THE UNIVERSITY OF MICHIGAN INDUSTRY PROGRAM OF THE COLLEGE OF ENGINEERING PERFORMANCE OF A SIMPLIFIED LOW-FLOW RATE COLD-WATER DEAERATOR KamilAksus Vincent Po Cramer Fo Go Hammitt April, 1962 IP-561

.' _ -,2Y 0~~~~~~~~~~~1 fA ^ *5 i

TABLE OF CONTENTS Page LIST OF FIGURES... o.......,..... o....... o............... iii LABSTRAC T, o o o... o o... o o.o... o... o o o......................... o o. o iv INTRODUCTION...,.................................. 1 DESCRIPTION OF APPARATUS................................... 1 TEST PROCEDUREo o o.............................oo 3 EXPERIMENTAL RESULTS o......... o.... o......... 6 DISCUSSION OF RESULTS....o o...... o...... o o.........o 12 1) Change in Flow Ratee................... o.. 12 2) Change of Absolute Pressure (Vacuum),...o......... 12 3) Change in Size (Diameter),.. o ao.o..oooo.o o.o.o 16 4) Change of Fluid Free Surface Area o o o,,o o.....o o 16 CONCLUSIONSo. O,........o o.......... o.. o o. o. o o.. o..O o oo 16 APPENDIX: VAN SLYKE APPARATUS. o....... o o o. o. 0 o 18 BIBLIOGRAPHY, o o o o o.....o.. O.o o o...... o o 0.......... o o. 25 AC(KNOWLEDGEMENT,..................................o.o..o. 26 ii

LIST OF FIGURES Figure Page 1 Cross-Sectional View of Main Body of Deaerating Unit,..,,.,.....,....... eo b...... o...a,..,. 2 2 Schematic of Complete Experimental Set-Upo,.. o....O 4 3 Flow Rate vs Differential Pressure (Sprayco H-5-D Nozzle)O. o o..o o o o. o o o o o.o 7 4 Percent Deaeration vs Differential Pressure 6" Diameter Deaerator, with Rashig Rings (Recalculated Data from previous investigationl)o....o o o o 13 5 Percent Deaeration vs Differential Pressure 6" Diameter Deaerator, without Rashig RingsO 6.OOa0o0 15 6 Percent Deaeration vs Differential Pressure 4" Diameter Deaerator. without Rashig Rings,, o..o... 16 7 Van Slyke Apparatus Schematic o.........o.... ooooo 19 iii

INTRODUCTION This report covers the preliminary investigations conducted to determine those factors covering the design of a simple, low-flow rate cold-water deaerating unit (1o0 - 1,5 gpm) to be used in a cavitation study research facility. It was found initially that a suitable unit could not be procured commercially. A previous study(l) showed that substantial deaeration could be achieved by simply spraying the water into a highly evacuated spaceo This report is primarily concerned with evaluating the effects of the following design parameters on the deaeration potential for such a unit1) change in flow rate 2) change of absolute pressure (vacuum) in evacuated space 3) change in size (diameter) 4) change in fluid free surface area DESCRIPTION OF APPARATUS Figure 1 shows a cross-sectional view of the main body of the deaerating unit. It consists of a 6" diameter by 36" long steel pipe, flanged at both ends, and fitted with a sight glass. The upper flange is drilled and tapped concentrically to receive a stainless steel "Sprayco H-5-D, Hollow Cone, Fine Spray, Medium Angle"* spray nozzle. An additional hole is provided for a vacuum-source connection; this hole is fitted internally with a baffle plate to prevent Manufactured by Spray Engineering Co,~ Burlington, Masso =1

-2VACUUM-SOURCE CONNECTION WATER INLET UPPER FLANGE NOZZLE SIGHT GLASS,, - / = 4,/ _XLEVEL MARKER 6"~~~ 1 Il/- MAIN PIPE I,,NSERT USED IN CERTAIN TESTS HOLES FOR WATER FLOW TO OUTER PIPE BOTTOM FLANGE WATER OUTLET Figure 1. Cross-Sectional View of Main Body of Deaerating Uni-t. Figr 1. Grs-Scioa Vie of Mai BOy ofeaRn PnIPE

