which
affect the design of the service section. The regeneration efficiency,
resistance to attrition, and structural properties of the resin determine
how often it must be replaced in the system.
94
-------
Ul
BACKWASH
OUT
1. Service or loading section
2. Service rinse section
3. Regenerant section
4. Regeneration rinse section
5. Backwash section
Figure 2. Flow pattern of a continuous countercurrent ion exchange system
-------
The length of the ion exchange zone and the residence time of the
resin must be determined for the design of the service section. These
are dependent on the process ion exchange rate and the quantity of ions
required to be removed from the waste stream.
The pressure requirements of the process liquid pumping system and the
available pressure drop are determined from the design of the service
section.
The method of moving the resin from one process section of the apparatus
to another is an important design consideration. Compressed air is used
to push a "slug" of resin and the "slip water" associated with it from
one section to the next. The ratio of "slip water" to the amount of resin
moved is kept at a minimum to increase resin productivity. This is
usually accomplished by installing a conductivity probe in the service
section of the apparatus. This probe senses the conductivity of the
solution in the service section and electronically controls a valve which
discharges the "slip water" from the system. If the conductivity of the
solution is too high or too low, in relation to a certain "set" conductivity,
the discharge valve is opened or closed accordingly.
The service rinse section rinses the resin free of the exchanged pro-
duct before it leaves the service section. The rinse flows countercurrent
to resin movement (as does the feed material) and can be controlled by
conductivity. With a proper "slip water" ratio, the service rinse will be
at a minimum and will not cause any appreciable dilution of the product.
When the resin is exhausted the compressed air, which is introduced above
the service section, moves the resin into the regenerant zone of the
apparatus. In this section, to obtain maximum utilization of the regenerant,
the regenerant also flows countercurrent to the resin movement. The cation
resins are regenerated with acid and the anion resins with caustic. The
materials used in the construction of the regenerant section must be re-
sistant to the corrosive properties of the regenerant.
-------
Rinsing of the regenerated resin is accomplished by introducing a
rinse into the apparatus above the point of regenerant introduction. The
amount of regenerant rinse used is again controlled by conductivity to
insure that the resin leaving the section is rinsed free of regenerant.
As the resin leaves the regenerant rinse section, it enters the
backwash section. In this section, the resin receives its final rinse
before returning to the service section of the apparatus.
The process liquid flow rate for a continuous ion exchange system is
six to eight gpm/sq ft of resin surface area. The normal operation time
of the system is two minutes in service and six seconds in resin movement,
with three to five seconds needed for delays to insure proper valving
sequence. With service flows operating 95 percent of the time, decreased
resin volumes are possible.
Comparing a continuous system to a fixed bed system utilizing the same
regenerant level, a capacity of 350 g/cu ft is obtained from a continuous
ion exchange unit compared to 268 g/ cu ft for a fixed bed system. Final
exchanged product dilution from a fixed bed is 28 percent, whereas in
the continuous system it is 17 percent.0304
4. PROCESS ECONOMICS
The capital investment costs of ion exchange plants are dependent
on the total volume of waste treated and,to a large extent, on the amount
of total dissolved solids (TDS) removed. These costs range from $850,000
for a 0.5 million gpd plant to $9.5 million for a 10 million gpd plant
(influent TDS of 1500 mg/1). For an influent TDS of 3000 mg/1 the capital
costs range from $1.25 million for a 0.5 million gpd plant to $10 million
I] 00
for a 5 million gpd plant.
The operating costs (less depreciation) for ion exchange plants are
largely determined by the regenerant chemical costs which are in direct
proportion to the rate of electrolyte removal. The operating costs for
97
-------
a 0.5 million gpd plant (influent TDS of 3000 mg/1) total 35.04/1000 gal
broken down as follows:
Labor
Materials
Chemicals
Utilities
Total
l.U
1.4
31.2
1.3
35. Ot
5. PROCESS MODIFICATIONS
The only major process modification currently in use is that of resin
pretreatment for the purpose of adsorbing certain nonionic organic sub-
stances from a waste stream. This modified process is called ion exclusion
and can be carried out with ordinary ion exchange resins. The resin is
presaturated with the same mobile ions (cations or anions, depending on
resin type) as are in the solution. It will then repel the ionic components
of the solution, while adsorbing neutral non-aqueous materials such as
alcohols, carboxylic acids, and ketones of relatively low molecular weight.
6. PROCESS APPLICABILITY
Ion exchange technology has been available and has been employed for
many years for removing objectionable traces of metals and even cyanides
from the various waste streams of the metal process industries. Objection-
able levels of fluorides, nitrates, and manganese have also been removed
from drinking water sources by means of ion exchange technology. Dangerous
levels of radioactivity have been controlled by ion exchange devices and
technology has been developed to the extent that the contaminants that are
removed and concentrated can either be recycled or readily transformed into
a harmless state or safely disposed.
98
-------
Ion exchange technology can be employed to remove, concentrate, and
Immobilize all of the metallic and non-metallic ionic species listed by
the United States Public Health Service as toxic or undesirable when present
in concentrations above certain levels. There are, for example, ion ex-
change techniques for removing the following potentially undesirable species
from water and waste streams; iron, aluminum, manganese, copper, zinc,
chromium, silver, nickel, cobalt, cadmium, barium, uranium, radium, mercury,
lead, fluorine, boron, nitrates, phosphates, arsenic, sulfides, phenol,
chlorophenols, glucose, and glycerine.0609
It has been reported that an acid mine-drainage water containing 3,000
ppm of dissolved solids has been successfully processed to yield a water of
quality suitable for industrial, agricultural, and domestic purposes by
using a modification of the DESAL process (Figure 3). The process is based
upon the use of an anion exchange resin, which functions in the bicarbonate
cycle to treat the acid mine-drainage water. The effluent water is aerated
and subsequently clarified to give a water of useful quality. The
composition of the treated and untreated water (Table 1) is based on data
supplied by Rohm and Haas Company.
' A large (500,000 gallons per day) modified DESAL process plant will be
operating late in 1972 treating acid mine-drainage effluents from the
coal mining areas in Phillipsburg, Pennsylvania. The treated effluent will
be used as a source of potable water for the local community.
With regard to National Disposal Sites,ion exchange could be used in
conjunction with other processes as is done in the modified DESAL facility.
The main purpose of ion exchange at the Site would be to concentrate and
remove specific hazardous wastes from various waste forms prior to long-
term storage or recycling.
99
-------
ION
EXCHANGER
8
O
LIME
COAGULATION
FILTER
RE GENERA NT
RECOVERY
IRON
MANGANESE
ALUMINUM
OXIDES
MINE DISCHARGE
PRODUCT
WATER
LIME
SLUDGE
LAGOON
Figure 3. Acid Mine-Drainage Treatment Plant (DESAL Process)
-------
TABLE 1
COMPOSITION OF UNTREATED AND DESAL—PROCESS TREATED
ACID MINE WATER0609
Concentration, ppm (CaC03)
Constituent Untreated Treated
pH 3.1 8.5
Ca 964.0 290.0
Mg 480.0 10.0
Fe 895.0 <0.1
Al 556.0 <0.1
Mn 91.0 <0.1
50= 3104.0 100.0
HC03- ~ 200.0
101
-------
7. REFERENCES
0285. Lund, H. F. Industrial pollution control handbook. 1 v. New York,
McGraw-Hill Book Company, 1971.
0304. Ross, R. D. Industrial waste disposal. New York, Van Nostrand
Reinhold Book Corporation, 1968. 340 p.
0609. Rohm and Haas Company. Application of ion exchange-pollution abatement
and control. Amber-Hi-Lites No. 123, July 1971.
1182. Datagraphics, Inc. Inorganic chemicals industry profile. Washington,
Environmental Protection Agency, July 1971. 167 p.
1289. Cecil, L. K. Water-1969. New York, American Institute of Chemical
Engineers. Chemical Engineering Progress Symposium Series.
