UNIVERSITY OF CALIFORNIA COLLEGE OF AGRICULTURE AGRICULTURAL EXPERIMENT STATION BERKELEY, CALIFORNIA THE SOLAR HEATER A. W. FARRALL BULLETIN 469 June, 1929 UNIVERSITY OF CALIFORNIA PRINTING OFFICE BERKELEY, CALIFORNIA 1929 THE SOLAR HEATER A. W. FAREALLi Many people living' in the interior valleys of California, where sun- shine is abundant, have found solar heaters practical and satisfactory for supplying hot water. Although many of the heaters are 'home made,' there are also a large number of commercial manufacture. The average household needs much water at temperatures of 120° to 150° F, and more would be used if the cost of heating could be reduced. Dairies and manufacturing plants also use much water at temperatures well below that obtainable with solar heaters. The University of California has had numerous requests for infor- mation regarding the construction and performance of solar heaters. These, together with a desire to assist the California farmer in obtaining- an economical means of warming water for his household and dairy, have resulted in a study of the problem by the Agricultural Engineering Division. The points investigated were : 1. The availability of solar energy. 2. Design of a practical solar water heater. 3. The heat-absorbing power of a simple glass-covered absorber with a plain iron coil. 4. The heat-absorbing- power of an open-type absorber without glass, and with a plain iron coil. 5. The heat-absorbing power of a glass-covered absorber with an iron coil embedded in concrete. 6. The heat absorbing power of an absorber without a glass cover but with the water coil embedded in concrete. 7. The characteristics of a recirculating type of solar heater. 8. The effect of insulation of the absorber upon the temperature obtained. 9. The operation of a solar heater under practical conditions. AVAILABILITY OF SOLAR ENERGY Studies were made of the United States "Weather Bureau records at various California and Arizona stations to determine the average hours of sunshine per month and the average number of clear and cloudy days during the year. Data from these studies are shown in tables 1 and 2. It will be observed that the percentage of sunshiny i Assistant Professor of Agricultural Engineering. Eesigned December 31, 1928. 4 University of California — Experiment Station days for the months of December, January and February at weather stations located in the Sacramento and San Joaquin valleys is rela- tively low in comparison to the Arizona stations. The latter are more typical of the weather conditions in the Imperial Valley. TABLE 1 Average Hours of Sunshine per Month at Various Weather Bureau Stations Station Average sunshine per month Average sunshine for January Average sunshine for July Average for December, January, and February. Hours per month Average for other nine months Hours Per cent of possible Hours Per cent of possible Hours Per cent of possible Hours per month Sacramento Fresno Phoenix 289 6 293 2 314.2 75 76 84 104.9 134.0 236.0 34 44 75 449 430 401 99 96 94 158.3 153.3 238.7 333 5 340 1 339.4 * Averages for California stations are for 1887-1925. Average for Arizona station is for 1895-1926. TABLE 2 Average; Clear, and Cloudy Days During the Year at Various Stations (Averages from continuous records for 15 or more years.) Yearly average Station Clear days* Partly cloudy days* Cloudy days* 224 175 231 168 236 296 77 134 71 112 89 53 64 56 63 85 40 16 *Clear=sun obscured for 0.0-0.3 of day; partly cloudy = obscured for 0.4-0.7 of day; cloudy = sun obscured for 0.8-1.0 of day. Figure 1 shows the hours of sunshine for each month at Fresno, California, and figure 2 shows the same for Phoenix, Arizona, in graphic form. The Solar Radiation. — The solar constant, K, which changes slightly from year to year because of disturbances in" the solar system, is usually evaluated by giving the quantity of heat in small calories received in one minute from the sun at its mean distance from the earth, by one square centimeter of a perfect absorbing surface pre- sented at right angles to the sun's rays. The actual radiation at the earth varies from day to day and from hour to hour. It is largely Bul, 469] The Solar Heater dependent upon the clearness of the atmosphere, that is, the freedom from smoke, water vapor, and dust particles. The energy as measured by different investigators varies considerably. Table 3 shows some of the values obtained. 