EFFECTIVENESS OF ORCHARD HEATERS ROBERT A. KEPNER BULLETIN 723 MAY 1951 CALIFORNIA AGRICULTURAL EXPERIMENT STATION THE COLLEGE OF AGRICULTURE UNIVERSITY OF CALIFORNIA Z 10 Ul U 0-*- > a a • a a o i a a a a > a a a a a a a a a a a a 7 0) 200 150 100 50 1 1— Jan. 2-3, 1942 1 1 p.m.- 1 a.m. Average of 2 balloon traverses above orchard. Feb. 14- 10 p.m.- 15, 1942 12:30 a.m. J / / f f / 30 35 40 40 Temperature °F Fig. 1. Temperature profiles in orchard and above bare, plowed ground before lighting heaters, for two nights with different inversions. (Taken in and near 28-acre orchard at West Covina.) which comes in contact with the foliage. Therefore, in an orchard, air cooling takes place throughout the entire height of the tree zone, and particularly at the tree tops. The foliage has a relatively low capacity for storing heat, and therefore cools rapidly by radiation, to a few de- grees below air temperature. The soil, on the other hand, has a high heat capacity; heat stored in the lower levels of soil during the daytime is conducted to the surface during the night and in this way retards the chilling of the top layer of soil. The greater cooling of the foliage ac- counts for the lower air temperatures within the orchard as compared to those above bare ground. Within the orchard, the retarding effect of the subsoil heat causes the soil surface to be a little warmer than the air which has been cooled by the foliage. For example in the left-hand graph of figure 1, the average soil surface temperatures were probably between 29 [6] and 30° F — colder than the air immedi- ately above the bare field (since the soil was cooling this air), but warmer than the orchard air chilled by the trees. Referring again to figure 1, note that the temperature increases rapidly with elevation within the lower 50 to 75 feet, and then increases more gradually up to the 300-foot limit of the readings. If the profiles were to be continued upward sev- eral hundred feet farther to a height above the influence of surface cooling, the upper portions of the curves would slope to the left due to the natural de- crease of temperature with altitude. In all following discussions, the differ- ences between the 60-foot temperatures and the 5-foot tree-center air temperatures in the unheated plot are arbitrarily taken as a measure of the magnitude of the in- version. On this basis, the inversion indi- cated by the left-hand graph of figure 1 is about 9° F, while the right-hand curve shows about 6° F. These tests would rep- resent conditions of moderately high ceil- ing, or a little less than average inversion. Cold air drift. One of the major fac- tors affecting the response from artificial heat, particularly in an isolated orchard, is air drift or air drainage. The air chilled by contact with the ground or foliage, being heavier than warmer air, will slowly flow downhill and underrun the warmer air. The filling of ground depres- sions by cold air from neighboring slopes increases the frost hazard in low spots. Conversely, the frost hazard is less on higher slopes where air drainage prevents the accumulation of chilled air. In the so-called "mass heating" of large areas, the air-drift effect diminishes except for border orchards. Even on nearly flat terrain, and with a relatively quiet night, the air drift across an orchard may have a velocity of 1 to 2 miles per hour at a 20-foot elevation. (Velocities within the tree zone will be perhaps one third as great as those at 20 feet.) For several heating runs at River- side, the amount of heat added to the in- coming air was calculated by air layers, based upon the temperature increase and measured wind velocity for each layer. On two of these nights, a 15-acre area was heated, with an average fuel input of about 34 gallons per acre per hour, and a wind velocity of 1% to 2 miles per hour at the 20-foot elevation. On one of these two nights, when the inversion was 7.5° F, 80 per cent of the total heat avail- able from the oil being burned in the en- tire upwind half of the area was needed to heat the incoming air, with 64 per cent of the total going into the air which en- tered the orchard above the 20-foot ele- vation and only 16 per cent being used to heat air entering below 20 feet. On the other night, the inversion was larger (14° F) and less heat went into the over- head air. The totals that night were 39 per cent into the air above 20 feet, and 28 per cent into the air below 20 feet. These examples demonstrate the im- portance of air drift in determining heat- ing requirements for isolated areas, and show the improved usage of heat with the larger inversion. They also suggest the importance of border heaters, which is discussed in more detail in a section start- ing on page 27. Wind. While air drift increases the heating requirements, it is a common ob- servation that when the wind velocity becomes great enough, the lower layers of air are warmed by mixing with the warmer air above. Such a wind may greatly reduce the inversion and boost orchard temperatures above the danger point within a period of a few minutes. In the field laboratory at the Citrus Ex- periment Station at Riverside, this effect became noticeable whenever the wind ve- locity at the 20-foot elevation exceeded about 2% miles per hour— a condition which would barely rustle the tree leaves. If, however, there is little or no inversion, wind will increase heating requirements. The soil as a source of heat: For the usual atmospheric conditions during radiation frosts, the net loss of heat to [7] the cold sky by radiation is about 20 Btu* per hour, per square foot of area, or about 900,000 Btu per hour, per acre.f This loss is equivalent to the perfect utilization of the heat of combustion from 6 1 /2 gal- lons of oil per hour per acre. In an unheated area, the ground is the main source of the heat lost by radiation. As the ground surface is cooled to below the daily mean temperature, heat is con- ducted up from the warmer soil at the lower levels. Any practice, therefore, which increases the soil thermal conduc- tivity will improve the availability of the soil heat, and