65(97):315, 1969.
1433. Kirk-Othmer encyclopedia of chemical technology. 2d ed. 22 v. and
suppl. New York, Wiley-Interscience Publishers, 1963-1971.
102
-------
WASTE DISPOSAL PROCESS DESCRIPTOR REVERSE OSMOSIS
1. INTRODUCTION
Osmosis is the physical transport of a solvent across a membrane
boundary, where the driving force for the transport is the difference
in the solvent activities on either side of the membrane. Such a
membrane is semi-permeable, i. e., it permits solvent molecules to pass
through it, but is impermeable to solute species (Figure 1) . Solvent
will diffuse through the membrane until its thermodynamic activity is
equal on both sides. This activity is a function of the solvent concen-
tration and the pressure head. The pressure head present at equilibrium
is defined as the osmotic pressure.
If external pressure is applied to the side of less solvent concentra-
tion (.Figure 2), solvent will flow in the opposite direction. This
is defined as reverse osmosis. This allows solvent to be extracted from
a solution, so that the solution is concentrated and the extracted solvent
is relatively pure. This idea is being exploited commercially in various
areas of water purification, such as the purification of acid mine
waters, tertiary sewage treatment, and the purification of tap
water in the microelectronics industry. Often feed streams are
subject to a reverse osmosis stage prior to passing through an ion exchange
stage (Figure 3).1441
2. OPERATING PRINCIPLES
The basic consideration in the operation of any reverse osmosis unit
is the amount of solvent material which can be passed through the membrane
in a given time. This solvent flux across the membrane is proportional to
the difference between the applied pressure and the osmotic pressure, or
Jw = A(aP - An) (1)
103
-------
"7
h
SEMfPERMEABLE
MEMBRANE
SOLUTION
SOLVENT
(WATER)
Figure 1. Osmosis; Normal flow
from low to high concentration.
rl
PRESSURE
PISTON
SEMIPERMEABLE
MEMBRANE
SOLUTION
SOLVENT
(WATER)
Figure 2. Reverse Osmosis; flow
from high to Tow concentration.
104
-------
RAW
WATER
1 PRETREATMENT 1
1 IF NECESSARY 1
L -r 1
j /
X
1 — r w
/
25
Fl
CHEMICALS
FORoH
CONTROL
/ STORAGE
t >
/-* S. ^ REVERSE MIXED
MICRON f \ OSMOSIS —& RESIN
LTER ^~ J UNITS BED
1 i
CONCENTRATE M|XED
FOR LARGE SYSTEMS P**'" ! RESIN
MIXED
BED
0
\ '
POLISH
MIXED
BED
i
FINAL
FILTER
4—
s
ANION _ CATION ^ ^
RESIN *~ RESIN ^
1 MICRON
FILTER
^
TO USE
AREAS
\ >*
TO USE
AREAS
FOR
SMALI
SYSTEMS
Figure 3. Flow Diagram of a Deionization System Incorporating Reverse Osmosis
105
-------
where A is a constant related to the properties of the membrane, AP is the
difference between the feed and product pressures, and An is the difference
between the osmotic pressure of the feed and product, respectively, with
respect to pure water.
The sin qua non of the reverse osmosis process is the membrane. The
exact mechanisms of membrane transport are not well understood, but two
basic membrane operating principles are possible in theory: (1) a
porous membrane which depends upon the size difference between water
molecules, the typical solvent, and hydrated ions to impede the flow of
the hydrated ions through the micropores, while allowing the smaller water
molecules through, and (2) a diffusive, homogeneous membrane which cannot
be described in terms of pore size, but which relies on physical and
chemical differences between water molecules and hydrated ions in a
poorly understood way to allow the selective passage of one and not the
other.
The diffusive, homogeneous membrane is far more efficient and is
used almost exclusively. Its characteristics impose the following
qualitative rules of thumb in the operation of any reverse osmosis
unit: 1M1
(1) The passage of ions (failure to reject) will depend on
the concentrations in feed and product;
(2) The fraction of ions passing through is relatively
independent of concentration at low concentration, i.e.,
when the mole fraction of solvent is almost 1.0;
(3) Univalent ions pass more easily than multivalent ions;
(4) Organic molecules of large molecular size will be
rejected very efficiently;
(5) Organic molecules of small molecular size will pass
through quite easily.
106
-------
3. PROCESS DESIGN
The process design is concerned with improving the factors associated
with maximizing the solvent flux (Eq. 1) and prolonging the membrane
lifetime.
The solvent flux, J is directly proportional to the pressure
differential, AP-Aii. AH is dependent on the characteristics of the feed
stream and is generally raised approximately one psi for every 100 ppm
dissolved solids in the feed stream. AP is limited by the pressure
which the membrane can withstand, and currently is limited to 600 psi in
commercially installed units. 1436>0851 Gulf General Atomic,0851' 1437
which is presently conducting research and development in this field
while installing units on a commercial basis has successfully operated
experimental units at 800 psi and now talks of 1500 psi as possible.
The membranes currently in use commercially and experimentally are
made of cellulose acetate modified in various ways. They are developed
and manufactured by the basic chemical companies, DuPont, Dow, Kodak, etc.
and further modified, packaged, and sold in reverse osmosis systems by
several firms, including Polymetrics, Inc. and Gulf General Atomic, Inc.
Mills describes the membranes as provided by the manufacturers:
"The modified cellulose acetate membranes, which are in general
use, are about 100 microns thick. The active part of the mem-
brane is a dense layer 0.2 to 1.0 micron thick formed during
the initial stages of the gelling or precipitation of the mem-
brane. This thin active layer forms the rejection surface of
the membrane and its properties can be changed by: (a) the
addition of additives to the casting solutions, (b) variations
in the solvent and nonsolvent portions of the casting solutions,
and (c) post-casting treatments. The remainder of the membrane
is a porous mass which supports the active surface. Cellulose
acetates can be characterized by their degree of acetylation
(o to 3); the grades of cellulose acetate which are preferred
107
-------
for membrane preparation have a degree of acetylation of
approximately 2.5; this corresponds to an acetyl content
of approximately 40 percent".
Choosing a membrane is still primarily a state of the art procedure,
involving extensive testing on the laboratory level up through several
stages of pilot plant development. 1436>1437>1443. 1446 For tMs reason
it is impractical to give generalized rules of thumb for membrane
selection in this description.
These membranes have rejection efficiencies of 95 percent for
univalent ions, 98 to 99 percent for multivalent ions, and 90 percent
for organic molecules whose molecular weights exceed 100.
Since cellulose acetate hydrolyzes readily in strongly acidic or
strongly basic solutions, the pH of the feed water must be kept between
3 and 8, and preferably between 5 and 7 (Gulf °851'1437 reports several
significant excursions with no ill effects). Sulfuric acid or caustic
solution are commonly used to adjust the pH to appropriate values and
facility for determining pH and metering in the solutions must be provided
in the design.
The membrane constant, A, in Eq. 1, usually varies between 0.015 and
? 1441 0851
0.020 gal./ft -day-psi at 70 F l^tl>"° and increases approximately
1.5 percent/F with temperature. A higher operating temperature is there-
fore preferred, but an upper limit of only 100 F is allowed in practice,
because the membrane is subject to hydrolysis above this temperature.
Membranes are subject to fouling caused by sedimentation, precipitation,
and bacteria. Gulf and McDonnell Douglas have successfully
combated fouling in experimental systems by running solutions of common
household detergents through the system.
The packaging of membrances to conserve space and maximize longevity
has achieved considerable sophistication. Tubular, spiral, and hollow
fiber configurations have been used successfully and are under vigorous
108
-------
investigation.1436'0782'1446 Packaging efficiencies as high as 1,500 ft2
of membrane in a cylinder 4 in. in diameter by 4 ft. in length are already
available commercially with a membrane guarantee for 1 year. With
a membrane constant of 0.02 gal./ft -day-psi and a pressure differential
of 600 psi, such a unit could treat over 15,000 gpd.
