UTGE UNIVERSITY OF WISCONSIN COLLEGE OF ENGINEERING ENGINEERING EXPERIMENT STATION Contributions from the Department of Steam and Gas Engineering TESTS ON THE RECIRCULATION OF WASHED AIR BY GUSTUS L. LARSON Associate Professor of Steam and Gas Engineering Reprinted from the Transactions of the American Society of Heating and Ventilating Engineers Vol. 22, pp. 11-51 MADISON 1916 Digitized by the Internet Archive in 2017 with funding from University of Illinois Urbana-Champaign Alternates https://archive.org/details/testsonrecirculaOOIars « W 1 / ALTGELD KALL TESTS ON THE RECIRCULATION OF WASHED AIR By G. L. Larson* INTRODUCTION In the light of more modern studies in ventilation it would seem that the real explanation of the ill effects of bad ventilation as not to be found in the chemical composition of the atmosphere breathed. The long debated idea that expired air contains or- ganic matter which is toxic has been abandoned by most physiolo- gists. The various phases of the chemical composition of air were discussed some time ago in a symposium on ventilation at the Chemist’s Club in New York City. There is unanimity among them regarding the chemical viti- ation of the air. Pure air contains nearly 21% of oxygen. This may be reduced to 17%, a proportion too small even to support ordinary combustion, before its diminution becomes harmful. Except in extreme conditions the amount of oxygen in the closest halls crowded with people practically never falls below 20%. Oxygen will therefore take care of itself and may probably be wholly left out of consideration in ventilating systems. We are reminded that it is necessary to go only a short distance up into the mountains to come under an atmospheric pressure such as to reduce the oxygen supply much more than it is reduced in crowded assemblies, and yet mountain air is especially healthful. The air, under usual conditions, contains about 4 parts of car- bon dioxide per ten thousand parts (0.04 per cent.) and the “standard” of desired purity for the air of dwellings was long placed as low as 6 parts per ten thousand. Experimentation in- dicates, however, that it does not become harmful to man until the carbon dioxide accumulates to above one per cent., or nearly forty times its usual amount. The air in crowded rooms very rarely reaches 0.4 per cent., so that evidently a quantity of car- Associate Professor of Cas and Steam Engineering, University of Wisconsin. 4 Tests on the Recirculation of Washed Air bon dioxide far exceeding the highest hygienic limit which has hitherto been set up as a “standard” can be breathed with impu- nity. It has also been stated that the bacteria in the air need not be considered in the problem of ventilation, since the comparative unimportance of the air as a vehicle of infection is becoming widely recognized. (See Journal of the American Medical Asso- ciation, “The Air as a Vehicle of Infection.” Feb. 7, 1914, p. 423.) In contrast with all the foregoing negative factors with respect to the discomfort or ill health hitherto associated with inadequate ventilation, we may now conclude with reasonable certainty that the symptoms of discomfort in a badly ventilated place are due to the physical condition of the air with respect to temperature, humidity and movement, and not to any chemical properties whatever. (The Journal of the American Medical Association, Nov. 7th, 1914.) The university engineers had the above points in mind in designing the ventilating systems of the Wisconsin High School, and the tests herein described were undertaken in order to ascer- tain the advisability of installing similar systems in the future buildings of the University. The system was designed by J. M. Smith, Operating Engineer of the University Heating Station, under the supervision of Prof. H. J. Thorkelson, Consulting Engineer for the University. The tests were undertaken especially with a view of throwing light upon the following questions : 1. What constitutes good ventilation? 2. Is proper ventilation a problem of supplying large volumes or merely a question of higher velocities with properly regulated humidity and temperatures? 3. Does recirculation give efficient as well as economical ven- tilation ? 4. What is the effect of the washer upon the bacteria and carbon dioxide content of the air? The writer desires to express his thanks to Prof. H. J. Thorkel- son for his valuable advice and encouragement and to Prof. Wm. Black for his advice and assistance at various times. He is also indebted to Mr. J. M. Smith, Operating Engineer, and to Messrs. Bauer and Blanding, of the Senior Class in Mechanical Engineer- ing, for valuable assistance rendered in preparing for and conduct- ing the tests. The writer desires especially to express his thanks to Mr. E. J. Tully, Chemist, State Laboratory of Hygiene, for his hearty co-operation and valuable assistance in conducting the bacteria tests. Tests on the Recirculation of Washed Air 5 6 Tests on the Recirculation of Washed Air DESCRIPTION OF BUILDING AND EQUIPMENT In preparation for the tests herein recorded some preliminary tests were made on a small air washer and recirculating system which was built for experimental purposes and installed in one of the rooms in the Engineering Building. With this apparatus it was almost impossible to duplicate conditions as found in prac- tice and as the tests were of a preliminary nature only, their results will not be recorded in this treatise. The tests which will be described here were made at the Uni- versity of Wisconsin High School. This is the newest of the buildings on the Wisconsin Campus, being used for the first time at the beginning of the present school year. Figure I shows a view of this building. It is substantially built of Bedford stone and pressed brick, and equipped with steel window frames and sash. Each window has at least one venting panel and the win- dows in the first, second and third stories have two such panels. These panels are designed with double contacts to insure clos- ing exactly to prevent the passage of air. Only the south and middle portions of the building have been constructed, the north wall being left in a condition to facilitate future extension. SYSTEM OF HEATING The building is heated with 8,530 square feet of direct radia- tion with the addition of 516 square feet of indirect radiation placed between the washer and the fan. The system is of the one pipe, direct steam type throughout. The direct radiators are of the Peerless pattern and they are supported on the walls by iron brackets. The indirect radiators are of the Vento cast iron type. This indirect radiation consists of four radiators set two radiators high and two wide and valved with hand valves in such a manner that one, two, three or all the radiators may be used as needed. The radiators are placed at such a height that ample space is left beneath them for by-passing the air to the fan. All of the radiators in the building, including the indirect coils and by-pass damper, are controlled by the national automatic temperature control system. All air controlled valves on radia- tors are equipped with hand screw and lock shield stems for permitting the valves to be closed by hand, so that in mild weather any of these radiators may be cut out. Steam is delivered to the building through a tunnel from the University Heating Plant. Tests on the Recirculation of Washed Air 7 SYSTEM OF VENTILATION The ventilation of the building consists of a blast fan dis- charging through ducts on the ceiling of the basement and rising to the rooms to be ventilated, entering the rooms near the ceil* ing. The vent ducts leave the rooms near the floor and are carried down to a system of tunnels below the basement floor which carry the air back to the fan through an air washer and indirect radiators. A sliding adjustable door is provided in the housing ahead of the air washer for the admission of outside air when necessary. The toilet rooms are ventilated by a system of exhaust venti- lation consisting of an exhaust fan in the attic and a system of ducts leading from the toilet rooms to the fan with a connection to each closet fixture. The ventilation of the Chemical Laboratory consists of an ex- haust fan in the attic with a