also increase the flow of heat into the soil during the daytime. Covering or loosening the soil surface is detrimental to heat transfer, while in- creasing the moisture content is bene- ficial. However, if a frost occurs within a few days after an irrigation, evaporative cooling from the wet soil surface may more than offset the advantage of the in- creased thermal conductivity. Typical relative temperatures at vari- ous soil depths and in the air just above the soil are indicated in table 1. The temperatures in table 1 are aver- ages for 7 nights during January and early February at Riverside, with mod- erate inversions. Each temperature was recorded electrically at 10-minute inter- vals and averaged for the period from about midnight to 3 a.m. Note that near the surface the soil under the tree is con- siderably warmer than the soil between trees. This is because the tree shields its soil from direct radiation to the sky. The air beneath the tree is nearly as cold as in the space, because it is being cooled by the leaves and because of air mixing due to drift and convection currents. At the center of the tree space the soil is con- siderably colder at the surface than at the 12-inch depth. It is this temperature dif- ference that causes heat to flow from the lower layers of soil to the surface during the night. Response is related to heater characteristics and placement in the orchard The output from heaters in an orchard is available partly as convective heat (products of combustion and air heated by contact with the stacks) and partly as radiant energy from the flame or from hot metal surfaces. Some of the factors which influence the effectiveness of the convective heat are (a) the amount of temperature inversion on a particular night, (b) the velocity and temperature of the products of combustion as they leave the stack, (c) the height above ground at which the stack gases are dis- charged, (d) the size of individual fires * One Btu (British thermal unit) is the quan- tity of heat required to raise the temperature of one pound of water one degree Fahrenheit. t Schoonover, Warren R., F. A. Brooks, and H. B. Walker. Protection of orchards against frost. Calif. Agr. Ext. Cir. Ill: 1-70. 1939. or volume of the products of combustion from each heater, (e) the wind velocity and the resultant mixing action, and (f ) the distance in from the edge of the heated area. In general, the inversion is the most important of the above factors. The smaller the inversion, the greater is the loss of convective heat above the tree tops. Since radiant energy can be trans- mitted between objects without heating the intervening air, this type of heat be- comes most important when the inversion is small and convective heat is not very effective. It is also important in border heating because with proper placing of heaters, the radiant energy received by the trees and fruit can result in their tem- perature being several degrees above that of the surrounding air. [8] Table 1 .—Relative Temperatures at Various Soil Depths and In the Air Above the Soil Unheated temperatures, ° F. in or under tree center at center of tree space Air temp. 5 ft. above ground 34.1 34.9 45.8 48.5 51.4 52.6 33.8 33.9 36.7 41.7 47.8 50.7 Air temp. 3 in. above ground Soil surface temp Soil temp., 2 in. below surface Soil temp., 6 in. below surface Soil temp., 12 in. below surface Radiant energy from heaters. In considering the effect of the radiant en- ergy which a heater develops, it is neces- sary to determine both the total radiant output, and the distribution of this output in relation to vertical angle. This was ac- complished in the laboratories at Davis Fig. 2. Laboratory set-up for determining the thermal characteristics of heaters, at Davis. s by mounting each heater about 10 feet above the floor as shown in figure 2, so that radiation readings could be made from various angles in a vertical plane all the way around the heater. From these readings it was possible to calculate for each test run the "radiant fraction" (de- fined as the amount of radiant energy put out by the heater, divided by the total energy available from perfect combus- tion of the fuel, and expressed as a per- centage). Figure 3 shows curves of total radiant fraction vs. burning rate, for several kinds of heaters. For most of these heat- ers, the radiant fraction tends to remain about constant regardless of burning rate, especially at the higher rates. Only a portion of the total radiant en- ergy from the heaters strikes the trees. The remainder goes either to the ground or to the sky. The amount which is inter- cepted by the trees depends upon the size and spacing of the trees, and the location of the heaters with respect to the trees. The energy radiated to the sky is prac- tically wasted insofar as orchard heating is concerned. The heat radiated to the ground is partially effective, since it re- sults in additional heating of the air in contact with the warmed soil. However, this heat is rather localized around the heater and much of it goes below the top levels of the soil. [9] 4 5 Burning rate, Ibs/hr. Fig. 3. This graph shows the relation between burning rate and the total radiant fraction from the heaters. Results are included for six different kinds of heaters. 