-3spray carry-over into the vacuum system. The bottom flange is drilled and tapped concentrically to accomodate a drain valveo Figure 1 also shows how a 4" diameter by 35" long steel pipe was concentrically fitted into the unit when experimental runs were made to determine gross effects of pipe diameter change on deaeration potential, In the previous experimentation,(l) a screen supported by a tripod was fitted into the lower half of the unit and covered with three pounds of 1/4" Rashig rings. This was done as the first phase of determining the gross effect on percent deaeration* of increasing the exposed fluid surface areao In Figure 1, the tripod and Rashig rings have been removedo Figure 2 shows a schematic of the complete experimental set-upo In operation, water is sprayed through the nozzle into the evacuated space, and collects in the pipeo Samples of water, taken before and after the run, are tested and the percent deaeration determined, Vacuum is maintained and regulated through the use of a'Penberthy 182A Water Jet Exhauster "o TEST PROCEDURE The experimental test runs were made using the following procedure (Figure 2) 1) Water was admitted into the storage tank and allowed to drain for a short period of time, after which a water * Percent Deaeration = Initial Air Content - Final Air Content x 100 Initial Air Content **Penberthy Manufacturing Co(, Detroit, Michigan

WATER FROM MAIN -PRESSURE GAUGE VACUUM GAUGE VENT!^ --— 1-~ —.[,VACUUM LINE /WATER JET EXHAUSTER -DEAERATOR BODY HIGH PRESSURE AIR INLET7 VENT 12 GALLON CAPACITY WATER DEAERATOR DRAIN STORAGE TANK TANK DRAIN Figure 2. Schematic of Complete Experimental Set-Up.

-5sample was collected in a clean, thoroughly rinsed glass bottle, and the bottle tightly cappedo 2) The vacuum system was started and the tank evacuated to a predetermined absolute pressureo Water was then admitted to the nozzle inlet valve at a predetermined pressure from which the differential pressure across the nozzle corresponding to a certain flow rate can be calculatedo For the low flow rates, this pressure was obtained directly from the tap water supply line to the nozzle; for the higher flow rates, the 12 gallon tank was filled with water and then pressurized from a compressed air line in order to reach values higher than those available from the tap watero 3) The inlet valve to the deaerator body was then opened and the water was allowed to spray into the evacuated space until a certain level was reached as marked on the sight glass, after which the inlet valve was closedo 4) The vacuum system was secured and the deaerator body vent valve opened to equalize the pressure inside the unit with atmosphereo The drain valve was then opened and approximately half the water allowed to drain out before a sample was collected in a clean glass bottle, and the bottle tightly cappedo 5) The air content in each of the water samples was then found using a Van Slyke apparatus, and the percent of deaeration obtained by making the necessary calculationso A sketch of the Van Slyke apparatus, theory of operation, test procedure, and sample calculations will be found in the Appendixo

EXPERIMENTAL RESULTS The nozzle flow-rate as a function of pressure differential across the nozzle, as specified by the nozzle manufacturer, was checked and found to be substantially correct. Hence, the manufacturer's specifications were used for the test calculations, and are shown in Figure 3. Table I lists the data from the previous investigation(l) in which the pipe diameter was 6" and the three pounds of Rashig rings were used as previously explained. The percent deaeration values were recalculated for this report due to a fault found in the Van Slyke procedure;* however, since the percentage change in values after this recalculation was quite small, the interpretation of the gross effects derived from these data remains the same, except as particularly noted later in this reporto Table II lists the data and calculated percent deaeration for a series of runs in the 6" diameter pipe, similar to those given in Table I, except in this case the Rashig rings have been removed. Tables III and IV list the data and calculated percent deaeration for runs using the 4" diameter pipe "insert" (no Rashig rings)o For the test runs listed in Tables I, II, and III, the Van Slyke "tap water" readings were of the order of 150 mm Hg,, or lo7% air by volume (saturation at atmospheric pressure)o During the high flow-rate test runs listed in Table IV, it was found necessary to pressurize the water with compressed air in order to obtain inlet pressures higher than the 50 psig available in the supply lineo For these runs, the Van Slyke "tap water" * Refer to Step 7b) in Van Slyke, "Test Procedure", Appendixo -6

