Figure 1.
Extraction and stripper columns
I
Figure 2.
Agitator mixed extraction and stripper columns
JOE DYKSTRA, 8. H. THOMPSON, and R. J. CLOUSE Oak Ridge Gaseous Diffusion Plant, Union Carbide Nuclear Co., Oak Ridge, Tenn.
Solvent Extraction System for Enriched UraniilJun The extraction column is more effective than the stripper column. Further work is needed to improve stripper column capacity
C O N S I D E R A B L E quantities of solutions originate from the cleaning of uranium compounds from process equipment a t the Oak Ridge Gaseous Diffusion Plant. These solutions contain widely varying quantities of uranium salts and impurities, and must be processed and purified in a solvent extraction system capable of recovering uranium to less than 5 p.p.m. loss. The da,ta and column evaluation reported are associated with early phases
of the investigation, and consequently may not be conclusive.
Solution Recovery Facility The uranium recovery facility employs mixer-agitated type contactors, 5 inches in diameter, with a n extraction medium of tributyl phosphate. This system (Figure 3) consists of two parallel extraction lines with a design capacity of IO pounds of uranium per
hour. Each line includes a pre-evaporator, amixer-agitated extraction column, a mixer-agitated stripper column, a postevaporator, a drum dryer, and a calciner, all interconnected to form a The continuous recovery unit (7). system is designed to process uranyl nitrate solutions having a wide range of uranium-235 concentrations. During 2 'years of operation the system has recovered uranium from solutions containing from 2 to 40 grams of uranium per VOL. 50, NO. 2
FEBRUARY 1958
161
SANITARY HtO
3.N NITRIC ACID SCRUB
I
MIXER AGITATED EXTRACTION COLUMN
C L E A N ORGAN'C
MIXER AGITATED STRIPPER
POST EVAPORATOR
I!
COLUMN
URANIUM OXIDE
(UO, 8
Figure
U,Or)
3. General recovery system
Table I.
Results of Tests 20-GaLa 35-Gal." Extraction Column 202 182 20 8.7 32
Mixer speed, single-stage flood, r.p.m. Most efficient col. operation, r.p.m. Below flood, r.p.m. Best height equiv. to theoretical stage value, inches Required to meet discard specification, inches
65-Gal.(l
50-Gal?
65-Gal.b
174 158 16 13.5 56
158 148 10 26.9
158 148 10 28.3
158 148 10 28.9
173 164
143 134 9 32
143 134
143 134 9 29.4
Stripper Column 204 172 32 23
Mixer speed, flooding, r.p.m. Most efficient col. operation, r.p.m. Below flood, r.p.m. Best height equiv. to theoretical stage value, inches a
9
24
9
43.1
4-inch horizontal baffle spacing. 2-, 3-, and 4-inch horizontal baffle spacing.
liter, with a maximum impurity content of 40 grams per liter. Even though the feed solutions have shown a wide range of densities, the raffinate discharged from the extraction column has consistently met the uranium discard specification of less than 5 p.p.m.
Contactor Design
FEED CONCENTRATION- 20 GRAMS U / L I T E R FLOW RATIO, AQUEOUS/ORGANIC/SCRUB - 5/5/1
55 /
60 /
65--70 75
/
__
---__
___-
----
-1
2, 3, 4" STAGES 5 0 G P H FEED e 6 5 G P H FEED
I -
i
A
/
,1
pI-----
I I
100 Figure 4.
162
4" STAGES 0 2 0 G P H FEED X 35 GPH FEED A 65 GPH F E E D
125 150 175 IMPELLER SPEED - RPM Extraction column experimental d a t a
INDUSTRIAL AND ENGINEERING CHEMISTRY
200
225
Extraction Column. The extraction column is an internally baffled, multistage, mixer-agitated type contactor (Figure 1) designed for mass transfer of uranium from the continuous nitric acid waste solution to the organic phase with a normal solvent to acid feed ratio to 1 to 1 (2). This column is 20 feet in length and 5 inches in diameter, and is constructed of borosilicate glass pipe to facilitate operational control. Settling zones, fabricated from 24-inch lengths of 5-inch stainless steel pipe, are installed a t the top and bottom of the glass columns to effect phase separation. INTERIOR COMPONENTS. A shaft with 48 flat-bladed mixers spaced on 4-inch centers.