500 400 300 ^200 100 0) 4 ■3> 5 > 1 i 1 Aremqe i 1 1 1 1 — I I I i Fig. 1. — Monthly distribution of hours of sunshine at Fresno, California, 1887-1925. 500 400 Average __J__ --■ 300 ^200 $ JK. I t IOQ I I 1 i i 1 % X •8 ! i 1 1 3 — Fig. 2. — Monthly distribution of hours of sunshine at Phoenix, Arizona, 1895-1926. TABLE 3 Value of Solar Constant as Determine© by Various Investigators Observer K* Place Viole Langley Savelief Angstrom C. G. Abbott 2.5 3 3 4 4.0 1.925 Top of Mt. Blanc Top of Mt. Whitney Kief (ground covered with snow) Ixele Numerous observations (considered most accurate) Expressed as small calories per square centimeter per minute. 6 University of California — Experiment Station For practical conditions in California the radiation at the earth's surface may be taken as roughly 5 B.t.u. 2 per sq. ft. per minute. The intensity of the sun's radiation is important because it varies considerably during the day. No specific data are available for Cali- fornia conditions, but data given in Bulletin 261 of the Vermont Experiment Station show that the radiation per square centimeter per minute at the earth on a clear day may vary from 0.94 to 1.36 calories between 9 :20 a.m. and 4 :53 p.m. From the foregoing, it is evident that the amount of energy which can be taken from the sun is variable. Data show that over a period of nine months of the year in California, there is an average of 333.5 hours of sunshine per month in the interior valleys, or an average of 11.1 hours of sunshine per day. During the other three months, the average is only 153 hours per month, or 5.1 hours per day. It would seem that solar heat could not be depended upon during this time. In localities such as Phoenix, Arizona, conditions would be much better, because the minimum average sunshine for the three winter months is 238 hours. DESIGN OF A SOLAR WATER HEATER One should consider the number and distribution of days of sunshine in a locality before deciding upon the installation of a solar heater there. It is possible to obtain records from the United States Weather Bureau office covering these points. One can then decide whether or not there is enough sunshine to pay to make the installa- tion. In the warm interior valleys of California it is usually consid- ered sufficient for water heating seven to nine months of the year. In general, it seems that there should be an average of at least six hours of sunshine a day during the period in which the solar heater is used. The solar heater is normally used for heating water for general household purposes, dairies, or summer resorts, or temporary quarters used only in the season of the year when there is an abundance of sunshine. The use will largely determine the type of construction. Figure 3 shows an artistically designed heater mounted upon the roof of a building in such a manner that it is attractive. Many installations, such as that for a dairy, need not be so elaborate. 2 B.t.u. — British thermal unit, or the amount of heat required to raise the temperature of one pound of water 1° F. Bul. 469] The Solar Heater The temperature obtained by a solar heater is dependent largely upon the relationship between the size of absorber and the amount of water used and upon the insulation of the system. By using' good insulation on the absorber and tank, and all connecting lines, one can obtain a water temperature approaching the boiling point. Insula- SSSsvMPwHHSSHHIIIHHP^i-. ^ .&»& • Fig. 3. — Commercial-type solar heater installed on a dwelling house in southern California. Note that the absorber is made in two sections. tion is of further value in making the apparatus more efficient on cold and cloudy days. It also protects pipes from freezing during moderately cold weather. Types of Solar Heating Systems. — The solar heating system may be constructed simply as shown in figure 4, or it may be more elaborate as shown in figure 5. — -~ To tap Solar absorber} Oil or electric owi/iory beater Fig. 4. — Diagram of a simple solar water heater system. Fig. 5. — Diagram of a pressure-type, recirculating, solar water heater equipped with auxiliary heater. The principal elements of the system are (a) the absorber, (b) the storage tank, (c) the connections, and (d) the auxiliary heater. 