J. Y. OLDSHUE Mixing Equipment Co., Inc., Rochester, N. Y.
Role of Turbine Impellers in Aeration of Activated Sludge The effect of variables-rate oxygen level-has
of gas flow, rate of oxygen uptake, and dissolved
been studied as applied to gas-liquid contact in the aeration
of activated sludge.
These results in a relatke sense can be applied to other
fermentation processes. The final decision as to the most effective combination
of air horsepower, mixer horsepower, and tank proportions for a given fermentation must depend on the economics involved.
Tm
biological oxidation of waste material depends upon supplying oxygen in sufficient quantity and concentration to the solid organisms suspended in the liquid. This is essentially the same gasliquid mass transfer mechanism that exists in all fermentation systems. The effect of various mixing variables on the aeration of activated sludge has been measured. O n the basis of this information, the most effective combination of air flow and mixer horsepower and the most effective tank shape for a given process result can also be predicted. The final decision on the best combination for a given fermentation must be considered in the light of the economics involved in each plant. Available Data on Aeration of Biological Waste
Throughout this discussion, the mixing vessels are considered to be equipped with standard baffies. For vertical,
2 1 94
cylindrical tanks, standard baffles are four in number and one twelfth as wide as the tank diameter. For square tanks, and tanks up to a length-width ratio of 11/2 to 1, two baffles are used; the width of each baffle is one twelfth of the tank width and the baffle is placed at the mid-point of the length dimension (Figure 1). The impellers are flat-bladed turbines, shown in Figure 2. Gas is admitted through a sparge ring beneath the lower impeller; the diameter of the sparge ring is equal to the impeller diameter. Gas flow is expressed as a superficial gas velocity, based on the cross-sectional area and at the average temperature and pressure in the tank. Data on the absorption coefficient, KLa, for the oxidation of an activated sludge are summarized in Figure 3. KLa is the rate of oxygen uptake divided by the difference in average oxygen concentration between the gas and liquid. These data are from two sources, the
INDUSTRIAL AND ENGINEERING CHEMISTRY
work of Eckenfelder ( 3 ) who used a 4-inch flat-bladed turbine and a 30-inch depth of submergence, the data collected in the research laboratory at the Mixing Equipment Co. (5, 7 ) , on the oxidation of activated milk sludge in a tank 10 feet in diameter with a liquid depth of 12 feet. Various sizes of flatbladed turbines, horsepower levels, and gas rates were used through the system. The results of these measurements are showm in Figure 3, which gives the smoothed data points from both sources of data, with sufficient accuracy for industrial calculation for a D I T ratio of 0.2, and liquid depths from 10 to 40 feet. For any gas-liquid contacting system, a curve similar to Figure 3 must be obtained. At present absorption coefficients, KLa, cannot be correlated from one chemical system to another. Thus, absorption data must be obtained in general for each new chemical system to be investigated in the mixing process.
IMPROVED FERMENTATION EQUIPMENT & DESIGN
BAFFLES
\
\
Figure 1.
Position of baffles and turbines
Cylinder mixing tank, four baffles, each T/12 wide
However, because of the large amount of data on gas absorption coefficients in sodium sulfite systems (2, 6) and in submerged fermentations (7, 4), it is possible to obtain necessary data with relatively few experimental runs. This report does not examine the data obtained for Figure 3 in great detail. Rather, it examines what can be done with this type of data in terms of working out practical industrial processes from this basic laboratory information.
O2uptake rate = K1,a
[c*
- 21 (l)
where c* is the concentration of oxygen in equilibrium with the average concentration of oxygen in the gas phase. Figure 4 shows rate of oxygen uptake, in parts per million per hour, as a function of air rate and mixer horsepower
for the case of a 20-foot-diameter tank with a 12-foot liquid depth and a 48inch-diameter turbine. I n a curve of this type, the partial pressure of the gas phase is influenced by liquid depth, and the contact time between the gas phase and the liquid is a function of liquid depth, so that this curve is specific for the dimensions listed above.