NUCLEAR TECHNOLOGY Four vertical baffles placed a t 90° intervals around the inner circumference of the column. Fifty-two circular horizontal baffles. The mixer and bafRe subassemblies together form 48 vertically stacked stages, each consisting of two horizontal baffles with a centered mixer. Stripper Column. The stripper and extraction columns (Figure 2) are of identical design. The stripper column effects rhass transfer of uranyl nitrate from the organic to the continuous water phase with a solvent-water feed ratio of 1 to 1. Interface Control. The columns operate with a continuous aqueous phase; therefore the interface level is located a t the top. Two immersed air bubblers, positioned 5 l / 2 inches above and below the selected level, control the interface to rt2 inches. Mixer Drive. The mixer shaft is powered with a '/(-hp. motor. The speed is regulated with an adjustable output rectifier and may be controlled between 0 and 400 r.p.m.
FLOW RATIO, AOUEOUS/ORGANIC
- l/l
50 GPH FEED 65 GPH FEED
75
I25 150 I75 IMPELLER SPEED
100 Figure 5.
-
225
200 RPM
Stripper column experimental d a t a
Column Evaluation The purpose of this inbestigation was to determine the effects of total column throughput and agitator speeds on column performance. Combined column flows were to range from 40 to 143 gallons per hour. As the initial step in this investigation, the following control conditions were established.
System flow ratios were maintained a t (organic-aqueous-strip-scrub). The scrub solution was 3.5 f 0.1N nitric acid a t room temperature. The column feed was maintained a t 3.5 f 0.1N nitric acid with a uranium concentration of 20 grams of uranium per liter and a n average total impurity
5/5/5/1
URANIUM Figure 6.
CONCENTRATION
content of 346 p.p.m. Typical purity analyses in grams per liter were : Aluminum Calcium Magnesium Copper
Nickel
0.1 0.2 0.02 0.006 0.02
IN AQUEOUS, PPM
Extraction column profile experimental data VOL. 50, NO. 2
FEBRUARY 1958
163
URANIUM
Figure 7.
CONCENTRATION
AQUEOUS, PPM
Stripper column profile experimental d a t a
The solvent phase consisted of a mixture of 20 & 0.57, rributyl phosphate and 80% Varsol. The solvent temperatures were maintained a t 70' to 75' F. The uranium concentration of the raffinate was maintained a t less than 5 p.p.m. to meet the discard specificatign. This specification was not met when determining maeirnum column throughput. The mixer speed was maintained a t selected rates with a maximum deviation of +2.5 r.p.m. Steady-state operation was considered to exist when three consecutive sets of samples taken from all sample points showing significant quantities of uranium agreed within 1 5 7 , (Figure 1). All column flows were maintained within f5Yo. ,4 continuous phase of 120" F. sanitary water was used 'in the stripper column. Flood condition was considered to exist when a single-stage flood was observed. The first observation was taken a t 10 r.p.m. below flood. Further observations were taken at increments of reduced speed. A single observation consisted of three sets of samples withdrawn from sample points 1 through 18 (Figure 1).