In addition to the membrane assembly, a pressure vessel to contain
the membrane is required along with pumps and suitable instrumentation.
Polymetrics 446 supplies an 18,000 gpd unit with associated equipment
in a volume of 12 ft by 4 ft.
4. PROCESS ECONOMICS
For the past several years, Gulf General Atomic °851 ' 1437 has been
operating a pilot plant at Norton, West Virginia for recovery of acid mine
waters. Sufficient data has been gathered to allow a realistic assessment
of the capital and operating costs associated with such a venture. These
costs may be regarded as typical for reverse osmosis plants for the purposes
of this description since details such as chemical additives to the feed
water to adjust pH, the addition of bactericides, etc. comprise a small
fraction of the total cost. Capital costs are estimated at approximately
$780,000 for a 1,000,000 gpd unit. Operating costs for a plant of this size
total 41.34/1,000 gal., broken down as follows:
Power 7.5$
Chemicals 3.2$
Operation and Maintenance 5.0$
Membrane Module Replacement 25.6$
TOTAL 41.3$
109
-------
5. PROCESS MODIFICATIONS
Modifications for the most part consist of doctoring the feed stream
with additives to Improve the rejection rate and Inhibit membrane fouling.
This 1s still considered to be an art rather than a science and reverse
osmosis unit manufacturers do a considerable amount of tailor-made develop-
ment work with every unit they sell. 435»1437 The membrane manufacturers,
of course, concentrate on Improving membrane properties and longevity, and
In concert with the unit manufacturers, strive to Improve the packing tech-
niques for greater rejection and acceptable pressures.
Often, preliminary laboratory Investigations Indicate that several
dissolved substances in a waste stream may be mutually Incompatible for
efficient reverse osmosis treatment. This then requires a change of membrane
or preliminary separation of the offenders prior to the reverse osmosis
treatment.
6. PROCESS APPLICABILITY
Almost any dissolved solid can be treated by reverse osmosis,
providing the concentrations are not too high (Table 1) and it 1s practical
1440
to adjust the pH to range 3-8. The main problem is membrane specificity
for particular species, as influenced by the presence of other species, pH,
tendencies toward fouling, additives, etc. when concentrations are high.
Reverse osmosis units capable of handling a wide variety of substances
of relatively high concentration simultaneously or in series are somewhat
beyond the present state-of-the-art, 43B«1439 although systems can be
designed for handling limited combinations of species. Typical commercial
reverse-osmosis units have capacities of 15,000 gpd and normally include
a prefilter and pH control pump in the integrated system (Figure 4).
At present several reverse osmosis pilot plants are in operation. For
example, Gulf General Atomic, Inc. 1437«1438»0851 has set up two pilot
plants of approximately 10,000 gpd feed water capacity - one in Norton,
West Virginia to process acid mine waters being deposited in Grassy Run
Creek and one in Pomona, California under a project jointly sponsored
110
-------
TABLE 1
MEMBRANE REJECTIONS AND PERMISSIBLE FEED WATER CONCENTRATIONS
Name Symbol
Maximum
Typical Allowable (%)
% Concentraction
Reject, in Feed Water
Sodium
Na-
94-96
3-4
Calcium
Magnesium
Potassium
Iran
Mangoneso
Aluminum
Ammonium
Copper
Nickel
Strontium
Hardness
Cadmium
Silver
Chloride
Bicarbonate
Sullate
Niliale
Fluoride
Silicate
Phosphate
Bromide
Borate
Chromate
Cyanide
Sulfite
Thiosullatc
Feirocyanido
Ca 2
Mfl-2
K-1
Fe-'
Mn-'
AI->
NH.-i
Cu->
Nl-'
Sr >
Ca and Mg
Cd »
AQ-I
ANIONS
Cl-'
HCOj-1
SO,-'
NO,-1
F-1
SiO}-'
PO.-'
flr-i
B.Oi-2
Cr04-J
CN-i
SO,-'
3,0,-'
Fe(CN)4-'
98-98
96-98
94-96
98-99
88-99
99
88-95
96-99
S7-99
UD-99
96-98
95-98
94-96
94-95
95-96
99-
93-96
94-96
9S-9T
99-
94-96
35-70' '
90-98
90-95' '
98-99
99
99-
•
•
3-4
•
•
5-10
3-4
8-10
10-U
—
•
8-10
'
3-4
5-8
— Sir—
3-4
—
10-14
3-4
—
8-12
«-12
8-12
10-14
B-14
* Musi walch lor precipilallon, other Ion controls maximum
concentration
• •Extremely dependent on pH, tends to be an exception to the
rule
ORGANICS.
Molecular Weight
_342
360
Biouhumicai Oxygen
L)cm,ind
Chemical" Oxygon
Osmond
Urea
(BOD)
90-99
(COO)
80-95
60
40-60
Rracii ci
to n sa
Bacteria 4 virus
Pyrogurl
_5.000-1pOOpO_
lood-lood
1°_°
~100"
•••Permeale is enriched in nalcnal due lo pretercn'i-H pa
-------
SUPPLY
WATER
PC
PRE-
FILTER
CENTRIFUGAL PUMP
pH CONTROL
PUMP
RECIRCULATION VALVE
CONCENTRATE
VALVE
INTEGRAL STANDARD POLYMETRICS SYSTEM
I
DRAIN
Figure 4. Typical, Coirmerciany Available Reverse Osmosis Unit 9
-------
by the Los Angeles County Sanitation District and the U.S. Environmental
Protection Agency. The purpose of the Pomona facility is to investigate
the feasibility of reverse osmosis as a tertiary sewage treatment process
to remove salts from industrial waste waters.
Results are available for the Norton plant for a 54-day test in 1969.°851
The plant was run at 600 psi with a 75 percent recovery, i.e., 3 gal. pure
product out of every 4 gal. feed water. Average rejection was 97.6 percent
of an initial feed water concentration of 3,600 ppm. The unit was in
continuous operation 85 percent of the time, with most of the down time
attributable to pump malfunction.
The Pomona operation may be summarized as follows: 13,000 gpd feed
water from an activated sludge secondary sewage treatment contain 700 mg/1
dissolved inorganic solids. As a pretreatment, 90 percent of the organic
matter is removed by carbon adsorption. The product water contains 30 mg/1
dissolved solids. The data obtained to date suggest that ion exchange is
more economical than reverse osmosis, which in turn is more economical than
electrodialysis. However, ion exchange is ineffectual in the removal of
viruses, and if reverse osmosis can be shown to be effective for virus
removal, it might prove the better choice. It is expected that reverse
osmosis has excellent rejection for viruses, since they are of very
large molecular weight.
Reverse osmosis is an appropriate method for concentrating wastes on
the hazardous materials list which form ions 1n aqueous solution, e.g.,
ammonium chromate. Also, organic materials of large molecular weight, such
as dyes or bacteria, which dissolve or form suspensions in water are readily
separable. Materials whose rejections by reverse osmosis are very poor are
mainly low molecular weight organic compounds which do not ionize 1n aqueous
solution, e.g., ethanol or urea. Of course, as a general rule, materials
which are marginally soluble or insoluble are not appropriate, since their
precipitation clogs the membrane.
113
-------
_
With regard to N*t.i.oR.a.l-0-i-&po?a*l-S'fte/|' reverse osmosis could be used
for concentrating scrubber liquors, as might originate from combustors.
The waste could then be further concentrated by evaporation, for example,
prior to long-term storage or recycling to the supplier.
114
-------
REFERENCES
0782. Bishop, H. K. Use of improved membranes in tertiary treatment by
reverse osmosis. PB-203 206. Newport Beach, California.
McDonnell Douglas Astronautics Company, Dec. 1970. 76 p.
0851. Kremer, S. S. Reverse osmosis field testing on acid mine waters at
Norton, W. Virginia. PB-198-936. San Diego, California, Gulf
General Atomic, Inc., July 1970. 44 p.