duct leading to the Chemical Labora- tory on the third floor. This duct has a register at both the floor and the ceiling. The exhaust fans discharge through globe vents on the roof. The general ventilation of the building is furnished by a No. 13 multi vane blast fan rated at 165 R. P. M. direct connected to a 10 H. P. 500 volt direct current motor. The fan is rated to de- liver 32,400 cubic feet per minute against a pressure of 24 an inch of water. The speed of the motor can be varied by field con- trol. The air washer is a Thomas “Acme” type. This washer con- sists of a spray chamber equipped with sufficient spray nozzles of approved type ; a settling tank supplied with a float valve, connected to the LTniversity water main, to maintain the water level and also an overflow and drain pipe connected to the sewer. It is equipped with a centrifugal pump which takes its suction from the settling tank and discharges through a basket strainer, to the spray nozzles. The eliminators are of the vertical type. The pump is direct connected to a 500 volt direct connected motor. The entire system is designed with the view of future enlarge- ment of the building and the capacity of the apparatus is suffi- cient to supply the entire building when completed. The ducts and tunnels are arranged so that the future wing of the building can be connected directly to the present system. 8 Tests on the Recirculation of Washed Air DESCRIPTION OF THE APPARATUS AND METHODS USED Steam and Power Consumption Tests The condensed steam from the building was taken directly from the steam traps and weighed. The condensate from the traps was at a temperature of about 210 degrees and it became necessary to run it through a condenser barrel to prevent loss from evaporation. Figure II is a view of the arrangement of bar- Fig. III. Tests on the Recirculation of Washed Air 9 rels and scales used. Figure III is the same view with the barrels and scales removed to show the condenser barrel and the steam traps. This view also shows the main steam line entering the building from the tunnel and the return piping. Calibrated me- ters of a standard make were used to measure the power con- sumption of the fan and washer motors. No particular difficulty was experienced in weighing the con- densate except that it was impossible to get a steady and uniform flow. This was undoubtedly due to the intermittent action of the thermostats and possibly, in some degree, to sticking of the steam traps. AIR MEASURING APPARATUS An anemometer was used to measure the air velocities. Since these instruments are often very unreliable great care was taken to calibrate it properly. A special apparatus was built for per- forming this calibration. It consists of a stand with a movable arm of such a radius that the anemometer moves around a circle twenty feet in circumference with one revolution of the arm. By means of a belt and pulleys any speed desired can be obtained. A lever arm on top of the stand is for starting and stopping the recording mechanism of the anemometer when the movable arm which carries the anemometer is in motion. Considerable difficulty was met in getting the true volume of the air entering the rooms. This will be explained later. CARBON DIOXIDE APPARATUS Haldane’s Portable apparatus was used to measure the carbon dioxide contained in the air. This apparatus is easily operated, and, while it is not the most accurate one on the market for measuring small amounts of carbon dioxide, it is accurate enough for most conditions met with in practice. Haldane, in his book, “Methods of Air Analysis,” states that his portable carbon dioxide apparatus will not vary more than one-half of one part in ten thousand either side of the correct result. The writer checked the apparatus at various times by measuring the carbon dioxide contained in the outside air, and at no time did the readings vary more than the above mentioned amount. It is safe to say that even with a little practice, read- ings can be obtained which will not vary more than one part in ten thousand either side of the correct residt. 10 Tests on the Recirculation of Washed Air APPARATUS FOR BACTERIA TESTS Two different methods were used to obtain a count of the number of bacteria in the air. The first consisted of drawing a measured amount of air through a sugar or sand filter. No con- sistent results were obtained from the use of sugar filters. The moisture in the air caused the sugar to stick to the inside of the filter tubes and it was removed with considerable difficulty. The time required to take samples varied very greatly with the sugar filters. The sand filters gave very consistent results as will be seen later. The second method consisted in the use of Petrie dishes and most of the bacteria tests were made in this way. APPARATUS FOR TRACING AIR CURRENTS Several methods were used for tracing the air currents in the rooms. The common method of using ammonia and tumeric paper was tried but the change in color of the tumeric paper from yellow to pink was so gradual that it was next to impossible to tell when the action commenced. Both of the other methods used were quite successful. Instead of using ammonia and tumeric paper, hydrogen sul- phide and lead acetate were used. A rubber tube from the hydro- gen sulphide generator was placed in the incoming duct and filter papers dipped in lead acetate were placed in various parts of the room. The filter papers turned black almost immediately upon coming in contact with the hydrogen sulphide gas. The odor of the hydrogen sulphide gas makes it inconvenient to use it, but on the whole it is fully as satisfactory .as ammonia. The air currents were also traced by using very light stream- ers of silk floss. This method proved very successful as very slight air currents can be traced in this way. OBSERVATIONS Air Measurements As has been stated before, considerable difficulty was met with in getting the true volume of air entering the rooms. Readings were taken at the ducts leading to the gymnasium. First a series of readings were taken at the register and then the register was removed and another set of readings taken holding the anemometer horizontally in the vertical duct leading to the Tests on the Recirculation of Washed Air 11 loom. The readings at the register showed an average of 482 feet per minute and those in the duct showed an average of 810 feet per minute. The registers are unusually heavy and the net area of the par- ticular size in the gymnasium is only 96% of the area of the vertical duct leading to the room. Therefore the velocity through the register should check very closely with the velocity in the duct, which is very far from being the case as shown above. Further readings were taken in a room which had a horizontal duct leading to it so that the anemometer could be left standing in the duct and readings taken after the register had been re- placed. In this room the velocity at the register was 386 ft. per minute ; the velocity in the duct, with the register removed, was 944 ft. per minute ; and the velocity in the duct after the register was replaced was 828 ft. per minute. In this room' the net area of the register is 88% of the area of the duct leading to the room. Note that the ratio of velocities in the duct, with and without the register in place, is 828/944 or 87.7%. As a check upon the above, readings of a similar nature were taken in another room. This room also had a horizontal duct leading to it so that the anemom- eter could be placed in the duct behind the register. Here the ratio of the velocities in the duct with and without the register in place was 86.8% and the ratio of the velocity at the register to that in the duct was only 35.7%. The net area of the register was 88% of the area of the duct. These tests show conclusively that the register deflected the air currents in such a manner as to give velocity values which were very much lower than the actual values, and that readings 12 Tests on the Recirculation of Washed Air taken with the anemometer placed against the register are abso- lutely unreliable. The above is a sketch of a portion of one of the registers. They are made of pressed steel and the section at AB shows the form of the stampings. The