4 5 Burning rate, Ibs/hr. Fig. 4. This graph shows radiant fractions transmitted from various kinds of heaters to the trees only, and also the total to the trees and ground. [10] The radiant fractions transmitted to trees of the size and spacing encountered at Riverside are shown by the group of heavy curves in the lower part of figure 4. The upper curves indicate the total energy radiated to the trees and ground combined. Table 2 summarizes the radi- ant fractions from the various kinds of heaters and the distribution of radiant energy for two sizes of trees. In general, with trees 12 feet in di- ameter and 12 feet high, the division of radiant energy between the ground, trees and sky is about equal. Larger trees inter- cept additional amounts of radiation which would otherwise be lost to the sky. Placing the heaters in the tree rows rather than in the centers of tree spaces also re- duces the loss to the sky, as indicated by the example included at the bottom of Table 1, but the distribution of energy striking the trees is not as uniform be- cause the heaters are so close to one side of the trees. The experimental coke heater (fig. 7) had the highest radiant fraction of the various kinds tested, while the Fugit was lowest. In general, the hot-stack or com- bustion-chamber type of heaters are bet- ter than other kinds of oil-burning heaters in regard to radiant energy output, be- cause of the higher metal temperatures and the larger vertical surface areas from which radiation takes place. Effect of heater placement. While the curves in figure 4 show the total amount of radiant energy received by the trees, they do not indicate the distribu- tion between individual trees located at various distances from the heater. If one considers a single Jumbo Cone heater placed in the tree row between two trees (12 ft. diam. by 12 ft. high) , 70 per cent of all the radiant energy received by the trees goes to the close sides of the two trees in the same row next to the heater. The nearest trees in the adjacent row on either side receive only 5 per cent each. Trees in the second row out receive prac- tically no radiant energy. Table 2.— Summary of Radiant Energy Outputs from Heaters Kind of of heater Burning rate, * lbs/hr. Radiant fractions, % of total fuel energy Total from heater 12' x 12' trees to ground to trees to sky 16' x 16' trees to ground to trees sky Radiant fraction at lower burning rates Heaters in centers of tree spaces (20' x 24' tree spacing) Exper. Coke.. Jumbo Cone . Return- Stack Kittle Hy-Lo230A Fugit 5.4 41.7 13.6 13.3 14.8 13.0 20.3 8.4 5.6 29.6 10.7 9.8 9.1 9.3 15.4 4.9 5.4 27.6 9.8 8.9 8.9 8.7 14.1 4.8 5.7 27.5 12.0 8.6 6.9 10.6 13.8 3.1 4.4 22.2 8.0 7.1 7.1 7.7 10.8 3.7 8.1 18.1 8.8 4.4 4.9 8.5 6.8 2.8 less slightly less less slightly more slightly more Heaters in tree row (trees 20' apart, rows 24' apart) Jumbo Cone 5.6 29.6 10.1 13.4 6.1 less * Calculations are for a particular test run at this burning rate. Radiant fractions remain about constant for burning rates above these values. [ii If the heater is placed in the center of the space between four trees, each of these four receives about 20 per cent of the total amount of radiant energy going to all trees. The next trees out receive only from 1 to 4 per cent each. Figure 5 shows the radiant energy dis- tribution for rows of Jumbo Cone heaters within a heated area. Both patterns show one heater per two trees in every row. At the left, the heaters are in the tree spaces, while at the right they are in the tree rows. The numbers below the diagrams indi- cate relative amounts of radiant energy received by all trees in any one row when all of the heaters are in operation or when only part of them are used. The most uniform distribution is ob- tained when the heaters are in the spaces and all heaters are burning. Each tree then receives equal amounts of radiation from two opposite sides. With only every other row burning in the spaces, all trees receive equal amounts, but concentrated on one side. With every other row burn- ing in the tree rows, there is nearly a 3 to 1 variation between radiant energy re- ceived by heated rows and by intermedi- ate rows. When only every fourth row of heaters is used (one heater per 8 trees) , the cold- est rows receive very little radiant energy and hence must be heated almost entirely by convection. Fortunately, the use of such a small number of heaters is usually associated with conditions of moderate or large inversion when convective heat can be of some benefit. Heaters in center of tree space 24' O O x O OOOOt O 0*0,0*0 o*o^- o x o o o o x o o o o x o o x o x o x o o o o x o o o o o o x o o x o o x o o x o o x o o x o o I 100 I 100 I 100 I 100 I 100 I Every row burning | 50 50 | 50 50 | 50 Every other row | 46 8 46 | 46 8 Every third row | 45 5 5 45 | 45 Every 4th row Heaters in trei 5 rows *TO o x O O X o o *o o O o o 6 6 o 6 o o o o 6 o 6 o 6 6 o 6 o 6 o o 6 o 6 o 6 6 o 6 o 6 o I loo I [looj [iog] [looj y_ooJ [looj | 73 1 27 [73] 27 \JT\ 27 I 71 1 14V 2 14V H 71 1 14!/ 2 14V2 [JT\ 13V2 2 13V 2 QT] 13% I, I 1 indicate heater rows in operation. Numbers are percentages. 100% is taken as the total radiant energy from a single row of heaters to all trees. Values given are for Jumbo Cone heaters at 5.6 Ibs/hr. in trees 12 feet high and 12 feet diameter. These values apply only for locations at least 3 rows in from the edges of heated area. Fig. 5. This chart shows distribution of radiant energy from rows of heaters to rows of trees, for various heater spacings. [12] Tests in a one-acre plot — how they were set up and what the results indicated In order to obtain an accurate com- parison of the effectiveness of various types of heaters based upon many tests under carefully controlled operating con- ditions, a one-acre square plot within a larger orchard at the Citrus Experiment Station was instrumented and groups of heaters were operated in it under various conditions. Heating runs were made on approximately 55 nights during three winters. These runs were all made on nights not quite cold enough to require protection, but under typical radiation frost conditions. Temperature inversions ranged from 4 to 20° F. Arrangements and kinds of heat- ers. Nine different kinds of heaters (il- lustrated in figs. 6, 7, and 9) were tested. They are as follows : 1. Hy-Lo 230A 2. Riverside Junior Louvre 3. Jumbo Cone 4. Exchange model, 7-inch stack 5. Experimental Return-Stack 6. Kittle 7. Fugit 8. Experimental coke 9. Experimental coke heater with radia- tion shields added The experimental coke heaters were tested with radiation shields added, as well as without them, in order to compare the relative merits of convective heat and radiant heat under extreme conditions. Each shield (fig. 7) consisted of two concentric cylinders of galvanized iron, 3 feet tall, open at both ends, and held above the ground about 4 inches to allow the entrance of air. The clearance be- tween the inner and outer