1433. Kirk-Othmer encyclopedia of chemical technology. 2d ed. 22 V. and
Supplement, New York, Interscience Publishers, 1963-1971.
1436. Personal communication. S. Haldow, Polytechnics, to M. Appel,
TRW Systems, Mar. 20, 1972.
1437. Personal communication. S. Kremer, Gulf General Atomics, to
M. Appel, TRW Systems, Mar. 24, 1972.
1438. Personal communication. Mr. Howorth, Los Angeles County Sanitation
District, to M. Appel, TRW Systems, Mar. 28, 1972.
1439. Personal communication, S. Linville, Aqua Media Company, to
M. Appel, TRW Systems, Mar. 20, 1972.
1440. Spatz, D. D. Reclaiming valuable mineral wastes. Pollution
Engineering, p. 24-26, Jan./Feb. 1972.
1441. Mills, A. R. Reverse osmosis for purification of water. Solid
State Technology, 13(7):41-45, July 1970.
1443. Reverse osmosis for waste water treatment. GA-8020. San Diego,
California. Gulf General Atomic, Inc., Mar. 1967.
1444. Merton, V., I. Nusbaum, R. Miele. Organic removal by reverse
osmosis. GA-8744. San Diego, California, Gulf General
Atomic, Inc., Aug. 1968.
1446. Advertising Brochure. San Carlos, California, Polytechnics.
1447. Advertising Brochure. North Hollywood, California, Aqua-Media.
1448. Why hollow-fibre reverse osmosis won the top CE prize for Du Pont.
Du Pont. Chemical Engineering. 78(27):54-59, Nov. 29, 1971.
115
-------
RADIOACTIVE WASTE SOLIDIFICATION
1. INTRODUCTION
The solidification of high-level liquid radioactive wastes is
utilized as a pretreatment for storage or disposal. Solidification
results in the conversion of the liquid waste to solids to provide greater
safety during handling, transportation, and disposal. The increased
safety associated with the solid wastes compared to the liquid is because
the solid form is less mobile, less soluble in water, smaller in volume,
and more rugged physically.
Several different processes have been developed for solidifying
.•
radioactive wastes. Generally, these processes are broken down into two
categories. They are high-level and intermediate-level radioactive waste
solidification processes, and low-level waste solidification processes.
The high-level solidification processes are used for the solidifi-
cation of the waste generated by reprocessing the spent fuel from nuclear
power plants. The spent fuel is ir> the solid form when it is removed
from the reactor and shipped to a reprocessing plant where it is processed
for the recovery of the usable fissionable materials. The high-level
waste is the aqueous effluent of the reprocessing solvent extraction
operation. The incentive to reprocess the spent fuel is three-fold: a
need to prevent depletion of the present uranium stocks; the economic
benefits realized which can represent a savings of approximately $38,000
per metric ton of fuel processed; the need to isolate and concentrate the
fission product waste.
Solvent extraction using nitric acid as a salting agent is the means
currently used for the first-stage removal of the fissionable materials.
The resultant high-level wastes are primarily aqueous solutions of
inorganic nitrate salts. Any differences in the waste streams generated
117
Preceding page blank
-------
will mainly occur due to variations in the amount and type of salts added
to the solutions during reprocessing. These variations can have a marked
impact on the solidification process conditions and on the nature of the
final product.
In the United States four processes have been developed for the
solidification of high-level wastes. They are pot calcination, spray
solidification, phosphate glass solidification, and fluidized bed
calcination. Basically, in each of these four processes heat is applied
to raise the temperature of the waste to drive off the volatile constituents
that will eventually cool to a solid. In some of these processes chemicals
are added to the waste stream to reduce the melt temperature to below
1,200 C. In all these processes, additional gas and vapor streams are
generated which require additional processing.
The intermediate-level wastes which are obtained from evaporating
second- and third-cycle wastes, cell and equipment decontamination,
solvent cleanup, and off-gas scrubbers will generally be treated the same
as high-level wastes. The volume of these wastes range from 200 to 500 gal.
per metric ton of fuel processed. These wastes are principally nitrate
solutions of sodium, potassium, aluminum, and iron, and often contain
sulfate, fluoride, and phosphate in addition. Their activity levels are
generally less than a tenth of a curie per gal. The low-level waste
solidification processes are adaptable to a number of different types of
waste streams. These include wastes generated at nuclear power plants
and fuel reprocessing facilities, along with the wastes resulting from
research laboratories and medical and industrial applications of radio-
isotopes. The wastes generated at nuclear power plants include radioactive
filter sludges, spent resins, and wastes which result from the treatment
of the condensate water and water from other sources in the plant. The
low-level wastes from the only operating fuel reprocessing plant is contained
in a liquid stream and contains only very small quantities of fission
products which are well below the Maximum Permissible Concentrations (MPCs)
listed in 10CFR20.
118
-------
2. OPERATION PRINCIPLE
Solidification of radioactive wastes is based on the principle of
converting the wastes from a mobile aqueous state to a solid state to
reduce the mobility of radionuclides after burial or storage. The
process for the solidification of the high-level radioactive wastes is
based on the addition of heat to volatilize the water, leaving a melt
that will eventually cool to a solid. To lower the waste melt temperature
and to enhance the solidified product, chemical modification of the
waste composition is required. Generally, compositions for waste
solidification require the addition of at least 70 mole percent of inert
chemicals to incorporate the wastes into a low temperature melt (less
than 1,000 C).
3. PROCESS DESIGN
This section is divided into two parts. In the first part the
desired characteristics of the solidified waste product are discussed.
These characteristics are of primary importance in evaluating the
adequacy of the solidification processes. In the second part the design
of the solidification processes for the high-level and low-level wastes
is described.
Solidified Waste Characteristics
The purpose of any solidification process is to achieve a substantial
increase in the safety associated with the handling, storage, transporta-
tion, and disposal of radioactive wastes. The increase in the safety of
the solidified waste is a result of the reduction in the mobility of the
radionuclides contained in the waste. To assess the safety of any disposal
technique, it must be recognized that solidification is only part of the
total disposal scheme. The other important consideration is the type of
disposal method utilized.
119
-------
teachability. A low Teachability of the solidified waste is
required to minimize the release of radioactivity to the environment
in the event that the waste container is breached and water should
come in contact with the solidified waste. The ideal leach rate of
the solidified waste can be assessed in terms of the specific activity
of the radionuclides present and their MFC values in water. 9 The
ideal leach rate is determined as follows:
• n _ (MPC) . ,V } . /I
L- R- - tory CT~J (
where
2
L.R. = leach rate, gm/cm - day
MPC = maximum permissible concentration of
radionuclide in water, curies/cm
S.A. = specific activity of radionuclide, curies/gm
V = volume of water contacted, cm3
T = time of contact, days
2
A = area of waste contacted, cm
To provide an indication of the ideal leach rate assume one square
centimeter of solidified waste were contacted by one thousand cubic
centimeters of water for one minute. This is equivalent to a flowrate
of 0.264 gallons per minute. For a waste containing only Sr90 with a
specific activity of 142 curies/gm and a 168 hour MPC value of
-12 3
1x10 curies/cm the leach rate of the solidified waste should not
-8 2
exceed 1x10" gm/cm - day. Since the ideal leach rate is a function
of the area of waste contacted, which decreases with time, the maximum
allowable leach rate would increase with time.
Stability. The stability of the solidified waste is important to
assure that the basic structure and properties of the waste remain intact.
The degree to which these properties change is a measure of the stability
of the waste. The basic structure and chemical properties of the waste
will change due to the decay of the radioactive materials to other chemical
elements. This effect is especially pronounced for reactor produced
waste since approximately 30 percent of the fission products present
120
-------
after 6 months out of the reactor will eventually decay to other chemical
elements. Since the change in the chemical composition of the waste is
impossible to prevent, the wastes should be held a sufficient time
before solidification to allow a majority of the short-lived isotopes
to completely decay. The exact time the waste should be held before
solidification is a function of the half-life of the radionuclides
present and should be such that any changes in the solidified waste
composition will not affect its basic structure or physical properties.