Design of an Aeration System
The curve shown in Figure 3 does not give information in the proper units for final process design. I t is necessary to convert this information to a mass transfer rate, pounds per hour of oxygen transferred per unit volume of tank, or as is common in biological systems, to express the rate in parts per million per hour. I n order to do this, it is necessary to determine the dissolved oxygen level to be carried in the system. I n activated sludge systems, the dissolved oxygen level must be above 0.5 p.p.m. in order to maintain satisfactory biological life. For purposes of illustration, 2 p.p.m. is set as the desired level of oxygen concentration. Any change from this value will cause a marked change in the preparation of the absorption rate curve. This represents a dissolved oxygen level of about 20Yo of the saturation value. By setting dissolved oxygen = c = 2.0
Figure 2.
Flat-bladed turbine VOL. 48, NO. 12
DECEMBER 1956
21 95
100
4;
I
I
1
' 1 1 1
TANK DIA. - 20FT.
1000
I
50 0 ~ 0 . 0 F2T / SEC
v
0.01
0
Y
t-
0.001
z
w
I \
0.004 0.002
_I
F' = 0 2 F T /SEC
K
I a 200 a
0.01
w
0.004
3 !
100
c
0 LL
0.002
W Y
LL
w
0.001
2 a
0 0
50
3
z 0
Z
tU 0
a
W
v,
0
c3
>X
m a
20
IO 2.5
HORSEPOWER
5
IO
.20
30
HORSEPOWER Figure 3. Absorption coefficient for waste treatment as function of horsepower and air r a t e
With a biological system, the rate of oxygen uptake is determined by the character of the growth, the character of the feed medium. the concentration of solids and the fluid regime created by the mixer. The concentration of activated sludge in a waste-treatment system is set at a nominal level based on experience with other parts of the system, and on uptake rates that are common in other types of aerators. However, in a curve similar to that in Figure 4, the problem of setting this uptake rate is largely one of the biology involved and requirements for settling, filtering, and other steps in the process. The major consideration is how mixer variables and air variables are affected by the selection of uptake rates. This can be predicted with accuracy from Figure 4. There are few data a t the present time on the effect of mixer variables on the rate of sludge uptake. I t is known that the rate of oxygen uptake is greater in a tank with a fluid mixer than if it is determined by a polarographic instrument. The design on the basis of uptake rates measured by standard procedures is conservative when applied to a turbine-impeller aerator.
combinations of mixer horsepower and air rate are possible. Thus, it is necessary to examine each uptake rate for a complete mixer design and then consider the optimum design for each case. Case I. Constant Uptake Rate. The rate of oxygen uptake can be calculated from a knowledge of the desired concentration of mixed liquor solids, the sludge loading, the sludge uptake rate per unit concentration, and the requirement of the sludge for oxygen. The rate of 80 p.p.m. of oxygen per hour (Table I) is relatively low, and several combinations of mixer horsepower and air horsepower can be used
Table I.
Effect of Uptake Rate o n Mixer D e s i g n at Constant Liquid Depth
As an illustration of how these data can be used for a particular case, three different uptake rates are considered. At each uptake rate, several different
2 196
Figure 4. Relation of oxygen uptake to mixer horsepower and air rate
INDUSTRIAL AND ENGINEERING CHEMISTRY
for the same uptake rate. It is even possible in this case to carry rhe analysis down to zero mixer horsepower for the purpose of comparison. This assumes approximatel>- 5% oxygen efficiency at zero mixer horsepower. I n Table I,A, the compressor horsepower has been calculated as 1.5 times the isothermal horsepower to supply the volume of air required at 5 pounds per square inch gage. The optimum combination of mixer horsepower and compressor horsepower is determined from two considerations, the minimum initial cost and the minimum operating cost.
Relation of Mixer Horsepower and Air Horsepower to Oxygen Uptake D.O., 2 p.p.m. Tank. 20 feet in diameter, 12 feet deep
Air Flow Compressor Oxygen Total Cu. ft./min. Ft./sec. HP. Eff., % HP. Uptake Rate 80 P.P.M. per Hour, 120 Lb. Oxygen per Day per 1000 Cu. Feet 12 5 12 0.02 0 360 6 10 9 0.01 3 180 2.5 25 10.5 0.004 8 72
Mixer
HP. A.