concave surface of the meshes will of course set up swirls in the air current but it would only be a guess to say what gen- eral direction they would take. As in hydraulic work, there will probably be a contraction of the air current after it has passed between the meshes. But in accounting for the low velocities obtained when the ane- mometer was held against the register face, it must be borne in mind that an anemometer is calibrated under conditions where the air strikes with equal intensity over the entire surface of the vanes. In a register such as the above, 56% of the face is composed of the meshes. These meshes create a great number of dead air spaces, and, Tests on the Recirculation of Washed Air 13 while the velocity through the meshes may be the same as the velocity in the riser, the anemometer will not show it because the vane area effected is much less than the vane area effected under calibrating conditions. Velocity measurements were taken in all rooms. The results are tabulated in Table I. The registers were removed in each case and enough readings taken in the duct to giv,e a fair average. The volumes were obtained by multiplying this average velocity by the net area of the register. That this method gives a fair value was shown later by velocity measurements taken at the suction side of the fan and in the two return tunnels from the rooms. Referring to Table I it will be seen that the total volume supplied to all the rooms was 15,241 cubic feet per minute. The readings taken at the suction side of the fan gave a volume of Table I Ventilation Test (V JJ CD 5 D z I 8 0 1 Size of meat I DUCT INCHES I5l t? a 0 Die o ju a iuZK > IU n fo < 0 y. a 0 OUJ H-l- !§ Sfta 3 OlU 0 «0 0- Volume of boom IN CU FT 5TUDCNT CAPACITY OF BOOM 8 i <0 h hi fig f t Oi| 0 O0!(U h h lU a 3 •Jgz 39 11 Oioa time in minutes eeq'd TO CHANGE A1B IN BOOM > h 5 I r ttJ > 5 JU Ql 3 4*4 4*8 0.1 II 0.098 6SZ 65 2012 31 €» 4*4 4*8 Oil) 0.098 652 65 2012 31 <3 8*12 16*12 0.666 0.587 86-9 510 16450 20 822 25.5 32 48 1 0 6*12 12*12 OS 044 SSI 375 13600 28 486 13.4 36 31 TWO TWO 2.66 2.51 765 15 12*32 24*32 266 2.51 712 3710 8226 0 22 55 106 9*12 12*18 0.75 0.666 740 493 9830 26 380 19.0 20 51 107 8*12 12*16 0666 0.587 759 445 8780 26 338 17.1 20 40 IOG 12*16 16*24 1.33 1.25 592 740 16600 127 130 58 22 48 109 8*12 12*16 0666 0587 514 478 9450 26 364 183 20 43 115 8*12 12*16 0.666 0.587 676, 397 8300 31 268 12.8 21 42 207 3*12 12*18 075 0.666 535 356 98oo 31 316 11.5 27 42 2 oa 8*12 12*16 0.666 0587 553 325 8780 26 838 125 27 38 2 03 12x16 16*24 133 1.25 918 1147 16609 1 27 130 9 14 47 2io 8*12 12*16 0666 0.587 614 360 9300 26 3 38 138 26 49 211 8*12 12*16 0.666 O.S87 455 267 8450 31 273 86 32 47 212 6*7 7* 12 0291 0. 257 477 123 3650 6 620 2 l.o 30 44 215 8*12 12*16 0666 0.587 480 282 8450 20 423 14.1 30 4/ TWO 2.91 274 367 • 214 12*35 Z4*3S 291 274 896 3455 79760 361 120 9.6 23 55 308 10*12 12*20 0833 0.733 510 374 10900 41 266 9.1 29 38 30°) 10*12 12*20 0.833 0.733 68 o 498 10900 25 437 20.0 22 40 3i o 6*6 6*12 025 0.2 2 245 54 3580 66 ill 8*12 12*16 0666 0587 429 252 8650 IS 279 16.8 34 42 312 8*12 12*16 0.666 0587 139 82 8450 14 563 5.8 103 30 315 6*7 7*12 0291 0.257 609 156 3650 23 314 8*12 12*16 0.666 0.587 395 232 8450 29 423 So 36 46 Totals Avebage*© 2584 A 33.1* °607 tsZ4l 3fc8fc84 °367 °I36 ®26 °44 Booms 6>.3io. 313 aoe 6 tosle cooms ^Obtained bv negiectins store rooms and boom3i2 14 Tests on the Recirculation of Washed Air 15,000 cubic feet per minute and. the readings in the return tunnels gave a volume of 14,120 cubic feet per minute. fable I shows that the velocity of the air entering the rooms ranged from 139 to 918 feet per minute with an average of 607 feet per minute. This average is about double the velocity usually recommended and in the opinion of the writer it is one of the desirable features of the system. The volume of air supplied per student is low compared with the amount that would have to be supplied on the old basis of keeping the carbon dioxide content down to six parts in ten thousand. The cubic feet of air per student varies from 5.8 to 25.5 with an average of 13.6. This seems low to one accustomed to the old figure of 30 cubic feet of air per student per minute but the air in the rooms always seemed fresh and clean and the amount supplied was apparently ample. (See questionaire submitted to the teachers in the build- ing.) STEAM CONSUMPTION TESTS Three steam consumption tests were made, of 24, 24 and 27 hours’ duration, respectively. The condensed steam was weighed every fifteen minutes and all temperature readings were taken every half hour. Power readings and humidity readings were also taken every half hour. The summarized steam weights and temperatures are given in Tables II, III and IV. Figures IV, V and VI are graphical logs plotted from Tables II, III and IV. The Vento heaters in the housing were not in use during any of the tests. There was no steam on them and the condensation weighed from them was only leakage through the valves. Test No. 1 was started at 7:00' A. M., February 2nd, and con- tinued to 7 :00 A. M., February 3rd. February 2nd was dark and cloudy and there was not much temperature variation during the twenty-four hours. The highest average outside temperature was 29.6° and the lowest was 22.1° with an average of 26.6°. The average inside temperature was 67.1° and the average pounds of steam per hour per degree difference in temperature between the outside and inside was 24.1. The graphical log shows clearly the relation between the hourly steam consumption and the outside temperature. Notice the effect of the fan upon the steam rate curve. There was a sudden decrease in steam con- sumption soon after the fan started and an increase immediately after the fan was stopped. This is also shown in the curve of steam consumption per hour per degree difference. Tests on the Recirculation of Washed Air 15 jjia asasa a aad booh aad WV3J.SJO SONflOd oo 5 v9 rrt (M <0 <0 to t-j N S v * ft s © jO 0 vt. 10 0 N N g © © Q N 0. (vl a N 0 cvj ft N 5 N N N N o N 0 “J 1M *0 <0 N 0 ft N ■^ONga^ddia aanxvaj3dW3_L N 41 h 0 ■vt U) 10 ■5 d © ft © S' <0 <0 •+ 10 (0 ft lO to © <0 © 10 to 3 to © ©. © © 3 to p vj 10 10 n <0 to 10 vS vO p p <0 to 3 anxvaad 3C3ISKIJ 3S>V»3AV 3 § v9 O' P. vt v9 N LO vO 1^ *0 v9 <0 >0 3 © © vfl < U) vD \9 to © vfi P 10 vfi <0 ■0 vfl N 0 vO 3 v9 P <0 v9 <0 vS 3 3 3 \9 G5 v0 o> vO 00 vf vO 3 vO N. >0 vO vS Vt vO >0 vj- v5 p. p •0 V0 ciwal 3a is -J.0O 39V33AV © N S N ft rO N © s9 CO m 5 N j* 8 ft © N v9 N P 8 © N 10 O' <9 ft © © vJ. K) © 0 © s p N P ft N P. N p h aj CSJ p 43 0 (5 sjoixdwnSNoO HV3ig IVXOJL © £ <0 10 £ ?! JO 3 © © vO * <0 § *0 vQ *0 vO 8 ~w © ft s CO 1 8 p o © j* © © © 0 p rf} N» c -i v9 © v9 s © vfi n LQ 5 5 ft 0 8 ft 0 © © oi > anon aaia/v\ nvxo_L © vfl Nj P s 00 s p 1 10 10 (0 C 0 N CO « <0 s CO © © 10 p v9 « N 0 ,0 © © N R R N © © M ■ s £ © N © © 8 <0 p 0 0 © an oh aad shod wnN3id WV3J.Q dO XHDI3/V\ U) 0 © 0 P 0 0 - - 0 o z N 0 0 - - - 0 0 0 - - N p 0 N «now aiad saoiviavg wohj iNvaxg' do j.w90/v\ > i I 10 r~ 5 0 H l/) © is » 0 r^» © © 0 5 N m P © R to <0 (0 (0 0 N r ft N CO to f? <0 ®0 «0 0 CJ 0 p 10 o $ N 3wix 12 12 co 8)' >0 o ■0 !2 12 2 Ni U) p iO % >0 \S © in CO P «5 © O 2 12 p © N >0 5 © 3 © 5) © 10 s <0 >0 65 5 0 p 1 16 Tests on the Recirculation of Washed Air table iii-a. Faph amdAir Washer Log FEOM 7JOAM TO 3:30 PM. PER 20. / SIS. TEST NO. 2. T IME Motor Logs Fam a/nd Washer Log Fam Motor Pump Motor Temperatures Humidity I a L P J O > <0 tu O' Cl X $ oo> (5 Z 01 f t)(0 fc-f 3 CO h 2 ca s CD 1- DBYBCILB >- Oi p 1L UJ . O 7.30 170 312 7S> 3.85 53 65 54- 58 60 55 44 77 a.oo 170 312 7.3 4.05 54 65 56 58 60 55 48 89 8.30 173 513 7.8 4.20 56 9.00 173 313 8 2 435 56 65 58 59 60 57 57 94 9:30 172 507 7.7 4.50 io:oo 174- 325 8.0 4.65 575 65.5 59 60 60 59 62 94 lo:3o 173 523 SO 4.65 11 :oo 175 323 3.1 475 58 65.5 60 60 60 58 64 IOO 1 1 :so 174 510 7.9 4.7o I2:oo 173 SOS 7.9 4.65 58 66 397 60 60 58 62 98 12:30 173 510 8 1 4.75 l:oo 174 520 8.0 4.80 58, 66 39.7 60 60 58 62 98 i:3o 175 525 80 4.73 2oo 174 325 82 5.00 58 66 59.5 60 60 59 62 97 230 176 525 8.5 5.20 3:00 173 5ZO 8.2 5.00 58 66 595 60 60 59 62 97 3:30 176 5Z5 82 5.00 58 66 60 60 60 59 62 100 Ave. 174 518 8.0 4.64 56 .8 65.6 585 5 as 60 573 585 944 Table II-A Fam amd Air Washer Log rBOM 730 A M. TO 330 PM. FEB. 2 IOI.S TEST No. 1 UJ 5 h Motor Logs Fam a/nd Washer Log Fam Motor Pump Motor temperatures HUMIDITY 5 tt £¥ P j 10 81 & 1 x < w UJ Of s* 5 < Befobe Washing After Washing got r iu W F II Befofe welshing; 0 gi tf <* Wet Bulb Dry Bulb Wet Bulb. Dry Bulb > Qt <1 Q ss 20 X . 