8 University of California — Experiment Station The general arrangement and action are as follows: The system is filled with water (or other liquid), which as it becomes heated from contact with the absorber a, rises and circulates through a pipe "X" to the top of the tank, b replacing the cold water, which settles down to the bottom and which in turn circulates to the bottom of the absorber, where heating takes place ; the cycle is then repeated, thus forming a continuous thermo syphon system. Warm water is drawn off from the top of the tank and unheated is added at the bottom, thus making warm water available before the entire tank is heated. The advisability of an auxiliary heater (see fig. 5) should be con- sidered, for it is usually best to use an oil, gas, or electric heater to warm the water when the sun does not shine. Many people find that hot water coils placed in the furnace take care of their problem very satisfactorily. The non-freezing type of system may be used in a cold country; it is ordinarily best, however, to drain the absorber in cold weather and not use it until winter is past. The Solar Absorber. — The absorber is the heart of the solar heater, for it must trap the sun's rays and change these from light into heat. A practical absorber should embody the following features: (a) The absorber surface should be placed with its face at an angle of 90° with the maximum light rays, (b) The absorber surface should be of such color and materials as to change this light into heat ; black sur- faces are best, (c) The absorber surface should be covered with a glass to serve as a valve to admit light to the surface and to restrict the escape of the heat, (d) The absorber should be insulated to prevent conduction of heat to media other than the water, (e) The water coil should be of sufficient area and conductivity to transmit the heat effectively to the water. It is not practicable to build an absorber which will automatically hold its surface at right angles to the light rays. Movable absorbers are intricate, expensive, and difficult to keep in order. For these reasons the stationary type is preferred, even though it must be made larger than the movable type. For general use the absorber should be at an angle of 35° to the horizontal at latitudes the same as Davis, California. Farther north it should be at a greater angle. The absorber should be located preferably on a roof with south exposure and protected from wind and shade. It may need to be built in two sections. For thermo-syphon action, it should always be installed with the bottom lower than the bottom of the storage tank, as shown in figure 6. Bul. 469 The Solar Heater The size of the absorber will be determined by the amount of water to be heated, the temperature rise required, and the time of exposure. Under normal conditions, one square foot of absorber surface should be allowed for each gallon of water to be heated to a temperature of 140° to 150° P. \\a*?/" \\X/9/ce* \ I 1 — — •- \\_ \ 7 Storage tank \X North House Fig. 6. — Diagram of a solar heater installed on a roof. The following bill of materials shows the approximate cost of a solar heater of a size to supply 40 gallons or more of hot water a day. Under favorable conditions it should supply much more than this amount. Materials required to build a solar water heater of 40 gallons per day capacity : MATERIALS COST Absorber box — 8, 18" X 48" single light window sash $12.80 2 pes. 2" X 4" X 12' S.1S.&1E 1.15 2 pes. 2"X4"X4' S.1S.&1E 0.35 12 pes. 1" X 4" X 12' T.&G. redwood 3.00 50 sq ft. 2" cork board 10.00 7 pes. V 2 " X 3" X 4' battens 0.70 8 pes. %" X 4" X 4' cleats 1.00 3 doz. 2Y 2 No. 12 wood screws 0.50 Assorted wire nails 0.50 y 2 sack Portland cement 0.50 i/ 5 cu. ft. lime 0.25 4 cu. ft. sharp sand 0.40 1 qt. black asphalt paint 1.00 1 qt lead and oil paint 1.00 Coil- 121 ft. %" standard black iron pipe (random lengths) 0.60 (10 pes. 10' length) (2 pes. 10' 6" length) 10 University of California — Experiment Station 11 %" close return bends 2 -20 24 pipe straps for %" pipe 0.50 2 %" tees - 1-00 2 %" plugs , 0- 30 Connections — 25 ft. %" galv. iron pipe 2.25 Tank— 1 40 gal. range boiler 12.00 24 sq. ft. 2" cork board 5.00 Total $66.00 B c"'Cork board or J-'TsCz \ J other insubtor ir\4"C/eat Fig. 7. — Detail of solar heater construction. A, Front view of absorber, showing method of placing coils and glass. B, Cross section of plain uninsulated absorber. C, Cross section of well insulated absorber. Note method of placing concrete about coils to improve conductivity. D, Method of using fins on pipes to increase efficiency. E, Cross section of insulated tank set-up. Insulation may be any