The exposure of the wastes to high levels of radiation can cause
radioactive gas evolution, swelling and discoloration of the waste,
and changes in its leach rate and physical properties. The exposure
rate and its effect on the solidified waste are dependent upon the
type of waste and amount of radioactive material contained. The level
of radiation damage is usually determined experimentally. Once the
level of damage is determined the amount of radioactivity or the amount
of radioactive material contained in the waste should be kept below the
level at which significant damage occurs.
The temperature stability of the waste determines the maximum
temperature at which the waste can be exposed before gas formation
and dispersion occurs. Generally, gas formation from solidified
wastes is not significant if the waste temperature does not approach
processing temperature. °715 For some of the high-level solidification
processes, a few exceptions have been indicated for calcine from feeds
with high sodium nitrate content (nitrogen oxide volatility) and for
some phosphate-sulfate melts (sulfur oxide volatility). °715
Thermal Conductivity. The maximum temperature of the solidified
waste must be maintained below prescribed limits to prevent the waste
from melting or vaporizing. To achieve these limits, a high thermal
conductivity is required to minimize the temperature rise between the
center and outer surface of the waste. The temperature difference
between the center and outer surface of a solid cylinder of
self-generating heat varies as:
121
-------
AT-1 . Q . D2
where AT = temperature difference
K = thermal conductivity
Q = heat rate per unit volume
D = diameter of cylinder
Thus, a higher thermal conductivity will linearly decrease the
temperature at the center of the waste and increase the amount of self-
generating heat that can be included in the waste. In addition, a higher
thermal conductivity also increases the size of the storage container
and allows for a decrease in the specific volume (liters/gram) of the
solidified waste.
Process Description
High-level Waste Solidification Processes. Four processes have been
developed for the solidification of high-level wastes. They are pot
calcination, spray solidification, phosphate glass solidification, and
fluidized bed calcination. The pot, spray, and phosphate glass processes
have been demonstrated on a full scale, engineering level by
Battelle-Northwest at Richland, Washington. °715' °716' °717' °527' °732
The fluidized bed process has been demonstrated in a large capacity
plant °694' °711 by Idaho Nuclear Corporation at Idaho Falls, Idaho.
A description of each solidification process is attached (Figures 1 and 2)
and basically, in each of these four processes, heat is applied to raise
the temperature of the waste to drive off the volatile constituents, leaving
a solid. Generally, chemicals are added to the waste stream to: (1)
reduce the melt temperature; (2) prevent the release of volatile
components; and (3) improve the properties of the final product.
Each of these processes generates additional gas and vapor streams
which contain reduced but still toxic levels of radioactivity. To
treat these waste streams for the removal of radioactivity, additional
auxiliary processing equipment is required. The auxiliary processing
system usually includes several separate cycles of evaporation plus
122
-------
SIX
ZONE
HEATING
VAPOR OUT
—^. WASTE
.FILL
LEVEL
.VIOLENT
BOILING
EARLY WALL
DEPOSITS
CALCINE
DEPOSITION
PATTERN
POT CALCINATION PROCESS
ATOMIZING
SPRAY
CALCIMER
THREE ZONE
FURNACE
MELT
WASTE
FEED NOZZLE
PULSE FILTER BLOWBACK
VAPOR OUT
ILTERS
MELTER
FURNACES
MELT RECEIVER
(Storage Pot)
SPRAY SOLIDIFICATION PROCESS
Figure 1. Pot calcination and spray solidification
123
-------
EVAPORATOR
AIRLIFT
OVERFLOW
VAPOR OUT
MfLT,
VAPOR OUT
WASTE
MELT RECEIVER
(Storage Pot)
PHOSPHATE GLASS SOLIDIFICATION
ATOMIZING GAS
GAS TO CLEAN-UP
CALCINER
FLUIDIZINr, GAS
TO STORAGE VAULTS
FLUIDIZED PEP CALCINATION
Fiaure 2. Phosphate glass solidification and fltndizpd hed calcination
124
-------
condensation followed by a stage of filtration. An acid fractionator
is also employed for the removal of nitric acid. This system is followed
by a caustic scrubber and further stages of filtration. Decontamination
factors of 109 have been obtained with liquid effluents.
The overall characteristics of the four types of solidified wastes
are summarized (Table 1) and a brief description of the design of each
process is presented below.
Pot Calcination. A batch process in which the processing vessel is
also the final container for the solidified waste. In this process the
liquid waste is added to the pot and heated in a multiple-zone furnace.
As the waste is heated, scale forms on the wall of the pot. The scale
grows in thickness until all the waste in the pot has been calcined
and reaches a temperature of 850 to 900 C. The pot is then cooled,
sealed, and taken to storage. The only major requirement for additives
to the liquid waste stream is for the retention of the volatile
components. The overall cycle capacity with a 12-inch-diameter pot is
14 liters/hr which represents an equivalent waste processing capacity
of 0.90 tonnes/day.
The basic elements required for the design of the equipment in pot
calcination are: a multiple-zone furnace for heating and cooling, a pot
to calcine the waste, and an off-gas line from the pot that can be
continuously washed down. The multiple-zone furnace must be designed to
provide different heating regions along the vertical portion of the
pot. Six zones for a pot filled with 6 ft of calcine have been found to
be adequate. The pots are made of corrosion resistant material to
survive the severe corrosion conditions during processing.
Three-hundred-series stainless steel, generally 304L or 310, has been
found to be adequate. Continuous washing of the off-gas line from
the pot is required to prevent plugging by entrained calcine or sludge.
The continuous washing can be accomplished by condensing the vapors
formed in the line.
125
-------
TABLE 1
CHARACTERISTICS OF HIGH LEVEL SOLIDIFIED WASTE0705
i-k
re
CTJ
Form
Description
Hardness
Friability
Bulk Density, g/«i
Thermal Conductivity, watts/
m- C
2
Leachabi 1 i ty , g/cm -day
Volume, liters/1000 MWd
(thermal)
Chemical Composition, mole 1
Fission product oxides
Inert metal oxides
Phosphorous oxides
Pot
Calcination
Calcine Cake
Scale
Soft
Crumbly
1.1 to 1.5
.19 to .26
10"1 to 1
1. to 2.5
15 to 80
10 to 50
-1.0
Spray
Solidification
Monolithic
Microcrystalline
Hard
Tough
2.7 to 3.3
.87 to 1.73
10'6 to 10'3
1.2 to 3
5 to 30
40 to 50
25 to 40
Phosphate
Glass
Monolithic
Glass
Very hard
Brittle
2.7 to 3.0
.69 to 1.73
10"7 to 10'4
1.5 to 5
5 to 25
10 to 30
^60
Fluidi zed
Bed
Granular
Amorphous
Moderate
Moderate
1.0 to 1.7
.17 to .43
10"1 to 1
1.5 to 5
5 to 50
10 to 90
^0
Maximum Stable Temperature, C t-900
Container Material
Stainless Steel
Phase separation at
•x.900
Mild or Stainless
Steel
Devi trifles at -^500
Mild or Stainless
Steel
•x-600
Mild or Stainless
Steel
-------
A major problem in the design of the pot calcination process is
retention of volatile components such as sulfate, ruthenium, and
mercury. Sulfate can be retained almost entirely in the calcine form
by adjusting the ratio:
Chem. equiv. (alkali + alkaline earth metals) •, ,
Chem. equiv. (nonvolatile anions)- '
Chemical additives to meet this ratio are usually calcium or sodium
nitrates. Since the above ratio also includes potentially volatile
alkali metal oxides, the following ratio should also be met:
Chem. equtv. alkali metals .- -,
Chem. equiv. (nonvolatile anions)
The volatilization of ruthenium can be decreased by the addition of
phosphate or hypophosphate ions as phosphoric acid. The retention of
mercury is difficult since it forms no compounds that are stable at
temperatures greater than 600 C. The mercury should be removed prior to
solidification by passing the waste through beds packed with copper
or aluminum. The copper beds are not feasible if the acid concentration
of the waste is greater than 0.5 M.