5$ mo *0 0 3 Z 0 r . £8 Is r h 0 o tf)Q 7-30 173 3o7 78 345 8:00 177 532 80 395 56 655 575 60 605 56 55 84 8 3o 172 508 79 390 565 65. 58 60 605 57 59 . 89 300 173 530 80 425 605 58 9-30 173 525 8.15 430 59 56.5 67 68 58 60 60 S 58 62 61 89 lO: OO 172 SIS 77 428 59 575 67 66 58.5 60 60 57 62 60 92 10:30 173 Soo 75 423 59 58. 675 665 59 60.3 60 58 60 60 92 1 1:00 174 512 ao 450 59 58 675 665 59 60.5 61.5 59 60 60 92 1 1-30 173 515 76 445 59 58. 67 665 59 605 61.5 58 62 60 92 1200 173 520 78 460 59 58.5 67 665 59.5 61.0 61.5 59 62 62 92 l2’3o 173 500 8.0 450 59 58.5 67 66,5 59 605 6/5 59 62 60 92 loo 17/ 515 75 450 585 58. 665 665 59 60.5 61.5 59 62 62 92 1:30 172 520 80 4 SO 595 58 675 665 585 60 6 IS 58 62 62 92 zoo 174 53o 30 4.55 59S 585 67 66 58.5 595 61.5 585 64. 65 94 ?.’3o 172 5Zo 75 445 58 59 £>£> 67 59 60 61.5 58 62 62 94 3:00 174 5 10 80 455 585 59 66.5 59 60 61.5. 57 64 64 94 330 174 512 80 455 58. 59 66 67 60 61 61.5 59 64 64 94 Ave. 173 516 785 433 58.5 58.2 66.6 666 587 60.5 60 9 58.0 615 617 9/6 *0 v 3< *- r of < ^ w h MO >JF 0 » oi 9 fg< 0|-fl < 5 ® s? pn- H<0 V ff) Z A tog* 2< z r Eo? . iLfOj; 1 a 0 1-1-3 2e 75 Tests on the Recirculation of Washed Air 17 8.00 9.00 10.00 I 100 12.00 LOO 200 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 I LOO 1200 LOO 200 3.00 4.00 5.00 6.00 7.00 Fig. 4 18 Tests on the Recirculation of Washed Air 8.00 9.00 10.00/ 1.00 12-00 LOO 2 00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 10.00 I LOO 12-00 LOO 2-00 3.00 4.00 500 600 700 Fig-5 Tests on the Recirculation of Washed Air 19 20 Tests on the Recirculation of Washed Air In all three tests there are irregularities in the steam consump- tion curve which can only be accounted for as resulting from the intermittent action of the thermostats, or from lack of positive action in the steam traps. At times the condensed steam would almost cease to flow and then would suddenly flow out in large quantities. If these high points were spread out over several readings preceding them the curve would give a better idea of the actual relations existing. For a more technical discussion of the exact results of these tests see Chapter on “Calculations.” Test No. 1 showed such a decided drop in steam consumption after the fan had started that it was thought advisable to check the results with a duplicate run. In Test No. 1 the fan was start- ed thirty minutes after the test was started and it was impossible to tell just what effect it had on the general shape of the curve. To overcome this difficulty the second test was started 12 y 2 hours before the fan was started. Test No. 2 was started 7 :00 P. M. February 27th and continued to 7:00 P. M. February 28th. February 27th was a bright and sunny day as was also February 28th. The average outside tempera- ture ranged from 17.9° to 39.4° with a total average of 26°. The extreme range of the outside thermometers varied from a mini- mum of 16.5° to a maximum of 50°. The average inside tem- perature (inside wall) was 66.8 degrees and the average pounds of steam per hour per degree difference was 17.2. The graphical log shows the effect of the fan on the steam consumption. There was a small drop in steam consumption due to the fan but not nearly as much as was shown in Test No. 1. In Test No. 2 the steam consumption continued to decrease after the fan had stop- ped. This is probably due to the fact that the average outside temperature curve was still rising and that the walls must have absorbed considerable heat from the high temperature on the sunny side in the afternoon. In the test just described thermometers were hung in the middle of the rooms as well as near the inside wall of the rooms. It is interesting to note the effect of the fan upon the heat distribu- tion as shown by these thermometers. It can be seen from the graphical log that the temperatures in the middle of the rooms averaged more than two degrees lower than the temperatures near the inside walls during the time the fan was not in opera- tion. These temperatures became nearly equal shortly after the fan had started. Tests on the Recirculation of Washed Air 21 In the two tests just described the fan was operated only during the time that the building was full of students. To eliminate the possibility that the decrease in steam consumption resulted from the animal heat of the students, a third test was run at a time when the building was unoccupied. This test, No. 3, was started at 8 :15 P. M., March 3rd, and con- tinued to 11 :15 P. M., March 4th. The outside, weather condi- tions were almost an exact duplicate of the conditions when Test No. 1 was made. The average outside temperature ranged from 23.6° to 31.8° with a mean of 27°. The average inside tem- perature for the run was 65.3° and the average steam consump- tion per degree difference per hour was 22.1. The fan was started at midnight and stopped the following .evening at 8:15 P. M. No definite decrease in steam consumption is shown by the curves when the fan was started but they do show a decided increase immediately after the fan was stopped. All three tests show a decided decrease in steam consumption during the time the fan was in operation. Test No. 3 shows more than Test No. 2 but not as much as Test No. 1. For the exact percentages see the chapter on “Calculations.” . HUMIDITY TESTS All humidity measurements in the various rooms were made with a sling psychrometer of the pattern recommended by the U. S. Weather Bureau. The humidity tests made on the air entering and leaving the washer were made with a Hygrodeik which had been checked with the sling psychrometer. The relative humidity in the various rooms as shown in Table I was taken on one of the coldest days in the winter. It varied greatly in the various rooms but the average for the entire build- ing was 44%. The humidity of the air as it enters the washer can be taken as the average humidity of all the rooms. That this assumption is correct is shown by the following measure- ments taken in some of the rooms on the day steam consumption Test No. 1 was made. Room Number Wet Bulb Dry Bulb Humidity Gymnasium 58.3 63.5 74 Room No. 10 54.5 65 50 Room No. 108 59 69 55 Room No. 209 59 65 70 Assembly 61.7 69 66 Room No. 309 55.5 66 51 Room No. 311 57.5 64.5 66 Room No. 211 60 69 59 Room No. 107 58.5 67 60 Average 28.2 66.4 61.2 22 Tests on the Recirculation of Washed Air The average humidity of all the above rooms was 61.2% as compared with 61.6% in the air entering the washer. In Test No. 2 the average humidity of the air entering the washer was 58.5%, and in Test No. 3 it was 61.5%. In each case the air had a relative humidity of about 95% as it left the washer. There can be no doubt but that a reasonably high humidity has a very beneficial effect upon the quality of the air. It not only improves the quality of the air for breathing but it makes it pos- sible to keep the rooms at a muqh lower temperature. An inspec- tion of any of the temperature log sheets will show that the average temperature was rarely above 67°, and often it was below 65° in some of the rooms, without a single complaint being regis- tered. The temperature of the air leaving the washer was usually about 60°. A series of temperature measurements taken in the registers of the various rooms showed that the air entering the rooms was also 60°. At first glance one would say that this was too cold. As a matter of fact it proved to be a benefit rather than a detriment, as it seemed to give more life to the air in the rooms and at the same time did not cause any uncomfortable draughts as would be expected. This test was made on March 30th, when the weather was quite mild. ’No test has been made during the summer so that the writer is unable to state what entering temperature might be expected if recirculation is resorted to during the summer months. However, in view of the fact that the air entered at 