-7 i.8.....-. I-' I I l=''' — 1.84___T 1.6 / | 0 MANUFACTURER'S DATA |.4 2I 0 20 40 60 80 100 120 DIFFERENTIAL PRESSURE ACROSS THE NOZZLE, API, PSI Figure 35. Flow Rate vs Differential Pressure (Sprayco H-5-D Nozzle) (Sprayco H-5-D Nozzle)

-8TABLE I RECALCULATED DATA FROM PREVIOUS EXPERIMENTATION (6"Diameter Deaerator, With Rashig Rings) Deaerator Readings Van Slyke Readings Nozzle Pressure ZP Flow Manometer Average Percent (psia) psi Rate Readings Manometer Deaeration Gpm "Hg. Readings Inlet Outlet "Hgo 36.7 3.2 33.5 o.98 99 36.7 3.2 33.5 8 9898 99 80.75 36.7 3.2 33.5.98 100 22.7 3.2 19.5.76 91 22.7 3.2 19.5.76 92 92 82.60 22.7 3.2 19.5.76 92 36.7 7.2 29.0.92 149 36.7 7.2 29.0.92 155 152 66.40 36.7 7.2 29.0.92 152 27.2 7.2 20.0.77 152 27.2 7.2 20.0.77 160 156 67.40 27.2 7.2 20.0.77 156 43.2 9.7 33.5.98 180 43.2 9.7 33.5.98 179 180 60.20 43 2 9.7 3355.98 182 29.7 9.7 20.0.77 173 29.7 9.7 20.0.77 174 174 61.70 29.7 9.7 20.0.77 175

-9TABLE II EXPERIMENTAL DATA LOW INLET PRESSURES (6" Diameter Deaerator, Without Rashig Rings) Deaerator Readings Van Slyke Readings Nozzle Pressure AP Flow Manometer Average Percent (psia) psi Rate Readings Manometer Deaeration GPM "Hg Readings Inlet Outlet "Hg 49.7.50 49.2 1.19 47 49.7.50 49.2 1.19 48 47.0 77.30 49.7.50 49.2 1.19 46 24.7.50 24.2.84 50 24.7.50 24.2.84 51 50.0 7500 24.7.50 24.2.84 49 49.7 7.34 42.36 1.11 110 49.7 7.34 42.56 1.11 108 108.6 29.40 49~7 7.34 42,36 1.11 108 24,7 7.34 17o36.72 112 24.7 7.34 17.56.72 114 113.5 25.6O 24.7 7.34 17536.72 115 49.7 13.23 36.47 1.03 142 49.7 13.23 36,47 1.03 141 141.0 4.24 49.7 13.23 36.47 1.03 140 24.7 13.23 11.64.58 144 24.7 13.23 11.64.58 143 143.3 2.64 24.7 13.23 11.64.58 143

-10TABLE III EXPERIMENTAL DATA LOW INLET PRESSURES (4" Diameter Deaerator, Without Rashig Rings) Deaerator Readings Van Slyke Readings Nozzle Pressure Average (psia) AP Flow Manometer Manometer Percent psi Rate Readings Readings Deaeration Inlet Outlet GPM "Hg "Hg 49,7.50 49.2 1.19 52 49.7.50 49,2 1.19 53 52 73520 49.7.50 49o2 1.19 51 24,7.50 29.2.84 54 24.7 o50 29.2 o84 56 55 70.90 24.7 o50 29.2 84 55 49,7 7.34 42.36 1.11 112 49,7 7~34 42.56 111 114 112 27.10 49.7 7.34 42356 1.11 110 24 7 7 34 17.36.72 116 24~7 7734 17o36 o72 115 116 23550 24.7 7.34 17o36 o72 117 49.7 13o23 36.47 1.03 142 49.7 13523 35647 1.03 144 143 2.94 49.7 13523 36.47 1o03 143 24.7 13523 11o64 58 144 24.7 1.23 11o64 58 145 145 1.18 24~7 13.23 11.64.58 146