Spray Solidification. A continuous process in which the liquid
waste is fed through an atomizing nozzle into a heated calcine tower.
The water is evaporated, dried, and the waste is calcined to a powder as
it falls through the calciner tower. It then falls into a melter where
it is melted at temperatures of 800 C to 1,200 C. The melt in the melter
flows through an overflow weir to a storage pot which is then cooled and
sealed. Process gases flow into the adjacent filter chamber where the
calcined powder collects as dust on the porous metal filters. The dust is
periodically blown off by sudden pulses of high pressure steam or air
and falls back into the melter with the main powder stream. The overall
cycle capacity is approximately 20 liters/hr of liquid.
The basic elements required for the design of the equipment in
spray solidification are: an atomizing nozzle; a multiple-zone
furnace; an off-gas cleaning system; a continuous melter and furnace;
127
-------
and a receiver pot and multiple-zone furnace. The atomizing nozzle
is designed to break the feed up into small droplets (80 to 100 microns
diameter) to permit drying and to calcine the feed. Commercial
internal-mixing nozzles have proved satisfactory. The multiple-
zone heater in the spray calciner is designed to maintain the
temperature in the spray tower higher than the walls of the spray
tower to prevent sticking to the tower walls. The calcine dust in the
off-gases can be separated by porous stainless steel filters with an
average pore size of 65 microns and a total area of 1 sq ft/(liter)/(hr)
of feed to the calciner. The continuous melter which is
positioned below the calciner to cbllect the powder must withstand
the effects of corrosion and high temperature (up to 1,200 C). Platinum
is the only known reliable material for the melter at the present time.
Other materials such as a chromium-alumina cermet and an alloy of
50 percent chromium-50 percent nickel are being examined. The receiver
pot and multiple zone furnace designs are similar to those in pot
calcination.
In spray solidification the feed composition has a strong effect
on its performance. Additives are required so that a melt-forming
matrix is formed which is stable at high temperatures. Phosphate is
generally used as the melt former. The melt temperature limits are
adjusted using a phosphate flux with an overall cation-to-phosphate
ratio (M/P) of 2 to 3. °725 In this relationship, M is the total
metal ion equivalent and P is the total molarity of phosphorous present.
The melt temperature increases with the M/P ratio approaching 1,000 C
at a M/P ratio of 3.
Sulfate and ruthenium are the two principal products most easily
volatized. The sulfate is retained in the melt by the addition of
calcium which combines chemically with the sulfate. The ruthenium
volatility is reduced by the elimination of the melt-making flux from
the feed and the addition of all the melt-making flux to the melter.
128
-------
Phosphate Glass Solidification. A continuous glass solidification
process which is carried out in three steps: feed preparation; low
temperature denitration-evaporation; and high temperature glass
forming. The feed preparation involves the addition of glass-forming
chemicals to the liquid waste. The liquid waste which contains all
the melt-making additives is first fed to an evaporator where it is
concentrated into a thick phosphate slurry. This slurry is fed to a
continuous melter where the material is heated to 1,000 to 1,200 C to
form a molten glass. The molten glass then flows into a storage pot
which is cooled and sealed. The process capacity is limited by the
melter which.has a capacity of 10 to 12 liters/hr compared to the
evaporator capacity of 35 liters/hr. °718
The basic elements required for the design of the equipment are:
a continuous denitrator-evaporator; a melter feed system; a continuous
melter and furnace; and a receiver pot and multiple-zone furnace. The
evaporator design must be capable of handling corrosive liquids,
providing low hold-up volume and maximum suspension of solids. The
f 7V
evaporator is made of titanium and is constructed of a special design.
The melter feed system is designed to pump and meter the waste flow
between the evaporator and the melter. Special designs using either an
airlift or slurry pump have been developed. The melter requirements
are the same as those for the spray solidifier except that the heat
transfer requirements are 50 to 100 percent higher. The receiver
pot design is similar to that for pot solidification.
The chemical additives are added directly to the feed and consist
mainly of orthophosphoric acid. To form a good quality glass, one mole
of phosphate is required per mole equivalent of nonvolatile cation. °732
If the ratio is less than one, a poor grade glass is formed and, if
greater than one, the excess phosphate is evolved until the ratio is
about one. If the mole percent of oxides of the alkali metals is less than
half the total metal oxides in the melt, additional alkali metals (i.e.,
sodium nitrate) must be added to obtain a glass that forms at 850 to
1,000 C and has good handling properties.
129
-------
In the evaporator about 0.5 to 1 percent of the ruthenium will
volatile but no appreciable sulfate or phosphate volatilization has
occurred. In the melter the principal off-gases are water, nitrates,
sulfate (when present in waste), and ruthenium. When sulfate is not
present,the off-gases from the melter are combined with the off-gases
from the evaporator, or condensed separately and recycled to the
evaporator to reduce the overall off-gas activity. When sulfate
is present, steam additions are required to help recover the sulfate
and to prevent corrosion by diluting the condensate and reducing
its temperature.
Fluidized Bed Calcination. A continuous process in which the liquid
waste is converted to granular solids by heating in a fludlzed bed of
the granular solids. In this process the liquid waste 1s injected
through an atomizing nozzle into the side of a heated bed (400 to 600 C)
of granular solids. The bed is continuously fluidized by gas flowing
upward through the bed. As the waste contacts the hot bed, it is evaporated
and calcined as coatings of the bed particles. The bed particles are
continuously removed from the calciner and transported to storage bins.
The calcine material that is entrained with the process gases is removed
from the gas stream by cyclone separators or filters and transported to
the storage bins. About 60 percent of the solidified waste product is
removed by the calciner and the remainder is removed during the processing
of the off-gas stream. The overall cycle capacity is 180 to 260 liters/
hour.
The basic elements required for the design of the equipment in
the fluidized bed process are: an atomizing nozzle; a calciner and
furnace; and an off-gas cleaning system to remove the calcine dust.
The atomizing nozzle is designed to break up the feed into small
droplets to permit rapid drying without caking. Commercial
external-miKing pneumatic nozzles are used with a gas volume of 500 to
800 times the volume of the liquid feed. °694 Air or superheated
steam is used for atomization. The calciner size is 3 to 6 times
the hourly volume of liquid feed to the calciner and has a length-
130
-------
to-diameter ratio in the range of 1.5:1 to 4:1. 5 The heat
requirements for the calciner is one kilowatt for each liter per
hour of liquid feed. The heat is provided by an in-bed heat exchanger
through which sodium-potassium (NaK) circulates. The sodium is heated
externally to about 740 C and circulated by an electromagnetic pump.
The calcine dust in the off-gas stream is removed by the use of cyclones
and porous stainless steel filters with periodic pulse blowback for
on-line cleaning. The pulse blowback system is similar to the system
used in spray solidification.
Additives to the feed solution are primarily required to retain the
volatile components and not to enhance the final solidified waste
product. Ruthenium volatility is reduced to about 1 percent when sugar
is added to the feed. 715 Sulfate and fluoride is retained with the
calcine (99%) and fluoride corrosion is controlled by the addition of
calcium in stoichiometric equivalence to the amount of fluoride
present. °715
Low-Level Waste Solidification Processes. Solidification
processes have been developed for the incorporation of low-level
radioactive wastes in asphalt, cement, and polyethylene. These
treatment processes are fully described elsewhere °547' °697' °704' 1017'
and will only be briefly described here. The treatment processes for
each of the solidification processes is similar. The waste and the
solidification material are added and mixed together in a predetermined
ratio. This ratio depends on the type of waste involved. Studies
have defined compositions of asphalt and polyethylene for incorporation
in organic and inorganic waste streams. These compositions are
attached (Table 2). Tests have resulted in a ternary operating diagram
for the incorporation of filter sludge and ion exchange resins in
cement. These results are also attached (Figure 3).