cylinders was about 4 inches. The coke heaters were placed inside of these shields. With the shields in place, the radiant fraction to the trees was only about one per cent of the total fuel energy, because the shields confined the radiant energy and caused it to be used in heating air which circulated up within the shields. Thus when the shields were used the coke heaters became almost 100 per cent con- vective heaters, while with the shields re- moved these heaters had the highest radi- ant fraction of any heaters tested. In general, heaters in the one-acre plot were placed in the centers of the tree spaces, with one heater per two trees within the plot. Border heaters were spaced one per tree on all four sides, giving a total of about 70 heaters for the entire heated plot. A few tests were made with Fugit heaters spaced one per three trees within the plot, but still with one per tree on the upwind borders. Test set-up and procedure. Tem- peratures from thermocouples at many locations in the heated plot (fig. 8), and in a similar upwind, unheated area, were recorded automatically at 10-minute in- tervals. The usual heating period during a run was about 3 hours. Temperatures were generally recorded for a 2-hour pe- riod just before lighting heaters, in order to establish natural differences between temperatures in the two plots. For determining the responses or heat- ing effects from different kinds of heaters, the 5-foot air temperatures at the centers of 4 adjacent trees in the middle of the heated plot were averaged and compared with corresponding temperatures in the unheated check plot. In addition, vertical profiles of temperatures taken near the center of the heated plot were compared with similar temperatures in the unheated plot. During some of the runs, leaf and fruit temperatures were measured in addition to air temperatures. While these tempera- tures are the critical ones to consider in determining the amount of protection re- quired during dangerous periods, they may vary several degrees for different exposures and locations on the tree. For example, fruit exposed directly to heater [13] a Fig. 6. Types of commercially available heaters tested. Upper left to right: Hy-Lo 230A (24-inch stack); Riverside Junior Louvre (18-inch stack); Fugit (1937 model) with generator parts shown. Lower left to right: Jumbo Cone; 7-inch Exchange Model; Kittle. radiation will be warmer than the air, while fruit on the "dark" side of a tree and exposed to the sky will be colder than the air. Since heat transfer to or from the air tends to average out variations be- tween individual fruits and leaves, air temperatures were used for all heater comparisons. The total amount of fuel consumed during each run was determined by [14] Fig. 7. Experimental coke heater (center). On either side are views of the radiation shields used in studies. Without shields this heater had a high radiant fraction; with shields, practically none. weighing the heaters individually before and after the run, using a specially designed portable scale unit (fig. 9) . Only one kind of heater was tested in the main plot during any particular night. How- ever, occasional runs with the Hy-Lo 230A heaters were made during each of the several winters of the study, as a check against possible changes in results from season to season. The results obtained with this heater were remarkably consist- ent from year to year. Vertical profiles of heating re- sponse. Figures 10 and 11 show the Fig. 8. Photo of the network of wires for thermocouples and other instruments in the one-acre plot at Riverside. Photo was taken during first year of tests; the trees were much larger by the end of the five-year period during which the tests were conducted. [15] Fig. 9. Portable unit used for weighing the heaters. With this equipment, one man could weigh and record 70 heaters in about 20 minutes. Heater on scales is the Experimental Return-Stack. temperatures and response from heaters in relation to the height above the ground. Figure 10 represents a night with a rather small inversion (6° F warmer at 60 feet than at 5 feet) , while figure 11 is for a large inversion (16° F) . In each case the graph at the left indicates air tempera- tures up through the tree space and through the tree center, for both heated and unheated plots. The graph at the right shows heating effects at different elevations, as determined by the differ- ences between the heated and unheated curves. Note that during each night there was no measurable heating effect above 60 feet. In each case the air in the tree tops (as shown by the dotted curves at 10 to 12 feet elevation) was colder than air at the same height between trees. This is due to the loss of heat by radiation to the sky from the relatively large tree-top area. Near the ground the air is warmer within the trees than in the space between the trees because the soil under the trees is warmer, being protected from sky radiation. Both graphs illustrate the phe- nomenon of the coldest air being at 2 to 5 feet above the ground, as explained on page 5. Relative effectiveness of different heaters. For direct comparison of the results from the various kinds of heaters, the average 5-foot tree-center heating effects for the individual test runs were adjusted by proportion so that they would [16] 50 ~j 45 I 40 4) o 35 O i 30 X 25 20 15 10 X _ Profiles at center of 1 -acre plot. Heat input to plot = 7.2 ■ 1 6 Btu hr. Average wind at 20-ft. ~~ elevation = 1 .8 m.p.h. ; Thru center of tree spac •j _ Thru tree c enter / /, 1 / Unheated temp equivalent to s. heating temps. X Tree t 3pS ^ ^' ii 'w^*. ^Tem ps. during heating r i V ,1 \ V | \ \ Y~ pace \ A k Tree center t 1 1 28 30 32 34 36 38 40 42 Air temperature °F 46 50 52 54 2 4 6 Heating effect, F Fig. 10. Vertical profiles of temperature and heating effects at center of one-acre plot, for a night with a small inversion. Hy-Lo 230A heaters were used in the plot during this test. 60 55 50 1 40 o o> » 35 o J0 o JE 30 *5 X 25 Profiles at center of 1 -acre plot. Heat input to plot- 7.25- 10 6 Btu'hr. Average wind at 20-ft. elevation = 1 .8 m.p.h. _ Thru center of tree space — . Thn tree c enter H 1 20 15 Unheated temps. equivalent to heating temps. ""*-^^, */ Tree tops^ ^* 10 5 1 i* / »> ^Te mps. c uring I eating k i | \ \ \ Tre j spac ) \ \ \ n Tree center """"* !