-11TABLE IV EXPERIMENTAL DATA HIGHER INLET PRESSURES (4" Diameter Deaerator, Without Rashig Rings) Deaerator Readings Van Slyke Readings Nozzle Pressure Average (psia) AP Flow Manometer Manometer Percent psi Rate Readings Readings Deaeration Inlet Outlet GPM "Hg "Hg 99~7.50 99.2 1.69 42 99.7.50 99.2 1.69 43 42 84.50 99.7.50 99.2 1.69 41 64.7 o50 64.2 1.38 45 64.7 o50 64.2 1.38 43 44 83520 64.7 o50 64.2 1.38 44 99o7 7.34 92.36 1.63 106 99.7 7.34 92036 1.63 105 105 38.20 99.7 73.4 92.36 1.63 104 64o7 7534 57.36 1.29 107 64.7 7.34 57o36 1.29 107 107 36,70 64.7 7534 57536 1.29 107 9907 13523 86.47 1.22 156 99 7 13523 86.47 1.22 157 157 12o55 99.7 1323 86.47 1.22 158 64.7 13o23 51.47 -1.58 156 64.7 13523 51.47 1.58 157 157 11o10 64.7 13523 51.47 1.58 157

-12readings were of the order of 175 mm Hgo due to additional air being dissolved because of the higher air pressures applied to the water (although time of contact was not sufficient to achieve saturation at these pressures)o DISCUSSION OF RESULTS The "Percent Deaeration vs Differential Pressure" data of Tables I and II are plotted on Figures 4 and 5, respectively, and those of Tables III and IV on Figure 6, for the following comparison purposes: 1) Change in Flow Rate: Figures 5 and 6 show that as the flow rate (ioe., nozzle differential pressure) increases, the percent deaeration increases. It will be noted that Figure 4 shows a slope in the opposite direction. However, it was found that these readings were erroneous due to a fault in the Van Slyke apparatus procedure, so that it is believed that the trend is in error. It would seem than an increase in inlet pressure would give better atomization, exposing more fluid surface (because of finer particle size) to the evacuated atmosphere, which in turn should lead to higher deaerationo It is therefore assumed that the slopes as shown on Figures 5 and 6, rather than Figure 4, are correct. 2) Change of Absolute Pressure (Vacuum): In all cases it is evident that as the absolute pressure in the evacuated space decreases (ioeo, vacuum increases) the percent deaeration increases, being highest when the vacuum reading is 29.5 in. Hgo Generally, the increase between 15oO in, Hgo and 2905 ino Hgo is very substantial, while at 3,0 in. Hgo very little deaeration is obtainedo

-13100 - -.- - CONSTANT VACUUM I 29.5 IN. Hg. 80 ll- - ------- z 0 Z 70 -J 4 70 L- -- -- ----------- -- -- - a __ 15.0 IN. Hg. j.._o_..~,.__3_QJ.0 IN. Hg. u 60 50 40 --—..10 15 20 25 30 35 40 DIFFERENTIAL PRESSURE, AP, PSI Figure 4, Percent Deaeration vs Differential Pressure 6" Diameter Deaerator, without Rashig Rings.

so CONSTANT VACUUM - 29,5 IN. Hg. 70 60 50 z 0 g. LU 0 0 20 30 40 0 60 DIFFERENTIAL PRESSURE A, P PSI Figure 5. Percent Deaeration vs Differential Pressure 6" Diameter Deaerator, without Rashig Rings. 6" Diameter Deaerator~ without Rashig Rings.

-1590 1 —1 ——........-,-.-.-.-. —.,,, -.-.-. I./ 8 0 _ __-,_,_ _ _ 00 60 60 50 0 FZ 40X_ cJ 30 201 0 I0,,..... z 0 0 20 40 60 80 100 DIFFERENTIAL PRESSURE, AP, PSI Figure 6. Percent Deaeration vs Differential Pressure 4" Diameter Deaerator, without Rashig Rings.