24
36
0 * 002
1.2
50
25.2
B. Uptake Rate 160 P.P.M. per Hour, 240 Lb. Oxygen per Day per 1000 Cu. Feet 0 7 14 30
720 360 180 108
0.04 0.02 0.01 0.007
24 12 6 4
5 10 20 25
24 19 20 34
C. Uptake Rate 40 P.P.M. per Hour, 60 Lb. Oxygen per Day per 1000 Cu. Feet 0 4
a
180
36 18
0.01 0.002 0.001
6 1.2 0.6
5 25
50
6
5 8.6
I M P R O V E D FERMENTATION EQUIPMENT & DESIGN
I
The minimum operating cost can be found by referring to Table 1,A. This considers the total horsepower used by the system; the minimum total is somewhere between 3 and 8 hp. for the mixer in this case. The minimum initial cost can be found by obtaining figures for the cost of the mixers plus installation and the cost of the air compression plus installation and plotting these costs as a function of mixer horsepower. This is shown schematically by Oldshue ( 5 ) . By plotting the total costs on the same curve, it is seen that the minimum initial cost occurs somewhere between 6 and 10 mixer hp. Oxygen efficiency has little to do with the final selection. Because the cost of the air itself is zero, only the cost of compressing need be considered. I n most activated sludge operations, 10 to 20% oxygen efficiency is obtained a t the minimum combination of mixer horsepower and air horsepower. Case 11. Variable Oxygen Uptake Rates. I n the design of a n aeration tank, various combinations of mixed liquor solids concentration and sludge loading can be used. Many of these figures have been set by the capacity of previous types of aeration devices. However, when fluid mixers are used, a tremendous latitude in rate of sludge uptake is available. I n sludge digestion tanks the rates of oxygen uptake are varied, but usually relatively low. Several mixer and air flow combinations can give the various uptake rates required. With sufficient information on these variables, the most economical combination of mixer and horsepower for each of several uptake rates can be obtained, and the most effective over-all plant design set up. Table I,B, shows various combinations of mixer horsepower and air horsepower for an oxygen uptake value of 160 p.p.m. per hour. The sort of plot shown in Figure 5 can be worked out to give the optimum combination of air flow and mixer horsepower: around 7 to 15 mixer hp. and 12 to 6 air hp. For an uptake rate of 40 p.p.m. per hour, Table I,C, shows various air and mixer combinations; the same type of optimum selection can be made.
Tank Height, Ft
.
Table 111. Effect of Tank Proportions F. SuperJicial Uas Velocity, Oxygen Air Pressure, Cu f t . / Ft./Sec. Ejf.? % Lb./Sq. Inch Qage min.
A. Constant Superficial Gas Velocity, 10 20 40
0.011 0.01 0.01.
9 17 25
10 20 40
0.005 0.01. 0.01.5
17 17 17
B.
Rotios
Air hp.
Mixer hp.
0.01 Foot per Second
4.3 8.6 17.3
1.8 1.0 0.65
0.95 1.00 1.05
1.4 1.0 0.64
1.0 1.0 1.0
0.52 1.0 1.4
2.1 1.0 0.43
Constant Volume of Air 4.3
8.6 17.3
Table I1 compares the optimum selection of mixer horsepower and air horsepower for oxygen u,Btake of 40, 80, and 160 p.p.m. per hour. These uptake rates are determined largely by solids concentration and sludge loading. Thus, aeration capacity need not be the limiting factor. Effect of Ratio Liqluid Depth to Tank Diameter
One of the most common questions confronting the designer is the proper shape of tank. I n most mixing operations, the ratio of liquid depth to tank diameter should be 1. However, in the aeration of activated sludge, it is often desired to maintain a liquid depth of 10 to 15 feet of sludge, so that relatively low pressure blowers can be used. With tank volumes from 40,000 to 400,000 gallons, the tanks will often have diameters or widths much larger than 15 feet. Multiple mixers must be used. Many industrial fermentations are carried out in smaller tanks. T h e limitation of 15-foot depth does not apply. In order to make a comparison, the following assumptions have been made : constant tank volume; constant process result, 100 p.p.m. per hour of oxygen uptake rate; and constant dissolved oxygen level of 2.0 p.p.m. per hour. Case 111. Constant Superficial Gas Velocity. Data from Figure 4 are used,