« I c / r 28 30 32 34 36 38 40 42 Air temperature °F 48 50 52 Heating effect, F Fig. 1 1. Vertical profiles of temperature and heating effects at center of a one-acre plot, for a night with a large inversion. Return-Stack heaters were used in the plot during this test. [17] all represent the same fuel input to the plot (50 gallons of oil per hour, or an equivalent amount of coke) . Figure 12 shows curves of heating effect in relation to temperature inversion for the different kinds of heaters tested. The experimental points representing in- dividual runs have been omitted from the graph. Each curve, however, is based upon at least 3 runs at different inver- sions. The results from 11 of the 55 nights were not used because of periods with wind or unsteady conditions during the tests. Of the 44 runs used in plotting these curves, only 4 were more than 10 per cent above or below the curves as drawn. The curves of figure 12 confirm field experience that regardless of the kind of heater, the response for a given fuel input is less for small inversions than for larger ones. When the inversion is large, there appears to be little choice as to type of heater, but at the smaller inversions there are definite differences in effectiveness. At 6° F inversion, for example, the re- sponse with Fugit heaters spaced one per 3 trees was only half as great as with coke heaters for the same heat input in the plot. Of the bowl-type heaters tested, the lazy- flame heaters were the least effective at small inversions, while the Jumbo Cone and 7-inch Exchange heaters appeared to be the best. At inversions smaller than those included on the curves, the differ- ences between heaters would undoubtedly be greater. If values of heating effect are taken from figure 12 for the various heaters at a small inversion (5° F), and values for radiant fractions to trees plus ground are taken from figure 4 for the appropriate burning rates, it can be shown that the effectiveness or response of the various heaters at this inversion was in almost direct proportion to the radiant fractions. Taking the extreme comparison of coke heaters with and without radiation shields, figure 12 indicates that even at an inversion as large as 10° F, the shielded heaters (with practically no radiant energy output) gave less than 40 4 6 8 10 12 14 16 Inversion, °F, in unhealed plot (5-ft. tree-center to 60 ft. above ground) Fig. 12. This graph shows the effect of temperature inversion, and kind of heater used, upon heating response measured at the center of the one-acre plot. [18] per cent as great a response as did the same heaters with shields removed (high radiant output). However, when the in- version was 17°, the response with the shielded heaters was 65 per cent as great as that with the unshielded heaters. It is probable that at an extreme inversion of 22 to 25° F, there would be little differ- ence in the heating response with or with- out shields. In other words, when the inversion is small, most of the convective heat (hot stack gases and heated air) rises above the tree tops and is lost. Under these con- ditions, radiant heating must be the prin- cipal means of protection. With a large inversion more of the hot gases are held down within the tree zone to supplement the effect of radiant heat. Therefore, the smaller the inversion, the more advantage there is in using heaters with high radiant fractions. These heater comparisons were made only in a one-acre plot, with responses being about half as great as might be ex- pected in mass heating. The one-acre results can, however, be applied directly to the outer portions of a larger heated area, since the response seems to be pretty much a function of the distance in from the edge regardless of the total size of the area (see page 26). Within the center portion of a large heated area there might not be as much difference between kinds of heaters, but they should still rank in the same order in regard to effectiveness. Tests in isolated orchards indicate need for special attention to border heating Five test runs were made with heated areas ranging in size from 7 to 28 acres. Of these, the two made in a 15-acre plot at the Citrus Experiment Station were by far the most complete in regard to in- strumentation and results. Since these runs were made during nights not quite cold enough to actually require protec- tion, it was possible to control the tests as desired and to install an adequate number of thermocouples in advance. The one run at Porterville and the two at West Covina were made during nights when there was general heating but not in areas adjacent to the test orchard. In each case, portable equipment was taken from Riverside and set up in the orchard just a few hours prior to the start of heat- ing, after it appeared reasonably certain that firing would be required. All of these tests were in orange or- chards with trees about 15 feet in diam- eter and 15 feet high, planted on either 22 x 22 or 20 x 24-foot centers. In each case, the heaters were placed in the tree rows with one heater per two trees within the orchard. In some cases, however, only part of the heaters were lighted. Fifteen-acre area at Riverside. During two nights in March, 1941, when minimum temperatures were about 35° F, a 15-acre square plot was heated, using all of the 45 heaters per acre within the area. During the last half of the 4-hour heating period on each night, the upwind border was increased from one heater per 2 trees to one heater per tree. On the other 3 sides, the regular spacing of one per 2 trees was maintained. Most of the heaters were equipped with 7-inch Ex- change stacks, and the average burning rate on each night was about % gallons per hour per heater. Thermocouples were placed at the centers of various trees along the middle row across the plot to obtain heating effects at various distances in from the edge