3) Change in Size (Diameter): Comparison of Figures 5 and 6 shows that for the highest vacuum and similar flow rates, the 6" pipe shows an increase in deaeration of about 5% over the 4" pipe, and there is little significant change at reduced vacuum. 4) Change of Fluid Free Surface Area: Comparison of Figures 4 and 5 shows that the use of Rashig rings, to increase the surface area and prolong the time that the fluid is in contact with the evacuated atmosphere, increases the percent deaeration (very substantially for the lower flow rates)o At the higher vacuum (29o5 in. Hgo) however, the increase in deaeration is only about 7% 8%. CONCLUSIONS From the results obtained it is concluded that: 1) The factor contributing most to the percent deaeration obtainable in such a simple unit is the degree of vacuum maintained in the evacuated space; the highest vacuum obtainable will give the greatest deaeration. 2) Increasing the flow rate also increases the deaeration factor over the range tested. For the equipment utilized in these tests a variation of the differential pressure across the nozzle from 30 to 80 psi results in an increase in percent deaeration of about 3% to 5%. 3) For the flow range specified (loO - 1,5 gpm), increasing the pipe diameter from 4" to 6" increases the percent deaeration by about 5%o

-174) The use of Rashig rings increases the percent deaeration considerably with low vacuums, but at the highest vacuum obtainable the increase in deaeration is only about 7% ~ 8%. In general, it is shown that at least 75% deaeration can be achieved in this simple unit with a single pass and starting with water saturated with air, With the addition of Rashig rings and larger pipe diameters, the effectiveness of such a unit could possibly be raised to above 85%0

APPENDIX VAN SLYKE APPARATUS The Van Slyke Apparatus is a device used to measure the total pressure of a known volume of air and water-vapor mixture, removed from a predetermined volume of watero The apparatus is shown schematically in Figure 7o It is made entirely of glass, except for 1) the rubber hose connecting the flask "F" with valve 2, and 2) the thick-walled rubber hose connecting the main glass tube with the water jacket containing bulb "A" and the thermometero Enough mercury is contained in the system to fill the glass tubing between valve 1 and valve 4, the flask "F" acting as a reservoir, A small, variable speed electric motor is used to agitate the water sample in the water-jacketed bulb "A"o -18

-19POS. "Y — 00 MANOMETER "M" (mm) 6 mi. BURETTE "B" FLASK >FLASK, — VENT "V" POS. "TS'-,,4 THERMOMETER (C) l~b A A I I/ GLASS WATER:S -50 X )3 } JACKET 50 POS. "BS"-.,-THICK WALL RUBBER HOSE RUBBER TUBING Figure 7. Van Slyke Apparatus Schematic.

-20A) Test Procedure: Start: (Make sure all valves are closed) 1) Place flask "F" at position TS: a) OPEN valve 2 (slowly). 2) LOWER flask "F" to Position BS: a) OPEN valve 4 (slowly) to vent "V", allowing mercury trapped above valve 4 to drop into bulb "A" -- CLOSE valve 4. b) OPEN valve 4 (slowly) to 6 mlo burette "B", allowing water in "B" (from last test) to drop into bulb "A" -- CLOSE valve 4. 3) RAISE flask "F" to Position TS: a) OPEN valve 4 (slowly) to vent "V" and allow mercury level in bulb "A" to slowly rise until approximately 1/8" of mercury is showing in vent "V" --- CLOSE valve 4. NOTE: This procedure will drive all the water from the previous sample out of bulb "A", and the tubing between bulb "A" and valve 4, b) OPEN valve 1 (slowly) and allow the mercury trapped in the vent line above valve 1 to drop into manometer "M" CLOSE valve 1o

-21NOTE: At this point, always check the glass tubing just below valve 3 to ascertain that no air is trapped in this section; if air is present, open valve 3 very slowly (so as not to blow the mercury in the vent above valve 3 into your eyes) and allow the trapped air to escape to the atmosphere. About 1/8" of mercury should always be kept in the vent line above valve 30 4) RAISE flask "F" to Position Y: a) OPEN valve 1 (slowly), allowing the mercury level in manometer "M" to slowly rise, until it just shows (1/8") in the vent line above valve 1 -- CLOSE valve lo NOTE: If the mercury stops rising before it enters the vent, very slowly raise flask "F" to a slightly higher position, allowing the mercury level in the manometer to rise in a corresponding manner --- close valve 1 immediately the mercury shows in the vent line. The system should now be completely filled with mercury from valve 1 to valve 4 --- check to make sure that no air (or water) is trapped under valves 1, 3, or 4. If air or water is found, secure all valves and repeat procedure from Step lo Testing Sample: 5) LOWER flask "F" to Position BS: a) Fill burette "B", with sample water, to 6 mil mark. NOTE: This operation is very critical --- extreme care must be exercised to avoid carrying air bubbles into the