--
and the gas velocity is set a t 0.01 foot per second. Three different liquid depths are considered: 10, 20, and 40 feet. The 20-foot depth is considered as having a ratio of 1.0. At a 20-foot liquid depth, the oxygen efficiency is 17y0, the air pressure required at the bottom of the tank is 8.6 pounds per square inch gage, and the volume of air, air horsepower, and mixer horsepower are all given the value of 1.0 in Table II1,A. If the liquid depth is decreased to 10 feet, the oxygen efficiency drops to 9%) because the contact time between the gas and the liquid is less. The air pressure required in the bottom of the tank drops to 4.3 pounds per square inch gage. The volume of air required goes up SO%, because the superficial gas velocity. is constant and the area has gone up to maintain the same total tank volume. However, the total air horsepower has dropped 5y0, while the mixer horsepower has gone up 4oYO. Thus the total mixer horsepower is higher a t the lower depth of 10 feet. Increasing the liquid depth to 40 feet increases the oxygen efficiency to 25%. The required air pressure is 17.3 pounds per square inch gage. The volume of air has dropped 35%) while the air horsepower has gone up 5y0. T h e mixer horsepower has dropped 36%. Thus the effect of increasing liquid depth is to:
LIQUID
RS
Table II. Aerator Design for Different Initial Uptake Rates (Tank 20-foot diameter, 12-foot liquid depth. Summary from Table I)
Uptake Rate, P.P.M./Hr.
Mker Hp.
Compressor HP.
Total
40 80 160
4 8 12
1.2 2.5 6
5.2 10.5 18.0
HP. Figure 5.
Mechanism of waste treatment
-
c) R, = p.p.m./hour = K L u(c* Re = uptake rate = f(c, area)
VOL. 48, NO. 12
DECEMBER 1956
21 97
Increase the oxygen efficiency. Increase the air pressure requirement at the bottom of the tank. Decrease the total volume of air required at a constant superficial gas velocity. Maintain approximately the same air horsepower within 2 ~ 5 7 7 ~ . Decrease the mixer horsepower. Table III,A, indicates that tank height should be relatively high. However, as the air pressure requirement at the bottom of the tank increases, the cost of the compressor increases, owing to the different type of compressor required. I n terms of the mixer, increased tank height means increased shaft length and there are certain limits on these from a practical operating standpoint. I t is probably desirable to go to relatively tall tanks with small diameters, as long as severe mechanical limitations on the air compressor and mixer are not introduced and excessive superficial gas velocities do not result. Case IV. Constant Volume of Air. For the example of a constant volume of air being admitted to the system, Table III,B, shows an oxygen efficiency of 17yofor a 20-foot liquid depth with an air pressure requirement of 8.6 pounds per square inch gage, and values of 1.0 for the relative volume of air required, 1.0 for the air horsepower, and 1.Ofor the mixer horsepower. Going to a lower liquid level of 10 feet, the superficial gas velocity drops to 0.005 foot per second, because of the increase in area for the tank to maintain constant liquid volume. The oxygen efficiency remains at 17YG. Air pressure drops to 4.3 pounds per square inch gage, the volume of air is the same, and the air horsepower drops 48y0. The mixer horsepower, however, increases
lloyo.