60° on a comparatively mild day, it is not likely that the temperature would reach an unreasonable value during the warmer months. CARBON DIOXIDE TESTS The following carbon dioxide tests were made on the two sides of the washer and in the various rooms mentioned. Tests on air entering washer December 14th. 3:20 P. M. : Sample No. 1 showed 11 parts carbon dioxide in 10,000. Sample No. 2 showed 10 parts carbon dioxide in 10,000. Tests on air leaving washer December 14th. 3:35 P. M. Sample No. 1 showed 6 parts carbon dioxide in 10,000. Sample No. 2 showed 6 parts carbon dioxide in 10,000. The tests tabulated below were made December 16th. TESTS ON AIR ENTERING THE WASHER Time 10:45 10:55 11:01 2:45 2:56 Average Parts in 10,000 8 6 8 10 10 8.4 Tests on the Recirculation of Washed Air 23 TESTS ON AIR LEAVING TIIE WASHER Time 10:15 11:20 11:30 11:42 11:53 2:15 2:30 Average Parts in 10,000 10 9 7 8 0 8 6 8.14 On January 12th the following results were obtained: Sample No. 1 taken 9 :55 A. M. on the air after being washed showed 6 parts in 10,000. Sample No. 2 taken 10:05 A. M. on the air after being washed showed 7 parts in 10,000. Sample No. 3 taken 10 :25 A. M. on the air before being washed showed 9 parts in 10,000. Sample No. 4 taken 10 :40 A. M. on the air before being washed showed 9 parts in 10,000. Sample No. 5 taken 10 :53 A. M. on the air before being washed showed 8 parts in 10,000. An average of the above set of readings given 6.5 parts in 10,000 after washing and 8.66 parts in 10,000 before washing. The above tests lead to the conclusion that the washer water absorbs a part of the carbon dioxide brought from the rooms by the air. It is impossible to say just what this rate of absorption is as the results varied considerably and the readings were not taken on the two sides of the washer at exactly the same time. More work needs to be done along this line. Tests were also made on the air in some of the rooms. In room No. 311 a sample was taken while the class was in the room and it contained 15 parts of carbon dioxide while a sample taken seven minutes after the class had left the room showed 13 parts of carbon dioxide in 10,000. On February 26th samples were taken in room No. 211 while 19 students were in the room. The first sample showed 15 parts, the second 13 parts, and the third 13 parts of carbon dioxide in 10,000. These tests were made in the afternoon, the room having been in use all day. On this same day samples were also taken in the auditorium near the close of an assembly period, and while there were 250 students in the room. The three samples taken showed 15, 20 and 17 parts of carbon dioxide, respectively. The carbon dioxide content shown by the above tests is con- siderably higher than would be considered good practice accord- ing to the old standard of 6 parts in 10,000, but it is lower than would be expected in view of the fact that all the air is recircu- lated and none is taken from the outside except the usual unavoid- able leakage. As far as could be detected the above amounts 24 Tests on the Recirculation of Washed Air of carbon dioxide had no detrimental effect upon the quality of the air in the rooms. The air seemed fresh and clean and no odors were detected. THE AIR WASHER AS A DUST REMOVER No actual dust counts were taken on the washer but it was readily seen that it was quite efficient in this respect. At the end of a week’s run the washer water was found to be very dirty and considerable sediment was found on the bottom of the tank. However, the writer doubts very much the claims made by most washer manufacturers to the effect that their washers will remove 98% of the solid matter in the air. BACTERIA TESTS These tests showed some startling and unexpected results. When using* recirculated water the washer supplied bacteria to the air instead of removing them and even when using new water continuously it did not show any marked efficiency as a bacteria remover. These results are directly contrary to the results re- ported by G. C. Whipple and M. C. Whipple of Harvard. (See the American Journal of Public Health, 1913, Vol. 23, pp. 1138- 1153.) The tabulated results of the tests are given in the following tables : Table V shows the results obtained with the use of sand and sugar filters. No conclusive results were obtained using- sugar filters, probably due to the trouble mentioned previously in the description of the apparatus. On an average, the sugar filters showed about the same number of bacteria on each side of the washer but the highest number would appear on one side of the washer in one test and on the opposite side in the next. The sand filters showed conclusively that the washer delivered bacteria to the air. The four sand filter tests checked each other very closely and showed an increase of bacteria in a ratio of about two to one. At the same time that the sand filter tests were taken, a series of Petrie plates were exposed for the same length ©f time on each side of the washer. These also show conclusively that the number of bacteria in the air is increased by the washer, but at a very much higher ratio than shown by the sand filters. The results of the tests with Petrie dishes are tabulated in Table VI. Figure VII is a plan of the washer and fan housing showing where the bacteria plates were exposed. The writer is unable to state just why the plates should show such a greater ratio above that shown by the sand filters. It can not be 6-0 Tests on the Recirculation of Washed Air 25 26 Tests, on the Recirculation of Washed. Air Table V Bactebia Tests on Air Washeb All samples taken with Sugar Filtegs andSand Filters fc IL 0 & UJ UJ Air before Washing Air after Washing < IL 0 111 5 F O “z Si F < Z iu 8 r u < < 0 <| K F 2 10 S 10 0 _J 8 z <2 Jg-S 2i-° to 5 P D Jan. is IO 2:06-2144 33 O .O . 232-337 4S 3 3 t l u Jan2S IO 230-252 22 25 2 27 230-252 22 2 0 2 r J JAN.26 IO 24S-3II 23 21 1 22 248-3: 1 1 23 11 0 11 1 V JAN.27 IO 232-228 56 12 II 23 232-32 3 51 20 0 20 ( ( 5 JAN28 '? 223-3:« 50 SO I SI 229319 50 2o 0 20 J <0 JAN?9 IO 2:30-3is 45 25 4 29 230-315 4S IO 0 10 FEB. 1 1° 2:43213 36 25 O 2S 243-319 36 33 0 39 FEB 2 IO 2 : 30-205 3S IO O 10 230-305 35 40 0 40 FE&3 IO CS4-22S 31 4S O 4S 1:542.25 31 198 2 Zoo FEB. 4 IO 242-225 43 6 2 6 242-325 43 16 1 17 V) Qi Feb.8 IO 3.00- 32S 25 30 2 32 3:00 325 25 70 4 74 FEB.3 IO 235300 2£ 33 O 33 23S-300 25 60 2 62 0 z 2 FEB-'O IO 255320 23 31 1 32 255-320 25 57 2 59 < FEB.I 1 IO 245310 25 36 0 36 245-310 25 66 O 66 mote T ME SAMPLE OF THE AIR BEFORE WASHING ON JANUARY 5- WAS taken directly in front of the washer and evidently water from the washer splashed into the filter tube. All other samples OF THE AIR BEFORE WASHING WERE TAKEN IN THE RETURN DUCTS WHERE IT WAS IMPOSSIBLE FOR THE WASHER WATER TO REACH THEM. The washer water was changed weekly. wholly explained from the different air velocities on the two sides of the washer. The average velocity of the air entering the fan was 940 feet per minute. The average velocity on the other side of the washer was 564 feet per minute in the north duct and 813 feet per minute in the south duct, and it can be easily seen that this will not account for all of the difference. The tests show no appreciable decrease in bacteria due to the washer water being changed each Monday. Evidently the wash- er walls were covered with bacteria from the water used the previous week and hence the high count at the beginning of the following week. After March 16th fresh water was used con- tinuously in the washer and tests were again taken March 25th and April 1st. These tests show an enormous decrease in the number of bacteria leaving the washer, but still most of the bacteria which came to the washer went through it. Tests on the Recirculation of Washed Air 27 sawsia i2*A3d mo NswvA s3aru.m3 a3wsv/v\ a/y no s-j-S-aj, viaajLovg 1A 3-iavi 0 z I 1 o< ,1U b li. < 0! < 3-LV-ld 33d sannoi/vi - vfl M N ff) ? 