-22water in the burette while filling. If any air bubbles are noticed, tap the burette gently to dislodge them and bring them to the surface. b) OPEN valve 4 (slowly) to burette "B" and allow the sample water level in the burette to drop to the 1 ml. mark on the burette --- CLOSE valve 4. c) Repeat steps a) and b) above until 20 mli of sample water has been measured off into bulb "A"o NOTE: After the 20 mlo of sample water has been measured off into bulb "A", check the tubing below valve 4 very carefully for the presence of entrapped air, If any air is noticed, 1) raise flask "F' to position TS, 2) open valve 4 (slowly)' to the burette "B" and allow the air to vent back into the burette (some water may come with it), 3) close valve 4 and lower flask "F" back down to position BS, 4) open valve 4 (slowly) to burette "B" and allow the water level in the burette to drop to the 1 mlo mark again, 5) close valve 4. 6) LOWER flask "F" slowly towards the floor, allowing the mercury level in bulb "B'" to slowly drop to a point about 1" above the mark "e" --- CLOSE valve 2. 7) RAISE flask "F" to position TS. a) Start shaker motor and agitate water sample in bulb "A" for 5 minutes, b) OPEN valve 2 (slowly), allowing water level in bulb "A" to slowly rise to mark "b" --- CLOSE valve 2o

-23NOTE: Make sure there is no water trapped in the section between marks "a" and "bTo If water is trapped: close valve 2, and procede again from step 6 onwardo c) READ, 1) mercury level in manometer "M", and 2) temperature of thermometero B) Sample Calculations: Nomenclature: Vs = 20 cco = predetermined volume of sample water Va = 2 cc, = known volume of air component in air + watervapor mixture at partial pressure "Pa" and temperature "t" Vac volume of air component "Va", corrected to standard atmospheric pressure "P" P =1 atmosphere = standard atmospheric pressure 760 mm, Hgo Pm total pressure of air plus water-vapor mixture Pv = water-vapor pressure, at temperature "t"o Pa = partial pressure of air component of mixture t = temperature of air plus water-vapor mixture 1) For Tap Water: Pm = 147 mmo Hgo t = 68~F (1) From Steam Tables: Pv = 0.690 in. Hgo = (0.69)(25o4) mmo Hgo (2) Pa = - Pv = 147 - (,69)(2504) = 12905 mm, Hgo

-24(3) Va = 2 cc. ~P 129 5 Vac = Va(p ) = 2(-6) = 0034 cc. (4) o Air (by volume) = Vac x 100 - (.34)(100) = 17% Vs 20 2) For Deaerated Water: Pm = 47 mm. Hg. t = 68~F (1) From Steam Tables: Pv = (.69)(25.4) mm. Hg. (2) Pa = Pm - Pv = 47 - 17.5 = 29.5 mm. Hg. (3) Va = 2 cc. Vac = Va(p) (2 2 = 0 0.776 cc. p 760 (4) % Air (by volume) = a x 100 = (.0776)(100) _ 388% Vs 20 (5) % Deaeration = 1.7 - 0.388 0 772 = 772% 1.7

BIBLIOGRAPHY 1o Ghose, So Development of a Low Temperature Deaerator, Term Paper in course MEo 600, Mechanical Engineering Dept., University of Michigan, Ann Arbor, Michigan, August, 1961. -25

ACKNOWLEDGEMENT The authors would like to acknowledge the assistance of Mr. John Ko Brigham, Jr, of the Mechanical Engineering Departmentin the construction and assembly of the equipment, and of the Industry Program of the College of Engineering of the University of Michigan in the reproduction of this report. -26