dry, light- weight material such as cork, asbestos, magnesia, rags, saw-dust, or rice hulls. Bul.469] The Solar Heater 11 The box shown in figure 7 is easily constructed. Figure 1A shows the general shape and the manner in which the coils are installed ; figure IB shows a detail of the cross section of a simple uninsulated box. The box is made for standard size window panes, which can be screwed down tight against the edges. The upper part of the sash can be covered over with sheet metal, as shown, to prevent entrance of moisture. Figure 7(7 shows recommended construction when insulation is used. It is similar to that in figure IB except that the bottom of the box is filled with 2 in. of cork board to which the coils are attached. Note that a concrete mixture is troweled in around the pipes to improve their conductivity. This mixture is made of 1 part cement, 4 parts sand, and % part lime. The concrete should fill the box to % the depth of the coils, as shown in detail of figure 7(7. The glass may be placed in 1 or 2 layers as shown. The double glass and 2 in. of insulation are desirable if high temperatures and efficient heat utilization are desired. Wooden or metal sash may be used. Narrow cross members which offer little obstruction to sunlight are preferred. A 21-oz. glass is satisfactory for ordinary sash, although lighter weight may be used if the sections are small. Care should be taken to see that the glass is tightly set in good putty in order to make an air and water-tight job. Coils may be made readily of stan- dard pipe, by the use of return bends. For most efficient absorption of heat the box should have the pipes as close together as possible, and then the intervening spaces should be filled with concrete, as shown in figure 7(7. Care should be taken to spread the coils as shown in figure 7 A, making the distance a always greater than distance b, in order to provide good circulation and to avoid air pockets. The size of pipe depends upon the capacity of the heater, but ordinarily %-in. pipe will be found satisfactory ; it is large enough to give good circulation and at the same time is easy to handle. Figure ID shows how pipe with metal fins could be used for increasing the surface exposed to the sun's rays. When pipe is cut for the coils, all except the bottom and top pipes should be made of equal length; these two, however, should be made long enough to extend 4 in. beyond the edge of the absorber box, so that union connections can be made. The joints should all be screwed up tightly and properly leaded, using white or red lead. The number of feet of pipe can be reduced by the use of finned pipes as mentioned above. The principal consideration, if high effi- ciency is desired, is to make the area of pipe exposed to the direct 12 University of California — Experiment Station rays of the sun as near the area of the glass as possible, and to use concrete or other conducting material to fill in the spaces between the pipes. Painting is desirable for two reasons : first, for protection from the elements ; and second, for increasing the efficiency of heat absorption. The outside of the box should be given 3 coats of a good lead and oil paint, red lead being used on all metal surfaces, as a primer. The outer coats can be any color which harmonizes with the surroundings. The pipes and inside of the absorber box should be painted a dull black, for this color is most effective in absorbing the light rays and converting them into heat. The connection between the absorber* and the storage tank should be as direct as possible and with no loops or sags which might cause air to trap in the system. Unions should be provided on both the inlet and outlet of the absorber in order to facilitate erection and repairs. The pipe should be of the same size as the pipe in the coils. It may be necessary to install a valve in one of the lines, preferably the discharge line between the absorber and the tank, to prevent back circulation at night. In a pressure system, the discharge line from the absorber to the tank should be brought in at the top of the tank if hot water is desired quickly. For an open system, such as that shown in figure 4, the dis- charge pipe from the ajbsorber to the tank must always be below the water level in the tank, or circulation and heating will not take place. The Storage Tank. — The storage tank may be an ordinary range boiler, as in the case of the pressure system (fig. 5) or it may