5 .S ^ N 3AV1cj a?d Via30Vg 3DV33AV lOOOvfio O^QOiO o 0 0 II o \8fOfQO 0 Q 52 S z UJ h 2 scnnow 0 ff 1 (VIN«l 0 «O^o viaQ-LOvg OOOO< 0 ®u)OqS ^-ONiOiOvO-O^'jJS < 33 xn/sjl(A 3 anSOdX 3 dO 3\ni± io io io iflo o o o o o f\jf\j(Vjcnj « 0 8 viaaxovg 9 vi) S ^ - i0 z: IU b < _l a saonoiA « ^ 8 via3xovg S3xOJiw aanscdxg jo 3 Wij_ 0 o 0 0 rQ rO (9 <0 N3XV1 33 31-1 aa 5 § $ ■§ -LS3J. dO 3J.VQ ? I >0 p z < “0 0 0 0 § y fu 0) (Vj 0 S bl 4 N Ji (XKinOO 3NU.VT39) viaaxsvg O 0 0 0 0 8 § 8 § § o ^ VI) If) (Vi 5 Aj.iaiQan_|_ Deooed Decided 2S 30 40 Kiomiui ^13d sano5 2 <5 X ? qS (Vj CO (VI - S >0 3UNIJ. 5 2: S £ x t < x x o- a o. S Si 3 $ 3 8 vo co id id co id XS3X| dO 3XVQ N N CO o - vj U 0 vj 0 O IU IU IU IQ 11 ij Q O Q Q Q O is? a i 28 Tests on the Recirculation of Washed Air After April 1st the water was changed daily and tests under these conditions were made on April 16th and April 23rd. The results show no advantage in changing the water daily as the number of bacteria leaving the washer is almost as great as when the water was only changed weekly. As a check upon the above results, Petrie plates were also ex- posed at the inlet and outlet registers of some of the rooms. Table VIII is a tabulation of these results. All of them show more bacteria coming into the rooms than going out. Bacteria tests were also taken on the washer water. Table VII shows the multiplication of bacteria in the water when it is recir- culated for a week and Figure VIII shows the same in the form of a curve. Table IX shows the results of plates exposed to the outside air. These were taken in comparatively still air and, when comparing them with the plates exposed in the ducts, the air velocity should be taken into consideration. Tests on the Recirculation of Washed Air 29 Fig. x. + 3 Fig. IX. 30 Tests on the Recirculation of Washed Air Figure IX is a photograph of two of the plates taken when the washer water was being recirculated for a week. No. 2 was exposed to the air after it had passed through the washer, and No. 3 was exposed to the air before it had passed through the washer. Figure X is a photograph of three plates taken after fresh water had been used continuously for ten days in the washer. No. 4 was exposed to the outside air. No. 5 was exposed to the air after it had passed through the washer, and No. 6 was ex- posed to the air before it had passed through the washer. All three were thirty minute exposures. Figure XI is a photograph of three plates exposed after the washer water had been changed daily for ten days. No. 7 was Tests on the Recirculation of Washed Air 31 exposed to the air before it had passed through the washer. No. 8 was exposed to the air after it had passed through the washer, and No. 9 was exposed to the outdoor air for the same length of time. CALCULATIONS CALCULATION OF WATER EVAPORATED IN WASHER The figures used below are average values from a six hour test that was made to ascertain the amount of water used by the washer. North Duct: Temperature of air 69.5°. Humidity 78%. The weight of water vapor per cubic foot at 69.5° and 78% humidity is 6.1252 grains. South Duct: Temperature of air 70.3°. Humidity 72%. The weight of water vapor per cubic foot at 70.3° and 72% humidity is 5.801 grains. AFTER WASHING Temperature of air 65°. Humidity 98%. The weight of water vapor per cubic foot at 65° and 98% humidity is 6.6466 grains. The weight of water vapor gained by air from North Duct is 6.6466 — 6.1252 = 0.5214 grains per cubic foot. The weight of water vapor gained by air from South Duct is 6.6466 — 5.801 — 0.8456 grains per cubic foot. Cubic feet of air per minute passing through the North Duct is 5,780, Cubic feet of air per minute passing through the South Duct is 8,340. Total weight of water gained by air from North Duct — 5780 x 60x0.5214 7000 x 8.3356 = 3J galIonS per h ° Ur ' Total weight of water gained by air from South Duct = >7000 x 8.335'6~~A 8340x60 x 0.8456> 7 - 25 gaU ° nS p6r h ° Un Total weight of water given to air by washer under above con- ditions of temperature and humidity = 3.1 -f- 7.25 — 10.35 gal- lons per hour. The washer tank dimensions are 8' x 5' x 1.25'. It was found by measurement that the water level decreased two inches in six hours, the make up water being shut off. Therefore by actual measurement the water given to the air by the washer under the above conditions was 8 x 5 x 2/12 x 7.5 = 8.34 gallons per hour. Tests on the Recirculation of Washed Air The discrepancy between the pounds of water actually meas- ured and the pounds of water calculated from the humidity tables is probably due to the inaccuracy of humidity measurement. The weight of washer water was obtained at the same time the above humidity measurements were taken. From these figures an interesting check on the volume of air passing through the washer can be made. The. average temperature drop in the air going through the washer was 4.9 degrees and the weight evaporated was 8.34 gal- lons per hour. 8.34 x 8.3356 = 69.52 pounds per hour. The temperature of the washer water was 64 degrees and the latent heat of vaporization at 64 degrees is 1,056 B. t. u. The volume of air which will be cooled 4.9 degrees by the evap- oration of 69.52 pounds of water is approximately 69.52 x 55 x 1056 , - — = 13,750 cubic feet per minute. 4.9 x 60 COST OF WATER FOR AIR WASHING To get the cost of air washing it is assumed that the cost of water is five cents per thousand gallons, as estimated by Mr. J. M. Smith from his cost data of operating University pumping plant, and that the washer is in service eight hours a day and five days a week. It is also assumed that, under the average run- ning conditions, the air passing through the washer absorbs ten gallons of water per hour and that the tank is filled to a depth of ten inches each time a change of water is made. Make up water is admitted to the tank by a float valve. Condition No. 1 : Washer water recirculated for a week, beginning with fresh water each Monday morning. Cost of water per week, 10 .05 [8 x 5 x — x 7.5 + 10 x 8 x 5] x = $0.0325 12 1000 Condition No. 2 : Washer water recirculated for 8 hours beginning with fresh water each morning. Cost of water per week, 10 .05 [ (8 x 5 x — x 7.5) x5 -f 10x8x5] x 12 1000 = $0.0825. Tests on the Recirculation of Washed Air 33 Condition No. 3: Fresh water used continuously. Under conditions No. 1 and No. 2 the water pressure on the spray nozzles averaged about 12 pounds per square inch gage. With 12 pounds pressure on the spray nozzles, measurements were taken to ascertain the cost of using fresh water continu- ously. With the drain from the tank closed it was found that it required two minutes for the water in the tank to rise eleven inches. Therefore the fresh water required when used continuously would be 5.5 8 x 5 x x 60 x 7.5 = 8250 gallons per hour 12 Cost of water per week, .05 8250 x 8 x 5 x = $16.50 1000 COST OF. POWER FOR RECIRCULATING WASHER WATER The average power required to drive the centrifugal pump was 2.41 K. W. The cost of this power for a week would be 2.41 x 8 x 5 x .025 — $2.41 The total cost of running the washer then is 2.41 .0325 = $2.4425 for Condition No. 1. 2.41 + .0825 = $2.4925 for Condition No. 2. and $16.50 for Condition No. 3. CALCULATION OF THE ECONOMY OF RECIRCULATION OF AIR In each of the runs made the air leaving the washer and enter- ing the rooms was at an average temperature of about 60°. The average outside temperature during the time the fan was running was 25.4°, 30.5° and 27° respectively for the first, second and third tests. The total amount of air delivered to the rooms in each case was approximately 15,000 cubic feet per minute. If this air had been taken from the outside instead of being recirculated, the additional steam required would have been 60 x 15000 — — — „ ^ x (60 — 25.4) = 590 pounds per hour for the first test. 55 x 960 The above calculation is based on the following figures : 34 Tests on the Recirculation of Washed Air The average steam pressure on the building during all three tests was 5.3 pounds gage and the average temperature of the condensed steam leaving the building was 209°. The quality of the steam entering the building averaged 98%. Therefore each pound of steam gave up .98 x 960 + 196.1 — (209 — 32) = 960' B. t.u. It is assumed that one B.t.u. will raise 55 cubic feet of air one degree of temperature. The additional steam required for the second test would have , 60x 15000 been ^ — x (60 — 30.5) = 503 pounds per hour and for the third test 60 x 15000 x (60 — 27) = 562 pounds per hour. 00 X J6U The actual weight of steam used by the building during the time the fan was running was 834.6, 573.1 and 824 pounds per hour for the first, second and third tests, respectively. The percentage of steam saved by recirculation of air in each 590 case is 590 + 834.6 & 41.4% for the first test. 503 503 + 573.1 562 and 46.5% for the second test. = 40.5% for the third test. 562 + 824 Since the fan in actual practice is only operated 8 hours out of 24, the saving from recirculation will not be as much as shown above. Running the fan 8 hours the saving for the entire 24 hour period 8 x 590 will be ---■ — - — — - — 16.8% from the first test and 8 x 590 % 23161 8x503 19.2% from Test No. 2. 