of the heated area. The permanently instrumented plot used for one-acre studies served as the unheated check sta- tion for these tests. In addition to the tree-center temperatures, vertical tern- [19] perature profiles were measured in the unheated plot and at the center of the heated area. Two captive balloons, oper- ated simultaneously at these two loca- tions, were used to obtain temperatures above the height of the thermocouples mounted on poles. Figures 13 and 14 show vertical tem- perature profiles up to 300 feet for the two nights. The difference between the heated and unheated curves is shown at the right on each graph and represents the heating response. For the night with 14° F inversion (fig. 13) , the graph indi- cates practically no heating effect above 50 feet; while on the night with only 7.5° F inversion (fig. 14), there was a definite response up to 100 feet but none above that elevation. When the inversion is small, there is less heating effect within the tree zone than with larger inversions, but heating extends to a higher elevation as shown by these graphs because the hot stack gases and warmed air rise more readily and must go higher before their temperature reaches the same value as that of the surrounding air. For instance, in figure 14 the air in the heated orchard at 30 feet was warmer (and hence lighter) than the unheated air at 100 feet; thus it would continue to rise until further cool- ing could bring about a balance of densi- ties. Figure 15 shows the variation of 5-foot tree-center heating effects across the heated area. The 20-foot air drift during these tests was 1% to 2 miles per hour and was consistently from east to west (from left to right on the graphs ) . The top graph contains two curves— one for the night with 7.5° F inversion and one for 14° F. The tree-center responses throughout the heated area were 15 to 20 per cent less on the night with the smaller inversion. In both cases, the heating effects obtained in the center portion of the area were about as large as can normally be obtained in general heating practice. During each of the two runs, the heat- ing effect at the upwind edge was only about 40 per cent as great as in the center portion, gradually increasing toward the center of the plot and then decreasing again beyond the center, although on the downwind side. Obviously the increase in heating effect from the upwind edge toward the center is due at least in part to the progressive accumulation of heat as the natural air drift carries the in- coming air past more and more heaters. It might be expected that the heating effect would continue to increase in this manner nearly to the downwind edge of the plot, but as indicated in figure 15 this was not the case. The response grad- ually decreased in the last ten or twelve rows and at the downwind edge was only 60 per cent as great as in the center por- tion. When a sensitive wind vane was placed at the downwind edge about 3 feet above the ground, it indicated a definite movement of air into the orchard at this height, directly against the pre- vailing air drift. Some indication of the vertical depth of this indraft may be obtained from re- sults for a few one-acre runs when vertical temperature profiles were measured in the second tree-row from the downwind edge, as well as at the center of the plot. Composite results for 6 nights whose average inversion was 6° F show that throughout the lower 20 feet of elevation, the heating effect at the downwind edge was about 15 per cent less than at the center of the plot; while above 25 feet there was no difference between the two locations. Six other nights with an aver- age inversion of 14° F had 25 per cent less heating effect near the downwind edge in the lower 15 feet and showed no difference above 20 feet. Thus for an isolated orchard, the rising stack gases and heated air create an up- draft which draws in cold air from all sides of the plot, at least if the prevailing natural drift is small. Reduced re-radia- ation from the overhead heated gases, to the orchard near the edges of the heated area, may also contribute to the lesser [20] 50 55 Air temperature °F Heating in tree zone ^y 65 5 10 Heating effect, F 15 Fig. 13. Vertical profiles of temperature and heating effect in the center portion of a 15-acre heated area. Test was made during a night on which there was a large inversion. 300 250 i 200 C o en 4> J 150 o a I 100 50 35 Average of 5 traverses, 2:50-5:55 a.m. Fuel consumption, 33 gal. per acre per hour r Unheated temperatures, adjusted for natural differences between plots o Balloon + Fixed station Temperatures at center of heated x Balloon • Fixed station 40 45 50 Air temperature °F 60 \ ^- Heating a Heating ir v' tree zone ' ^ 65 5 10 Heating effect, °F 15 Fig. 14. Vertical profiles of temperature and heating effect in center portion of a 15-acre heated area. This test was made on a night during which there was a moderately small inversion. [21] EFFECT OF INVERSION 12.5 Plot J Plot K Plot L PlotM EFFECT OF BORDER HEATERS, 7.5 INVERSION Plot J Plot K Plot L Plot M Tree rows Fig. 15. These graphs show tree-center heating effects in the middle row, across a 15-acre heated area. Average burning rate was 0.75 gallons per hour per heater, using 45 heaters per acre. These results do not include the lighting period. response. Regardless of the cause of the reduced heating effect, it is evident that extra border heaters are needed on all sides of an isolated orchard, although in greatest numbers on the upwind side. The bottom graph in figure 15 indi- cates that when the number of heaters on the upwind border was doubled (to one heater per tree), the response increased by about 50 per cent at the edge, and by lesser amounts throughout the upwind 10 or 12 rows. No tests were run with heavier border concentrations or with extra heaters on the downwind or side borders. Seven-acre area at Porterville. On the night of December 14-15, 1940, when general heating was necessary in the Tulare district, portable equipment from Riverside was set up and operated in a 7-acre heated tract northwest of