be a barrel or tank, as illustrated in figure 4. The size of the tank must conform to the water requirements and absorber area. For ordinary household use, a 40 or 60-gallon tank is desirable. For high efficiency, insulation should be provided around the tank, as shown in figure IE. A satisfactory method is to apply 2 in. of asbestos, sil-o-cell, magnesia, or cork insulation material with metal housing. A cheaper method is to build a wood box about the tank and fill in between the tank and the box with a coarse dry material such as sawdust, ground cork, or rice hulls. All water lines, especially those to and from the absorber, should be insulated with asbestos or magnesia. Care of the Heater. — Proper care of the solar heater consists mainly in keeping the absorber glass clean and free from dirt. If it is mounted on a roof the best method of cleaning is to hose it off at weekly or monthly intervals. The absorber box must be kept tight and well painted with a good lead and oil paint. Neglecting to drain the water in freezing weather may result in burst pipes. Bul. 469] The Solar Heater 13 TESTS OF SOLAR HEATERS AND ABSORBERS In order to study the application of solar heaters for water heat- ing- a number of tests have been conducted. The apparatus used consisted of a solar absorbing device, as shown in figure 8. The general set-up was an absorber mounted on a skid and so arranged that its angle of incidence to the sun's rays could be varied as desired Drilled hole G/oss Pipe cop ORIFICE DETAIL. Concrete, pointed block- CROSS - SE C-TION SOLAR HEATER TEST APPARATUS SIDE. VIEW Fig. 8. — Diagram of apparatus used in testing solar absorbers. by adjustable arm D. A float chamber (as shown) was fitted with a float valve B and an orifice C. The absorber had the following' dimen- sions: glass area 10.8 sq. ft., length 41 in., depth 1% in., and width 41 in. Provision was made to add cork insulation if desired. The coil was made of %-in. black iron pipe, welded together as shown in figure 9, with the pipe in an oblique position. During a part of the tests the pipes were embedded in concrete, as will be described later. The coil area was 10.8 sq. ft., constructed from 13 pieces of %-inch pipe 36 inches long and 2 pieces of 1-inch pipe 40 inches long. The float chamber was connected to the University Farm water supply of approximately 40 lbs. pressure, and water entered at A 14 University of California — Experiment Station (fig. 8), passed through valve B down through the orifice C of 0.10-in. diameter, and on to the lower part of the absorber E. It was then forced up and out to the overflow at F. The funnel at C was higher than the top of the absorber, so that there was always sufficient head to force the water through, without backing up and running over at the funnel. The orifice (detail drawing in fig. 8) shows how this part was constructed through the use of a regular galvanized iron pipe cap. The apparatus was placed in a position on the ground about 15 feet from the south side of a large building, so that it was more or less protected from the north winds. Fig. 9. — Set-up of apparatus used in the study of solar absorbers. Thermometers used were all of the standard chemical type and calibrated for correctness within the ranges for which they were to be used. Thermometer No. 1 (fig. 8) was used to measure the air temperature and was hung in the shade of the platform of the appa- ratus. Thermometer No. 2 was placed in a tee in the water inlet at the bottom of the absorber. Thermometer No. 3 was placed in a tee at the water outlet of the absorber. Thermometer No. 4 was used to take the temperature of the water at the orifice. The temperature variation at this point was not sufficient to cause a great variation in flow through the orifice. Thermometer No. 5 was laid against the bottom, or back, of the absorber, so that its bulb was % i ncn ou ^ from the surface (fig. 9). Other apparatus consisted of a y 2 -ga\lon receptacle for checking the flow of water, also a scale for weighing the water flow. Bul. 469] The Solar Heater 15 Method of Conducting the Tests. — The apparatus set up as described, was operated for a number of days on preliminary runs, before the test data were taken. The data obtained show a flow of from 1.22 to 1.30 lbs. of water per minute, with an average of 1.27 lbs., which has been taken as the flow for all calculations. A good agreement of the data taken on successive days shows that