8 x 503 + 16988 THE EFFECT OF THE FAN UPON THE STEAM CONSUMPTION TEST NO. I February 2nd and 3rd The total weight of steam used by radiators when the fan was in operation was 6676.5 pounds, or an average of 834.6 pounds per hour for 8 hours. Total weight of steam used when the fan was not in operation was 16483.5 pounds, or an average of 1030.2 pounds per hour for 16 hours. The difference in average steam Tests on the Recirculation of Washed Air 35 consumption is 1030.2 — 834.6 = 196.2 pounds per hour in favor of the fan. Since some of the above saving may be due to unequal outside temperatures during the two parts of the run, it is best to base the relative performance on the basis of per degree difference of temperature. Average pounds of steam per hour per degree difference of tem- perature while fan was in operation is 19.8 pounds. Average pounds of steam per degree difference of temperature per hour during time when fan was not in operation is 26.64 6.84 pounds. The difference is 6.84 pounds, or a saving of = 2o.ol 25.7% in favor of running the fan. TEST NO. II February 27th and 28th Total weight of steam used by radiators during the time the fan was in operation was 4,585 pounds, or an average of 573.1 pounds per hour for eight hours. The total weight of steam used during the time the fan was not in operation was 12,403 pounds or an average of 775 pounds per hour for sixteen hours. The difference in average steam consumption is 201.9 pounds per hour. Average pounds of steam per hour per degree difference of temperature during time that fan was in operation = 16.18 pounds. Average pounds of steam per degree difference per hour dur- ing time the fan was not in operation — 17.66 pounds. 1.48 The difference is 1.48 pounds, or a saving of ^ ^ ■= 8.38% in favor of running the fan. TEST NO. Ill March 3rd and 4th Total weight of steam used by radiators during the time the fan was in operation was 16,485 pounds, or an average of 824.2 pounds for twenty hours. Total weight of steam used by the radiators during the time the fan was not in operation was 6,369 pounds, or an average of 910' pounds per hour for seven hours. Difference in average steam consumption — 86 pounds per hour. 36 Tests on the Recirculation of Washed Air Pounds of steam per hour per degree difference during time the fan was in operation = 21.54. Pounds of steam per hour per degree difference during the time the fan was not in operation = 23.81. 2.27 The difference is 2.27 pounds, or a saving of —--—=9.53% io . 81 in favor of running the fan. A COMPARISON OF THE SAVING MADE BY KEEPING THE AIR IN MOTION AND THE COST OF OPERATING THE FAN Will it pay to operate the fan continuously? To answer this question the following calculations have been made : The cost of steam delivered to the building is approximately 27c per thousand pounds and the cost of power is approximately 2 */ 2 C per kilowatt hour. According to the figure obtained from the first test a saving of 25.7% in steam consumption was made by running the fan to keep the air in motion. The total pounds of steam used during the sixteen hours that the fan was standing idle was 16,483.5 pounds. Had the fan been running during this period a saving of 16,483.5 x .257 or 4,235 pounds of steam would have been obtained. 4235 Cost of steam saved ~ ^qq q~ x = $1-144. 516 x 7.85 Cost of operating the fan = fqoo — x 16 x .025 = $1.62. The saving in the second test would have been 12403 x .0838. or 1040 pounds of steam. 1010 Cost of steam saved - x 0.27 /ToooN Cost of operating the fan - — Ax 16 x .025 = $1,656. (51o X oj The volts and amperes used in the above calculations are aver- age values of the power consumption of the fan motor. The third test would show about the same relative cost as the second test. The second and third tests probably give the fairest idea of the saving of steam obtained by keeping the air in motion. Obviously the cost of power for the fan more than overcomes the saving resulting from its use. In all of the tests the air temperature was reduced about six degrees in passing through the washer. If the washer had been : $0.28 Tests on the Recirculation of Washed Air 37 shut down during the time there were no students in the building, a greater saving of steam would have resulted from keeping the air in motion. But even under such conditions, the saving would not have been enough to overbalance the cost of power for the fan. Take Test No. 2 for comparison. In this test the saving that would result from running the fan and washer for the additional sixteen hours was shown to be 1040 pounds of steam. The drop of temperature in the washer was six degrees and the air passing through it was 15,000 cubic feet per minute. The extra steam con- sumption required to replace the heat taken away by the of 1631 pounds - for the " sixteen hours. * 6 x 15000 x 60 washer water was — — — —7 r — £= 102 pounds per hour, or a total 55 x 960 of 1635 pounds for the sixteen hours. Therefore, with the fan running and the washer cut out, the total saving of steam would have been 1040 1635, or 2675 2675 pounds. The cost of this steam would be ■- x 0.27 = $0.72 as compared with $1,656 for power to run the fan. The steam cost used in the above calculations is below the aver- age and there may be conditions where high steam cost and low power cost will justify the running of the fan continuously. Lack of time made it impossible to check these last calculations with an actual test of steam consumption for recirculated air with the washer cut out. It would be interesting to see how it would work out in actual practice. STEAM USED PER SQUARE FOOT OF RADIATION PER HOUR The steam consumed by the radiators per square foot of radia- 23161 tion per hour in the first test was 24 x 8530 — ■ — 0.113 pounds 16988 24 x 8530 = 0.083 pounds for the second test 22854 and — = 0.0993 for the third test. /v i X 0O0U The average for the three tests is .0984, or approximately l/10th of a pound of steam per square foot of radiation per hour. Since an average radiator will condense l/4th of a pound of steam per square foot per hour under extreme conditions the re- sults show that the building has ample radiating surface for any conditions which may arise. 38 Tests on the Recirculation of Washed Air TOTAL COST OF HEATING AND VENTILATING THE BUILDING FOR A TYPICAL DAY OF 24 HOURS Data taken from Test No. I .0325 Cost of washer water = = $0.0065. 5 2.41 Cost of power for washer pump =- — — — $0.48. _ 518x7.85 Cost of power for ventilating fan = x 8 x .0252 — $0.81. Cost of steam for heating = 23161 x .27 = $6,253. Total cost per day = $7.55. To the above total cost should be added the cost of running the two small exhaust fans in the attic. The hot water tanks were cut out during the tests so that no data is available for the cost of steam for heating water. The recirculating system was put into operation at the be- ginning of school, September 1914, and was in operation through- out the whole school year. The principal tests were made in the months of January, February and March, and the last of the bac- teria tests was made in April. The questionaire was submitted to the teachers on May 13th and the reports were handed in by them about three days later. This was about two weeks before school closed, the teachers having had nine months to form their conclusions in regard to the system. An interesting point in regard to the system came out about two weeks ago (Sept. 1915). The fan motor burned out and it took several days to get it into running condition again. The teach- ers were not aware that anything had happened but complaints soon began to pour in that there was something wrong with the ventilation. QUESTIONAIRE SUBMITTED TO TEACHERS A questionaire was submitted to the various teachers in the school in order to get their opinion of the system and its effect upon the students. The questions were as follows : 1. Has your room been sufficiently heated? 