in the extraction column and stripper column. Twenty Gallons per Hour. This test was made with a total throughput of 44 gallons per hour for the extraction column and 40 gallons per hour for the stripper column (Table I). Thirty-Five Gallons per Hour. Conditions included a total extraction column flow of ?7 gallons per hour and a stripper column flow of 70 gallons per hour. Sixty-Five Gallons per Hour. Con-
220-
I
'
1 I
Test Data
The first group of extraction and stripper column evaluation tests (Figures 4 and 5) were made with the horizontal baffles spaced on 4-inch centers, with aqueous feed throughputs of approximately 20, 35, and 65 gallons per hour
1 64
IN
1
1 140b
; I I
I
220
1
~
FLO'A RATIO, ORGANlC/AQUEOUS/SCRUE -5/5/1
2
a 9c:200t
180; -1 _I
OPERABLE
w
l
r
l
a
CONDITIONS
FEED CONCEUTRATON - 20 G R I N S U/LITER FLOW RATIO, ORGAfUIC/AQUEOUS/SCRUE- 5/5/l
160-
,
I
2 50 75
25 AQUEOUS FLOW RATE-GPH.
--
1
FEED COSCENTRITION - 25 ORAVIS U/LITER ~
Figure 8. Extraction column experimental d a t a
INDUSTRIAL AND ENGINEERING CHEMISTRY
ditions included a total flow of 143 gallons per hour for the extraction column and 130 gallons per hour for the stripper column. The data show that with the extraction column the height equivalent to a theoretical stage value consistently increased below flood. The majority of the raffinate results obtained during this test were greater than 5 p.p.m., indicating the column operated near maximum capacity. The increased quantity of uranium retained in the stripped organic phase indicates that the stripper column had approached capacity. The second group of extraction and stripper column evaluation tests were made with a new horizontal baffle arrangement. The new baffle arrangement for the extraction column included twelve 2-inch stages a t the bottom of the column, above which were six 3-inch stages and above these eighteen 4-inch stages. The baffle arrangement in the stripper column included twelve 2-inch stages in the bottom of the column, above which were six 3-inch stages and above these thirty-two 4-inch stages. The inside diameter of the horizontal baffle was changed from 3 to 3.5 inches. The data obtained from this test are also shown in Figures 4 and 5 and include tests made a t 50 and 65 gallons per hour aqueous flow. Fifty Gallons per Hour. These data were taken at total flow rates of 110 gallons per hour in the extraction column and 100 gallons per hour in the stripper column. During this test, premature flooding was experienced in both the extraction and stripper columns. The test, therefore, does not appear to be too reliable, except for the small segment of these curves shown in Figures 4 and 5 as a solid black line. Sixty-Five Gallons per Hour. This
1
'\
I
\ COUDITIONS
I1
)\\
OPERABLE CONDITIONS
~A
I40 + 0
i
I\\
~-
~
1
25 50 75 AQUEOUS RATE- GPH
Figure 9. Stripper column experimental data
NUCLEAR TECHNOLOGY Table 11.
taken with the horizontal baffles spaced on 4-inch centers. Figures 6 and 7 depict the uranium concentrations obtained a t different levels in the extraction and stripper columns when each column was operating under optimum conditions for the four aqueous throughputs. The abscissa shows uranium concentration in parts per million and the ordinate, the length of column necessary to reach the desired discard specification. Flood data experienced during the first group of tests using horizontal baffles spacing of 4 inches is shown as Figure 8 for the extraction column and as Figure 9 for the stripper column. The third group of tests were made to determine the effects of scrub solution throughputs in the scrub section of the extraction column. During these tests, 20 grams per liter uranium solution was used and contaminants such as aluminum, copper, iron, sodium, and nickel were added so that scrub characteristics could be studied. The effects of aqueous feed throughputs are shown in Table 111 and the decontamination factors in Table IV. Maintenance of uranium discard specification in the raffinate was essential
Equilibrium Data
(20 grams uranium/liter)
Equilibrium Organic, Aqueous, CoeffiMg. U/M1. Mg. U/Ml. cient Extraction Equilibrium Data 15.00 1.43 0.071 0.0039 0.00027
1E
2E 3E
4E 5E
1.09 0.050 0.0026 0.00018 0.00005
13.7 28.6 27.3 21.7 5.4
Stripper Equilibrium Data IST 2ST 3ST 4ST 5ST 6ST 7ST
21.0 5.56 0.10 0.05 0.0025 0.0015
21.0 21.0 4.20 0.17 0.032 0.048