131
-------
TABLE 2
ASPHALT AND POLYETHYLENE MATRIX COMPOSITION FOR ORGANIC
AND INORGANIC WASTES °547
Haste
Organic
25 (trl butyl phosphate)
25 (tr1 butyl phosphate)
25 (trlbutyl phosphate)
25 (trlbutyl phosphate)
30 (trlbutyl phosphate)
30 (trlbutyl phosphate)
30 (carbon tetrachlorlde)
30 (tetrachloroethylene)
Inorganic
40 (NaB02)
60 (NaB02)
50 [Ca(B02)2]
55 [C«(BOZ)2]
20 (NaB02)
30 (NaB02)
40 (NaB02)
20 (HaB02)
40 (NaB02)
Composition, wtX
Natrlx Material
38 (asphalt)
35 (asphalt)
37 (asphalt)
37 (asphalt)
70 (OVLT polyethylene)
70 (DYDT polyethylene)
70 (OYDT polyethylene)
70 (OYDT polyethylene)
60 (asphalt)
40 (asphalt)
50 (asphalt)
45 (asphalt)
80 (DYLT polyethylene)
70 {DYLT polyethylene)
60 (DYLT polyethylene)
80 (OYDT polyethylene)
60 (DYDT polyethylene)
Attapulglte-150: a oagneslum aluminum silicate drilling clay
Filler
37 (Attapulglte-150)*
40 (Grundlte)*
38 (Attacote)t
38 (Attaclay)'
None
None
None
None
None
None
None
None
None
None
None
None
None
produced by Minerals and Chemicals
Connents on
At 160 C
Pours
Pours
Pours
Pours
Fluid
Very fluid
Very fluid
Very fluid
Very fluid
Very fluid
Fluid
Fluid
Fluid
Fluid
Fluid
Very fluid
Very fluid
Phillpp Corp.
Project
At 25C
Finn, but
Soft and
Firm, but
Soft and
Very good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
Good
bleeds
tacky
bleeds
tacty
Grundlte: a high-strength bonding clay produced by the Illinois Clay Products Co.
'Attacote: a magnesium aluminum
Attaclay: a magnesium aluminum
silicate coating clay produced
clay produced by Minerals and
by Minerals and Chemicals PhlUpp
Chemicals Phillpp Corp.
Corp.
-------
40 60
SOLIDS (WT X)
OPERABLE AREA FILTER SLUDGE
OPERABLE AREA
ION EXCHANGE RESINS
r 40 9* "INCLUDES 5 LB OF
% 50* NaOH FOR BEAD
^ RESIN
Figure 3. Cement solidification process ternary operating diagram
for filter sludge and ion exchange resin wastes
133
-------
The solidification process for asphalt and polyethylene consists
of adding the waste to the asphalt or polyethylene in an evaporator with
a wall temperature between 130 and 160 C. Mixing is accomplished by
agitator paddles. The paddles turn with a downward forcing action that
forces the mixture out the bottom and into the disposal container. In
the solidification process for cement the waste and cement are generally
mixed directly together in 55-gal. drums.
4. PROCESS ECONOMICS
For the four high-level solidification processes only cost studies
for pot calcination were found to be available. °705 These cost studies
were made using an economics model based on a discounted cash flow tech-
nique. This type of model requires that the income received must provide
for the recovery of the investment by the establishment of a reserve account
to pay all waste management operations that remain to be completed after
all income has ceased. Using this model the costs of pot calcination
were calculated as a function of the age of the waste and size of the
storage container. These costs are presented for the case in which the
center-line temperature of the waste is held at less than 900 C when
standing in air (Figure 4). These costs ranged from $4,200/metric ton of
fuel for 1-year-old waste in 6-in.-diameter pots, to $460/metric ton of
fuel for 30-year-old waste in 24-in.-diameter pots.
Cost estimates for the incorporation of aqueous low-level waste
in asphalt and polyethylene have been made. These cost
estimates were made using a price of $0.015/1b for asphalt and
$0.20/lb for polyethylene. The total processing cost was $0.34/gal.
of waste for asphalt and $0.80/gal. of waste for polyethylene.
5. PROCESS APPLICABILITY
The high-level waste solidification processes have been developed
especially for the spent fuel wastes generated from nuclear power
134
-------
6-in - DIAMETER CYLINDER
•MM?
12-in -DIAMETER CYLINDER
'
24-in -DIAMETER CYLINDER
OF HASTES (years)
Figure 4. Pot calcination costs as a function
of the age of the waste0705
135
-------
plants. These wastes include the reactor-produced fission products as
well as the actinides or transuranium elements.
Since these solidification processes result in a waste volume
reduction their applicability to certain types of wastes can be limited
by the waste heat generation rates. This is a problem not only for the
wastes generated by the presently constructed thermal power reactors,
but the size of this problem will increase with the introduction of
the fast breeder reactors. To assess the magnitude of this problem
calculations were made to determine the maximum allowable heat that
would be stored in a particular storage container for each of the four
solidification processes. The reactor-produced fission product heat
generation rates are included as a function of time for both the current
thermal power reactors and the proposed fast breeder reactors (Figure 5).
The fast breeder reactor heat generation rates are approximately five
to ten times higher than those of the current thermal power reactors.
The maximum container heat content was calculated assuming a
maximum waste temperature of 900 C and a storage environment of 20 C
air, which corresponds to a container wall temperature of approximately
450 C. For these conditions, the heat content of a cylindrical storage
container is shown as a function of container diameter and waste thermal
conductivity (Figure 6). Combining 14iese data with those previously shown
(Figure 5) the maximum heat content and corresponding age of the fission
products (storage time before solidification) were determined for each of
the four solidification processes. These data are presented for a 6-in.-
diameter container (Table 3). For the present thermal power reactors the
required age of the fission products or the storage time before solidifi-
cation varied from 0.2 to 2.1 years and for the fast breeder reactors
from 2.7 to 9.5 years. For an 8-inch-diameter container the required
age for the breeder reactor would vary from 8.4 to greater than 10 yrs.
These data were presented for storage in air and the age or storage time of
the waste would be reduced if the containers were stored in water or in an
enclosed, air-conditioned room. These data indicate that the waste heat
content can limit the use of these solidification processes especially
136
-------
: FAST REACTOR
TJ^IOO.OOO MWd/tonne at 200 MW/tonne
PROBABLE MAXIMUM FOR THERMAL REACTOR
FUELSi45,000 MWd/tonne at 30 MW/tonne
t ; ! • '
TP
CURRENT THERMAL REACTOR
20,000 MWd/tonne at 15 MW/tonne
TIME SINCE REACTOR DISCHARGE (years) -
tt-m-n i 111 rr
10'
Figure 5. Fission product heat generation rates
0716
137
-------
600
CENTERLINE TEMPERATURE 900 C
WALL TEMPERATURE 450 C
6-1n-DIAMETER
I!
500
T
1
ft'
*
£ !!:400
ft o
<
Ji
S-in-'DIAMETER
il
tt
-
-rrt
I
:..w ....200
to
*t < 4.