2. Has the ventilation been satisfactory? 3. Has the air in the room been too moist or too dry? Tests on the Recirculation of Washed Aii< 39 4. lias the moisture in the air had a good or bad effect upon the quality of the air in the room? 5. If the moisture has had a bad effect please state what the effect is. 6. Has the room seemed close? 7. Have you noticed any unpleasant odors? 8. Have the students been bright and alert or (frowsy and list less? 9. How does the ventilation in this building compare with that of other buildings in which you have taught? 10. How does the general health and alertness of the students in this building compare with that of students in other buildings in which you have taught? RESULTS FROM THE QUESTION AIRE Question No. 1. All of the fifteen teachers answered “yes”. Question No. 2. Twelve teachers answered “yes.” Teacher J answered “nearly so”, Teacher M, “no”. Teacher N, “In the morning, yes; in the afternoon, no”. Question No. 3. Teacher A — “Rarely too moist”. Teachers B, G, I and K— “No”. Teachers C, D and H — “Neither”. Teacher E — “If anything, too moist at times”. Teacher F— “Yes”. Teachers J and N — “A little dry”. Teachers L, M, and O — “Just right”. Question No. 4. Ten teachers answered “Good”. Teacher A — “Not strictly bad at any time”. Teacher B — “Apparently good”. Teacher F — “Did not notice.” Teacher L — “Have not noticed any bad effects”. Question No. 5. No statements given by any of the teachers. Question No. 6. Teachers A and C — “Only when it is first opened in the morn- ing”. Teacher B — “Very seldom.” Teachers D and G — -“No”. Teacher E — “No, until recently”. 40 Tests on the Recirculation of Washed Air Teacher F — “Not usually”. Teacher H- — “In warm weather if windows are closed, yes”. Teacher I — “At times windows have had to be opened”. Teacher J— “A little”. Teacher K — “No, except in the morning before the fan had run sufficiently”. Teacher L — “Sometimes”. Teacher M — “In room No. 115, no ; in room No. 311, yes”. Teacher N— “Yes”. Teacher O — “Not often”. Question No. 7. Twelve teachers answered “no”. Teacher M — “Room No. 115, no; room No. 311, yes”. Teacher N— “Yes”. Teacher O — -“Seldom”. Question No. 8. Six teachers answered “Bright and alert”. Teacher G — “Depends more on students than on room condi- tions”. Teacher H— “Neither, some of each in each class.” Teacher I — “Not drowsy and listless as a rule”. Teacher J — “Both, but listlessness not due to ventilation”. Teacher K — “We have both kinds, but the room conditions did not make them so”. Teacher L — “Attitude varies”. Teacher M — “Both”. Teacher N — “Bright in morning; drowsy from 2:30-3:30 P. M”. Question No. 9. Six teachers answered “Better”. Teacher A — “Infinitely better”. Teacher B — “Better (have taught in six schools”). Teacher C — “Most favorably”. Teacher D — “Best in my experience”. Teacher E— “There is much more moisture in the air”. Teacher K — “Excellent”. Teachers M and N — “Favorably”. Question No. 10. Teachers A and G — “Better”. Teacher B — “Much better”. Teachers C, I and L — “Favorably”. Teacher D — “Have not noticed health especially. Should call alertness better than usual class”. Tests on the Recirculation of Washed Air 41 Teacher E — “Have no data”. Teacher F — “Don’t know of any”. Teacher H — “Not so alert here, health better”. Teacher K — “About the same, perhaps somewhat better”. Teachers M and O — “About the same”. Teacher N — “Better than elsewhere”. CONCLUSIONS 1. The tests show that it is both unnecessary and uneconomic cal to supply large volumes of air to obtain good ventilation. 2. That 15 cubic feet per student would be ample providing it enters the room at a fairly high velocity and carries the proper amount of moisture. 3. With humidity ranging from 50 to 70 per cent., the occu- pants of the rooms are perfectly comfortable at a temperature of 65 degrees or even less. 4. With humidity of about 60 per cent., the air can enter the rooms at a temperature of 60 degrees without creating any dis- comfort ; in fact it seems to give life to the air and aids in the efficiency of ventilation. 5. Carbon dioxide content as high as 20 parts in 10,000 does not have a bad effect upon the ventilation. 6. Ventilation by recirculation is both efficient and economi- cal. At the end of a year’s run the teachers are almost unanimous in their praise of the system. 7. With a recirculating system such as this, it requires from 40 to 50 per cent, less steam to heat the building while the fan is in operation than would be required if the air was drawn from outdoors for the same length of time. 8. Air movement keeps the temperature uniform in various parts of the room and decreases the amount of steam required for heating. The tests show a minimum saving of about 8 per cent, due to this air movement and this would be true whether the system is a recirculating one or otherwise. 9. The air washer absorbs a considerable amount of the car- bon dioxide contained in the air passing through it. 10. The air washer is apparently quite efficient as a dust re- mover but it does not remove bacteria from the air when the washer water is recirculated. The tests show that it actually supplies bacteria to the air under such conditions. 11. In spite of the poor showing of the washer, the air enter- ing the rooms carries no more bacteria than outside air when the 42 Tests on the Recirculation of Washed Air relative velocities in which the plates were exposed is taken into account. The writer plans to continue these tests this coming winter with a view of ascertaining the effect of recirculation upon the oxygen content of the air. Further tests will also be made to determine more accurately the amount of carbon dioxide absorbed by the washer, and experiments will be carried on with various methods of eliminating bacteria from the washer water. ADDENDUM* A few words of explanation may be added in regard to the low values of carbon dioxide found in the tests. The writer makes no assumption that the building is air tight. The tests were made under actual conditions as they exist at the school. The windows are of such construction that less leakage takes place than is usual in most school buildings, and careful observa- tions of the building showed that none of the windows were open during the test or during most of the school year. But of course leakage took place, and in accounting for the low values of carbon dioxide found, the following points must also be borne in mind : Room No. 15 (see Table I), which is the gymnasium, was oc- cupied only late in the afternoons, and whatever leakage that took place into this room during the day was distributed through all the other rooms of the building by means of the recirculating system. The same is true of the Auditorium, Room No. 214. The Auditorium was used only for assembly periods two or three times a week and then only for an hour a day. At all other times the leakage into this large room was distributed through all the other rooms. And, during assembly periods all the class- rooms are unoccupied and a large part of the leakage into them finds its way into the Auditorium. The same action takes place from Rooms No. 108 and No. 209. These rooms are small auditoriums which are only used for short periods once or twice a week. There was an average attendance of about 250 students in the building throughout the school year. The figures in Table I are based upon the student capacity of the rooms and not upon the actual attendance. The average attendance for the year for the various class rooms is given below. * Added on the recommendation of the Publicatiton Committee. Tests on the Recirculation of Washed Air 43 Room Number Student Capacity of Room Average Attendance 8 20 16 10 28 16 106 26 20 107 26 11 109 26 20 115 31 26 207 31 22 208 26 17 210 26 20 211 31 23 218 20 17 308 41 20 309 25 16 311 15 14 312 14 14 314 29 16 Average 26 Average 18