the vari- ables were well controlled. No attempt was made to measure the solar radiation constant during these tests. It is believed, however, that the radiation was so nearly the same on the bright, clear days during which the tests were made that a fairly close comparison of data can be made. It is evident that of the variables tested, all were of such magnitude that their effect was easily noticeable. A series of tests was laid out, calling for the measurement of tem- perature characteristics, and of heat absorption, using various com- binations of absorbers and glass. In carrying out the tests, the water was turned on at 8 a.m. each morning and the temperature of the various thermometers read at 1-hour intervals until 5 p.m. The flow was determined by catching the water for 2 minutes at the overflow (F, fig. 8) of the absorber. Special procedure is described under each test series. A Simple Glass-covered Absorber with an Iron-Pipe Coil. — The object of this series was to determine the heat-absorbing characteris- tics of an absorber, glass covered and painted black, fitted with plain black iron coils. The apparatus consisted of the standard absorber box described above. The black iron coil pipes were attached to the back of the absorber box, which was made of %-in. matched wood flooring. No special insulating material was used. The glass cover was standard 21-oz. window pane. TABLE 4 Tests of Plain Glass-covered Absorber With Black Iron Coil Temperature, ° F B.t.u. absorbed per minute B.t.u. Test Atmo- sphere Absorber Water absorbed per minute per square foot In Out Rise of absorber A-l 58.9 65.0 69 6 64.5 94.9 104.7 105.4 101.6 65.7 72.0 73.7 70 4 81.1 88.0 89.5 86.2 15.4 16 15.8 15.7 A-2 A-3 19.9 1 85 Water flow was at the rate of 1.27 pounds per minute. Tests started at 8 a.m. and ended at 5 p.m. 16 University of California — Experiment Station The absorber was exposed to the sun at right angles to the rays at 12 o 'clock noon. The standard set-up and procedure were followed. Table 4 and figure 10 show the data obtained. From table 4 it will be noted that the average temperature rise for the day was 15.7° F, with a minimum of 15.4° F and a maximum of 16.0° F. This would make the heat absorption by the water 19.9 B.t.u. per minute, or 1.85 B.t.u. per minute per sq. foot of glass area. // 12 I Time of day ■pig, 10. — Temperatures of the absorber and water during the test of a plain glass-covered absorber fitted with plain iron pipe water coil. Figure 10 shows that the temperature of the water and absorber, as well as the temperature rise of the water, varied during the day. It will be noted also that the absorber temperature rose rather quickly and reached a maximum of 126° F between 12 o'clock noon and 1 o'clock p.m. As the angle of the sun became more oblique, the rate of absorption was reduced. It is significant that the maximum tempera- ture rise of the water was 22° F, which was at 12 m., and that the minimum was 8° F at 8 a.m. There seemed to be a considerable difference in temperature between the water and the absorber air. This difference reached a maximum of 50° F at 12 m. to 1 o'clock p.m. Apparently more coil surface or a better heat-transferring medium was needed to keep this absorber temperature down and prevent excessive loss from conduc- tion through the walls. Bul. 469] The Solar Heater 17 An uncovered Absorber with Plain Iron Coil. — The object of this series was to obtain data regarding the effect of the glass cover on an absorber, and to determine the efficiency of an uncovered absorber. The apparatus was the same as in the first series except that no glass was used on the absorber. The procedure also was the same. TABLE 5 Tests of Plain Absorber Without Glass Covering Temperature, c F B.t.u. absorbed per minute B.t.u. Test Atmo- sphere Absorber Water absorbed per minute per square foot In Out Rise of absorber B-l 72.0 70.1 70.5 70.86 78.0 80.5 79.4 79.3 74 3 76.2 76.3 75.6 83.0 85.6 85.4 84.6 8.7 9 4 9.1 9.06 B-2 B-3 Average... 11.5 1.064 Water flow was at the rate of 1.27 pounds per minute. Tests started at 8 a.m. and ended at 5 p.m. 170 1 Water r/ow - i.cs pouna^ per m/nan Absorber- g/oss removed, Jbhck. s 2. *ofionar% A 90' to sun of noor h % Area - i ^qc/o re mer er. \ x*IIO 5 ^90 Averages Absorber y\ 'afer ot *1 H •— ■"■""* _X- — — ■— ■* ■ — — — . " ? __z_ r~ Woicx in± "ZT^^ XAir- . ~Z^- — ■ ___——- __ .50 ^Water temperature rise i - r i l '*'———.». 