Porterville. The heated area was part of a larger unheated orchard which ex- tended upwind from the heated portion, providing a good location for an un- heated check station. In the heated area, temperatures were measured only at the center of the plot. At both the heated and unheated stations, 5-foot tree-center tem- peratures were measured in order to de- termine average heating effects, and fixed thermocouples were mounted at eleva- tions up to 28 feet on portable poles. A captive balloon was operated at the heated station. Within the plot, lazy-flame heaters were spaced one per two trees, while on all four sides of the heated area the spacing of border heaters was one per tree. Starting at 11:30 p.m. all of the border heaters and every other row of the heaters within the plot were lighted. At about 3:30 a.m. the remainder of the heaters were lighted, and heating continued until about 7:30 a.m. Figure 16 shows the heated and un- heated tree-center temperatures during the night, as well as the 60-foot tempera- tures and the 20-foot wind velocities. During the first heating period, the aver- age response was 4.1° F; with all heaters [22] burning, it was 8.8° F. The average burn- ing rate was 0.7 gallons per hour, and the temperature inversion was 13° F. The average vertical temperature pro- files obtained between 3 :00 and 6 :00 a.m. during this run are not shown. However, they were similar in shape to the ones for the Riverside run with similar inversion (fig. 13) but with temperatures 12 to 14° F colder at all heights up to the 300- foot limit of the readings. There was no indicated heating effect above 100 feet and very little above 50 feet. Twenty-eight acre area at West Covina. Two runs were made in an iso- lated orange orchard near West Covina on nights when light or moderate general firing was required. The terrain at this location is very flat, since the orchard is located on the floor of a valley. Even so, the average 20-foot wind velocity during the two nights was 1 to 1% miles per hour. Since there was no adjacent or- chard, it was necessary to locate the un- heated check station in a bare, plowed field. In addition to the regular heater spac- ing of one per 2 trees within the orchard, . 7.5 i 5o s > T? 2.5 there was a border row along the north- east side with one heater per tree. The southwest border had a few extra heaters, but the other two had only the regular spacing. On the night of January 2-3, 1942, all of the border heaters and every fourth row within the plot were burned from 1:30 a.m. to 8:00 a.m. During the night of February 14-15, the first heaters were lighted at about 1 :00 a.m., following the same pattern as for January 3. Addi- tional heaters were lighted at 4:00 a.m. and still more at 6:30 a.m., as indicated in the upper part of figure 17. After 6:30 a.m., all of the odd-numbered rows of heaters within the orchard were burning, in addition to the borders. For both nights the average heater burning rate was about 0.5 gallons per hour. The average inversion was 8° F on the first night and 5.5° F on the second night. Vertical temperature profiles (not shown) indicated no response above 100 feet on either night. The lower part of figure 17 shows, for the second night, the tree- center heating effects at various distances in from the northeast edge of the orchard. Although this was the side with a border Wind at 20-ft. elev., ( unheated station S r«" y^ 11 p.m. 12mdt. 1a.m. 2 3 4 5 6 7 8 a.m. December 14-15, 1940 Fig. 16. These curves show wind velocity and air temperatures in a 7-acre heated plot at Porter- viile, California, during the night of December 14-15, 1940. [23] 20% / 30% *r Drift (1 mph at 20' elev.) Row No. 50% N 21 rows to edge 21 %Q D 0*0 0>*D o-o o*a 0*0 0>0 D»0 o»a y»0 0*0 0*0 0*O o kDOOOOOOOOOODOOQOOOQOODODODO 19 *0 O-O O'O O'O 0«D 0*0 O'D OO OO 0»O OO OO OO D-D '[)[)0^oyoooooooooooooooooyoo 17 "O O 0*° 0*3 O»0 0*0 O'O o»0 0«0 <>0 '0»O 3>Q O O !>£> O «0 0D ODOcJODC'OS O O o O (J O D O O O O O O O 15 «0 o»D 0»£> OO OO OO 9»0 OO 0»0 0»O OO 0»O OO D«0 33 rows to edge * First lighting — every 4th row and borders — 390 heaters 14 rows to edge Second lighting — all but 4 of rest of odd rows — 540 heaters total Third lighting — rows 3, 9, 23, and 35 — 660 heaters total 5.5° F average inversion, 5 to 60 feet elevation /N.E. edge Heating effects in row 20 (1 t.c. in row 19) Center of orchard 10 15 20 Tree numbers from N.E. edge of orchard Fig. 17. This chart shows the lighting schedule (above) and the tree-center heating effects across a 28-acre heated orchard at West Covina, during a test made on the night of Feb. 14-15, 1942. [24] 10 O 2 Curve applies to response in center portion of heated areas larger than 5 to 7 acres, without regard to border heating requirements. Fuel requirements at or near borders will be greater than indicated by this curve. 2 4 6 8 10 12 Inversion, °F, in unheated plot (5-ft. tree-center to 60 ft. above ground) 14 16 Fig. 18. This shows the effect of temperature inversion on fuel requirements per degree of response, in the center portion of 7- to 28-acre heated areas. of one heater per tree, the drift entered from this direction during only about one-third of the time; usually the drift was more nearly parallel with this border. The three curves represent the three periods during which different numbers of heaters were burning. The heating effects indicated in figure 17 were measured in a row next to one in which heaters were burning during the entire time. These results indicate a nearly uniform response in this row when only every fourth row of heaters was burning. As the number of heaters was increased, the difference between the center response and that near the edge increased, perhaps because of additional indraft caused by the stronger orchard stack action with more heaters. The re- sponse in the outside row was greater, presumably because of the nearly parallel wind and the direct heating by radiation from the heavy row of border heaters, while the indraft of cold air flowed past these trees and prevented the next trees in from being warmed as much. Tree-center temperatures