12-in-DIAMETER
1
jlOOc
I
ffi
I
•it
fcffl
11
i
m
0.4
tftpj
0.8
:-T
1.2
1.6
2.0
WASTE THERMAL CONDUCTIVITY, (watts/m - C)
Figure 6. Waste heat content as a function of waste
thermal conductivity and pot diameter
138
-------
TABLE 3
ALLOWABLE HEAT CONTENT OF HIGH-LEVEL SOLIDIFIED WASTE
CJ
CD
Thermal Conductivity, watts/m- C
Range
Average
Maximum Heat Content (6" Dia. Pot)
Watts/ liter
*Watts/ metric ton of fuel
Age of Fission Products, years
Current Thermal Reactor
Maximum Thermal Reactor
Fast Reactor
Pot
Calcination
.19 to .26
.23
75
2250
2.1
5.5
9.5
Spray
Solidification
.87 to 1.73
1.30
425
12.750
.2
1.0
2.7
Phosphate
Glass
.69 to 1.73
1.21
395
11,850
.2
1.0
2.8
Fluidized
Bed
.17 to .43
.30
ICO
3000
1.7
4.2
6.9
' *Solidified Waste Volume = 30 liters/tonne
-------
with the fast breeder reactor fuel wastes. To overcome this problem, several
of the more active fission products such as Sr , Ru , Cs , and Ce
may have to be removed from the waste prior to solidification.
A comparison of the four solidification processes is presented
(Table 4). From the standpoint of process and flexibility, pot calcination
is superior to the other three processes, but the other three processes
are superior in terms of the solidified waste product. A comparison
shows that spray solidification or phosphate glass solidification come
closer to matching the desired characteristics of the solidified waste
developed in Section 3. It must be recognized that solidification is
only part of the total disposal scheme and that the final comparison or
selection must be made in conjunction with the disposal method to be
utilized.
The low-level waste solidification processes are applicable to
those wastes that are disposed of by direct burial in the ground. In
general, these solidification processes are more adaptable to the
short-lived isotopes (6 months to 30 yr) and less hazardous materials.
The source of these types of wastes are the secondary streams generated
at nuclear power plants and fuel reprocessing facilities, along with the
wastes resulting from research laboratories and medical and industrial
applications of radioisotopes.
Cement, asphalt, and polyethylene are used for the solidification of
these types of wastes. Cement products are only moderately insoluble
compared to the asphalt and polyethylene products. The leach rates of
asphalt and polyethylene products are 50 to 100 times lower than those
of cement products containing similar amounts of waste salt. A comparison
between a good asphalt product (about 60 Wt. % salts) and a good cement
product (about 25 Wt. % salts) shows that the weight percent of Cs137
leached in 10 weeks from the cement product was 58 percent and from the
asphalt product was 0.56 percent.
140
-------
TABLE 4
COMPARISON OF POT, SPRAY, PHOSPHATE GLASS, AND FLUIDIZED BED SOLIDIFICATION PROCESSES
Pot Calcination
Advantages Disadvantages
1. System Simple
2. Type Batch
3. Amenability to Scale Up Limited
4. Flexibility for Haste
Variations Minimum
5. Specific Restrictions None
6. Product teachability High
7. Product Thermal
Conductivity Low
8. Product Toughness Soft
Phosphate Glass Fluldlzed Bed
Spraj Solidification Solidification Calcination
Advantages Disadvantages Advantages
Complex
Continuous Continuous
Good Good
Moderate
Requires
platinum
melter
Requires good
waste flow
control
High ruthenium
volatilization
Low Lowest
Highest High
Tough
Disadvantages Advantages
Complex
Continuous
Fair
Moderate
Requires
platinum
nelter
Requires good
concentrate
flow control
Does not
retain sulfate
In solid
Brittle Moderate
Disadvantages
Complex
Moderate
Requires complex
heating scheme
Amount of self-
generatlng heat
limited
High
Low
-------
Cement can be used with almost all waste streams but the products
formed are Inferior to the asphalt and polyethylene products in terms
of solubility. The integrity of the cement product also decreases
greatly as the concentration of soluble salts increases. 1017 When low
concentrations of waste salts are used, the volume of the final cement
product is much larger (60%) than the original waste volume. With
asphalt products volume reductions of almost two have been obtained. °697
In addition, most operations with cement are dusty and cumbersome and
good mixing is difficult to achieve.
In general, asphalt and polyethylene are better materials than
cement for solidifying low-level wastes except for waste streams
that contain significant amounts of oxidizing agents. Economically,
asphalt is preferred to polyethylene since its processing costs are
less. Asphalt is not recommended for wastes that contain high
concentrations of oxidizing salts. Tests have shown that mixtures
containing relatively large amounts of nitrate or nitrite burn
vigorously and that a mixture containing 65.6 Ut. percent nitrate
(90 Wt. % NaNO*) with a porosity of approximately 50 percent can be
detonated with blasting gelatin. Polyethylene can probably be used
with oxidizing salts, but further studies are needed to assess the
safety of the system. For organic wastes polyethylene is recommended
since a high percentage of the wastes can be accommodated without the
use of an inert filler that is required with the asphalt-organic
mixture.
142
-------
6. REFERENCES
0547. Fitzgerald, C. L., R. E. Godlee, and R. E. Blanco. The feasibility
of incorporating radioactive wastes in asphalt or polyethylene.
Nuclear Applications and Technology. 9: 821-828, Dec. 1970.
0551. Smith, C. W., and W. B. Bigge. LWR fuel recovery: prospects, plans
and problems. Nuclear News, 62-66, July 1970.
0562. AEC Manual, Chapter 0529. Safety standards for the packaging of
radioactive and fissle materials. Aug. 22, 1966. 17 p.
0563. Recommendations of the International Commission on Radiological
Protection, ICRP Publication 2. Report on Committee II on
permissible dose for internal radiation (1959). Peryamon Press,
1959. 232 p.
0694. Lohse, G. E., and M. P. Hules. Second processing campaign in the
waste calcining facility. Idaho Nuclear Corporation, IN-1344,
Mar. 1970. 48 p.
0697. Godbee, H. W., R. E. Blanco et al. Laboratory development of
a process for incorporating of radioactive waste solutions and
slurries in emulsified asphalt. Oak Ridge National Laboratory,
ORNL-4003, 1967. 45 p.
0704. Goode, J. H., and J. R. Flanary. Fixation of intermediate-level
radioactive waste concentrates in asphalt: hot-cell evaluation.
Oak Ridge National Laboratory, ORNL-4059, 1968. 32 p.
0705. Staff of the Oak Ridge National Laboratory. Siting of fuel
reprocessing plants and waste management facilities.
ORNL-4451, 1970. 500 p.
0711. Thompson, T. K. Development of in-bed combustion heating for
calcination of radioactive wastes. Idaho Nuclear Corporation,
IN-1278, 1968. 25 p.
0715. Schneider, K. J. Status of technology in the United States for
solidification of highly radioactive liquid wastes.
Battelle Northwest, BNWL-820, 1968. 63 p.
0716. Schneider, K. J. Waste solidification program, process technology-
pot spray, and phosphate glass solidification processes.
V. 1. 1969. 232 p.
143
-------
REFERENCES (CONTINUED)
0717. Blusewltz, A. G., J. E. Mendel, K. J. Schneider, and R. J. Thompson.
Interim status report on the waste solidification demonstration
program. Battell Northwest, BNWL-1083, 1969. 80 p.
0718. McElroy, J. L., J. N. Hartley, and K. J. Schneider. Waste
solidification program, phosphate glass solidification
performance during first radioactive tests in waste
solidification engineering prototypes. V. 5. Battelle
Northwest, BNWL-1185, 1970. 110 p.
0725. Band, W. R. et al. Waste solidification program, spray
solidification performance during first radioactive tests
in waste solidification engineering prototypes. V. 6.
Battelle Northwest, BNWL-1391, 1970. 110 p.
0732. McElroy, J. L. et al. Waste solidification program,
phosphate glass solidification performance during final
radioactive tests in waste solidification engineering
prototypes. V. 7. Battelle Northwest, BNWL-1541, 1971. 150 p.
1017. Sodbee, H. W., J. H. Goode, and R. E. Blanco. Development of a
process for incorporation of radioactive waste solutions
and slurries in emulsified asphalt. Environmental Science
and Technology, 2(11) .-1034-1040, Nov. 1963.
2149. Code of Federal Regulations. Title 10—Atomic Energy (Revised
as of January 1, 1971). Washington, U. S. Government Printing
Office, 1971. 429 p.
144
------- |