20 ^ 10 % O I 9 10 II 12 J £ A.M. 7/me of day 4 PM. Fig. 11. — Temperatures of the absorber and water during the test of a plain absorber, without glass, but with iron pipe water coil. Table 5 and figure 11 and figures 14, 15, and 16 show the data obtained. From table 5 it is seen that the average temperature rise per day during all tests was only 9.06° F, with a minimum of 8.7° F and a maximum of 9.4° F. 18 University of California — Experiment Station This shows a heat absorption by the water of 11.5 B.t.u. per minute or 1.069 B.t.u. per square foot per minute. Figure 11 shows how the various temperatures ranged during the day. It is evident that the absorber temperature is very low, in fact lower than the water temperature in the hottest part of the absorber ; heat, therefore, is not only not absorbed from the air but is actually given off or lost to the air from the pipes. The only really effective heating is that which is absorbed from the light rays which strike the pipes directly. It appears that the efficiency of this type of absorber is very low, but it could be greatly improved by making the coil surface against which the light impinges of practically the same area as the absorber. A flat tank used in place of the coil would be typical of such a design. The addition of material such as concrete between the pipes would also tend to conduct the heat to them more readily and thereby increase the efficiency. A Glass-covered Absorber with Iron Coil Embedded in Concrete. — The object of this series was to determine the heat-absorbing char- acteristics of the absorber, when the pipes were embedded in concrete. The same equipment was used as in previous tests, but the coil was embedded in a 1-4 mix of concrete with the surface enameled black. The procedure was the same as in the first series. The theory of embedding the coils is that concrete, having a higher conductivity than air, will carry the heat to the coils, with a lower temperature differential, so that for a given amount of heat absorbed in the pipes, a lower absorber temperature will be required. This results in less radiation from the exterior of the absorber, and conse- quently a larger proportion of the heat liberated in the absorber is utilized in heating water. TABLE 6 Tests of Glass-covered Absorber with Coils Embedded in Concrete Temperature, c F B.t.u. absorbed per minute B.t.u. Test Atmo- sphere Absorber Water absorbed per minute per square foot In Out Rise of absorber C-1 C-2 C-3 70 3 70.9 74 1 105 104.9 106 5 79 79.3 81.1 98 98.8 99.1 19 19 5 18.0 24.13 24.76 23.00 71.89 23.96 2.24 2.29 2.13 Average .... 2.23 Water flow was at the rate of 1.27 pounds per minute. Tests started at 8 a.m. and ended at 5 p.m. Bul. 469] The Solar Heater 19 Any material which has good conductivity and can be put in place would serve the same purpose as concrete. A better method might be to have metal fins soldered or welded on to the pipes. It is evident from table 6 that there is a good absorption of heat with concrete, for the average temperature rise was 18.8° F. With the average flow of water 1.27 lbs. per minute, the absorption would be 23.96 B.t.u. per minute or 2.33 B.t.u. per square foot per minute. Table 6 shows average values for three similar tests on three different days. The absorption was very nearly the same on the different days. Figure 12 shows a typical time and temperature curve for this series. It will be noted that the heat absorption was very nearly constant from 10 a.m. to 2 p.m. 150 ^130 no $90 % %70 I Water flow- 1.27 pounds per minute. Absorber - g/ass covered, coils embedded in concrete, painted black, stationary , 90' to noon sun. 1 square meter ^s^Average water in ftbfer in^ ■* -— ^ ~_ Water temperature^ rise 8 9 AM 10 II 4 P.M. 20 ^ 10 § o i$ ib i a 3 Time of day Fig. 12. — Temperatures of the absorber and water during the test of a plain glass-covered absorber, fitted with iron pipe water coil embedded in concrete. It should be pointed out that the absorber temperature and the outgoing water temperature were almost the same. A good rate of absorption of heat by the concrete and pipes is therefore apparent. The absorber temperature also is relatively low; evidently the heat radiated to the atmosphere is less than in the first series. The tempera- ture of the outgoing water also held up well into the late afternoon. The minimum water temperature rise during the day was 4° F at 8 a.m. ; the maximum was 28° F at 12 m. and 2 p.m. The