Raising the liquid depth to 40 feet increases the superficial gas velocity to 0.015 foot per second, gives 17YGoxygen efficiency, and increases the air pressure requirement to 17.3 pounds per square inch gage. The volume of air remains the same, while the air horsepower goes up 40%. The mixer horsepower drops
57%. Thus, on this basis of comparison, the taller tank with the smaller diameter requires a lower total horsepower for the same mixing process. laboratory and Pilot Plant Studies
If data such as those shown in Figure 3 are available, no further pilot plant or laboratory studies are required. If this information is not available, laboratory work must be considered. The most basic measurement-that of the absorption coefficient, KLa-requires measurement of the rate of oxygen uptake, usually in parts per million per 2 1 98
hour, and of the dissolved oxygen level. The uptake rate can be measured by polarographic methods. in which the samples are removed from the tank and the depletion of the oxygen is measured at various times. This. however, gives the uptake rate under zero mixer horsepower input, and does not take into account any possible aid by the mixer on the rate of oxygen uptake. Instead of the polarograph. other standard determinations can be used. When applicable oxygen is absorbed by the liquid, it is possible to measure the concentration of oxygen in the exit gas. which will give the rate of oxygen uptake in the system under turbulent fluid conditions. The polarographic method for the measurement of dissolved oxygen is probably best. Specific Pilot Plant Procedures. To obtain data for Figure 3, six runs are normally sufficient. Three runs are made at a constant air rate at various mixer horsepower inputs; then at one of these mixer horsepower values, air rates higher and lower than the first air rate are used. Usually, with this information, sufficient accuracy can be obtained for projection to full scale equipment. ACTIVATED SLCDGE PROCEDURES. Four runs should be made at approximately 0.1, 0.5, 1.5, and 5.0 hp. per 1000 gallons, at an air flow of 0.005 foot per second. At the horsepower level of 0.5 hp. per 1000 gallons, at a n air flow of 0.005 foot per second. At the horsepower level of 0.5 hp. per 1000 gallons. the gas rate should be varied at 0.01 and 0.15 foot per second in addition to 0.005. INDUSTRIAL FERMENTATIONS. Four runs should be made at a level of approximately 1.0. 3.0, 10.0, and 20 hp. per 1000 gallons a t a gas velocity of 0.05 foot per second. At a level of either 3 or 10 hp. per 1000 gallons, the gas rate should be changed to 0.01 and 0.15 foot per second. Mechanism
The transfer of oxygen from the gas phase to the solid phase goes through two distinct steps (Figure 5). The first step is the transfer of oxygen from the gas to the liquid, which is designated by R, and is equal to the product of the absorption coefficient, KLa. and the concentration difference across the gasliquid interface. The greater the difference in concentration between the gas phase and liquid phase. the greater will be the absorption rate. R,
=
p,p.m. per hour
=
RLU( c x - c ) ( 2 )
The mixer influences the absorption rate only by changes that it can make in KLa. Therefore, it is very important to separate KLa from the absorption
INDUSTRIAL AND ENGINEERING CHEMISTRY
rate: R,, in evaluating the effect of mixing on a biological system. The value c* is obtained from knowledge of the oxygen concentration in the inlet and exit gas. If c,,~. is the saturation concentration of oxygen in sewage a t total oxygen pressure of 14.7 pounds per square inch absolute c * = [c,,t. for
pure oxygen]
x (3)
The transfer of oxygen from the liquid phase to the solid phase, R,, depends upon the dissolved oxygen level, the fluid turbulence in the system, and the area of transfer between the liquid and solid. The mixer can influence the rate of oxygen uptake of the solids (parts per million per hour) by its effect on the dissolved oxygen level, c. and by increasing the area for absorption. Nomenclature = =
= =
= =
= = =
dissolved oxygen level, p.p.m. oxygen saturation value corresponding to average partial pressure of oxygen in gas stream entering and leaving aerator diameter of impeller impeller speed rate of oxygen transfer between gas and liquid rate of oxygen between liquid and solid, also called rate of oxygen uptake tank diameter ratio of impeller diameter to tank diameter oxygen uptake rate divided by difference in average oxygen concentration between gas and liquid
literature Cited (1) Bartholomew, W. H., Karow, E. O . , Sfat, M. R.,IND. ENG. CHEM.42, 1801, 1810 (1950). ( 2 ) Cooper, C. M., Fernstrom, G. .A, Miller, S. A,, Ibid., 36, 503 ( 1 944). (3) Eckenfelder, W. W.? Sewage and Ind. M’bstes 24, 10, 11 (1952). (4) Karow, E. O., Bartholomew, W. H.: Sfat, 14. R.,J. Apr. Food Ciiem. 1, 4 (1953). ( 5 ) Oldshue, J. Y . ?“Aeration of Biological Systems Using Mixing Impellers,” Manhattan College Waste Treatment Symposium, April 1955. ( 6 ) Oldshue, 3 . Y., “Application of Mixers to Bioengineering Processes,” Rose Polytechnic Inst., Bioengineering Symposium, Terre Haute, Ind., 1953. ( 7 ) Oldshue, 3. Y., “Theory and Design of Mixers for Aeration of Waste!” 10th Ind. Waste Conference, Purdue University, M a y 1955. RECEWED for review January 11, 1956 A C C E P ~June D 11, 1956