average was 19° F. The orifice temperature varied from 62° to 94° F. The absorber temperature maximum was 118° F at 1 p.m. and the minimum was 85° F at 8 a.m. The mean temperature difference between absorber and water was 7° F. The maximum was 9° F at 8 a.m. and minimum 4° F at 5 p.m. 20 University of California — Experiment Station The conclusion seems justified that embedding of the coils in con- crete improves heat transfer to the coils, decreases the temperature of the absorber, reduces the heat lost to the outside of the absorber, and results in transferring to the water a greater percentage of the solar energy. It offers a method which is inexpensive and can be easily installed by any worker. It improves the efficiency of the heater, and at the same time provides a system which is adapted to pressures. An Uncovered Absorber with Iron Coil Embedded in Concrete. — The object of this series was to determine the heat-absorbing charac- teristics of the absorber without a glass covering, but with pipes embedded in concrete. The apparatus and procedure were the same as in the preceding series except that the glass covering was removed. The supposition in this case is that the sunlight striking the pipe surface will be absorbed as heat ; the sunlight striking the concrete in which the coils are embedded will also be absorbed as heat, which will then be transferred more readily to the pipes than if they were merely in contact with air. Such reasoning indicates that on account of this better conduction to the pipes, a larger percentage of heat will be absorbed and utilized as useful heat than if such a conducting medium were not used. Concrete is not so good a conductor of heat as metals, and probably a higher efficiency would be obtained if the pipes were fitted with metal fins or if they were replaced altogether by a shallow tank upon which the sun's rays could strike directly. Evidently some heat is radiated out to the atmosphere from both the concrete and the coils, but the amount carried to the pipes must be greater than that radiated to the atmosphere, on account of the rela- tively good conductivity of the concrete as compared to air. Table 7 and figure 13 show results of this test. TABLE 7 Teist of Absorber, Without Glass But With Coils Embedded in Concrete: Temperature, ° F B.t.u. absorbed per minute B.t.u. Test Atmo- sphere Absorber Water absorbed per minute per square foot In Out Rise of absorber D-l 77.6 78.9 69.8 78.2 88.9 88.0 77.7 88.4 83.2 81.6 77.7 82.4 99.0 96.8 89 9 97 9 15 8 15 2 12 2 15 5 D-2 D-3* Average... 19.70 1.83 Water flow was at the rate of 1.27 pounds per minute. Tests started at 8 a.m. and ended at 5 p.m. * Not averaged. Bul. 469] The Solar Heater 21 The heat absorption was satisfactory, with an average of 19.7 B.t.u. per minute for the entire absorber, or 1.83 B.t.u. per square foot per minute. The average daily temperature rise was 15.5° F, with a maximum daily average of 15.8° P and a minimum of 15.2° P. The maximum temperature rise during the day was 22° F, which occurred at noon. While this test series is not strictly comparable with series B, the data indicate that the heat absorption is greater with the pipes embedded in concrete than when they are exposed to 150 %/30 1 1 1 1 1 /Voter flow - I.Z7 pounds per m/nufe. Absorber - g/ass removed, coils embedded in concrete, pointed - black, stationary, 90' to sun of noon. Area- J square meter. 12 I a Time of day Fig. 13. — Temperatures of the absorber and water during the test of a plain absorber, without glass covering, but fitted with iron pipe water coil embedded in concrete. air. The relative average heat absorption was 19.70 B.t.u. for the former as against only 11.5 B.t.u. for the latter. The data in test D-3 (table 7) are not averaged, for they were taken on a cool day, when a strong wind was blowing directly against the absorber surface. They are shown only for contrast. The effect of weather conditions was very noticeable on these days, for the average temperature rise of the water dropped from 15.5° to 12.2° P. This is as would be expected, because the atmosphere came into direct contact with the absorbing surface. Figure 13 shows temperatures of the various parts of the absorber throughout the day. Figures 14, 15, and 16, show a comparison of the general tempera- ture characteristics of the four types of absorbers studied. 22 University of California — Experiment Station %/JO I &90 70 «*<&- L. & f^ ^fZ^ e X Oe