during both of the nights indicated that when only every fourth row of heaters was burning, trees in the three cold rows each had about 25 per cent less response (0.3° F less) than trees in a row containing lighted heaters. Relation of response to size of heated area. The effect of temperature inversion upon fuel requirements per de- gree of response at the centers of the 7-acre, 15-acre, and 28-acre orchards is shown in figure 18. The fuel inputs in gallons per hour per acre were based upon the average heater spacing within the plot, without regard to border heaters. Where the number of heaters used per acre was changed during a particular run, separate points were plotted for each period. During one of the runs at West Covina there were 3 such periods with increasing numbers of heaters in oper- ation, while at Porterville there were 2. Although the actual fuel inputs for different runs varied from 6 to 34 gallons per hour per acre, and the plots were in [25] 3 different localities with sizes ranging from 7 to 28 acres, the results fit a single curve rather closely. This indicates that in the center portion of a heated area, beyond the influence of border condi- tions, the response is relatively independ- ent of the size of the area heated. The curve indicates a considerable increase in fuel requirements as the inversion de- creases below 12 to 14° F. For example, the fuel requirements at an inversion of 4° F would be 50 per cent greater than at 10° F. In tests with the 15-acre plot (fig. 15) the reduced response due to border effects extended in 10 to 15 rows from the edge. The responses were about constant throughout the center portion of the or- chard and were about as large as are normally expected even in mass heating. Thus, although the largest heated area involved in these tests was 28 acres, it is probable that the fuel requirements indi- cated by figure 18 would apply to larger areas on reasonably flat terrain. The rule used by some growers, that they normally will need to light 8 heaters per acre to obtain a response of about 1° F, is consistent with the fuel require- ments indicated by the curve for an in- version of 10° F (assuming a burning rate of V2 gallon per hour) . The curve in figure 18, however, should not be used for areas of less than 6 to 8 acres, because the reduced response due to border effects is likely to extend in to the center of areas smaller than this limit. Regardless of the size of heated area, the fuel requirements near the edges will be greater than those indicated in figure 18 for the center portion. Conversely, if heaters are uniformly spaced, with no extras on the borders, the response near the edges of the orchard will be less than in the center. The magnitude of this re- duction of border response is influenced considerably by the total rate of fuel con- sumption per acre per hour. For example, in 3 runs with heater spacings of 11, 22, and 45 per acre, in 15-acre or 28-acre orchards, the following results were ob- tained: Total fuel rate, gallons per acre per hour Response at third tree in from upwind edge 5.5 90% as great as in center of area 11.0 70% as great as in center of area 33.7 50% as great as in center of area The greater reduction in border re- sponse with larger fuel inputs appears to be largely a result of greater inflow of cold air due to the increased updraft created by the heaters. Within a large heated area (beyond border influences) the tree-center response is limited by in- creased radiation to the sky from the warmer trees and air, and by reduced outflow of heat from the soil as its surface temperature rises (decreasing the tem- perature difference between the soil sur- face and the lower depths). Results obtained from comparable tests with the different sizes of heated areas indicate that the response at a given distance in from the upwind edge (but not beyond the center) is about the same regardless of the size of the plot. For example, with 7-inch exchange heaters and similar conditions, the tree-center heating effect at the fifth row in from the upwind side of the one-acre plot (at center of plot) was almost as great as in the 15-acre plot at the same distance in. Likewise, the response obtained near the center of the 7-acre orchard at Porter- ville (12 rows in) agrees quite well with results obtained in the 15-acre plot. Since the response appears to be a function of the distance in from the edge, the extensive results obtained in the one- acre plot regarding responses and relative effectiveness of different kinds of heaters should be directly applicable to a strip around the outside of a larger isolated orchard. The importance of border con- siderations becomes more apparent when one considers that with 90 trees per acre, a strip only 5 trees wide all around a 15- acre area, or 7 trees wide around 30 acres, represents half of the orchard. [26] From the tests, border heating recommendations can be made for isolated orchards While the tests did not include enough over the first two or three rows in and large-plot runs to actually check the best not be all concentrated on the outside, arrangement of border heaters for a uni- This will allow a greater percentage of form response, sufficient information is the tree surfaces in the outer rows to be available to serve as a basis for recom- "seen" by the heaters and thus be heated mendations in this respect. The following directly by radiation, general considerations should be kept in (c) Avoid excessive burning rates for mind in regard to border heaters: border heaters; use larger numbers of (a) High radiant output from the heaters properly distributed heaters at normal is especially important in border areas, burning rates. In the center portion of the so that the outside trees may be heated orchard, burning rates should be kept as directly, without the necessity for warm- low as possible to minimize updraft which ing all of the inflowing air before it gets draws in cold air along the borders, into the orchard. If a grower has more (d) Border heaters are usually needed than one kind of heaters, those with on all sides of an isolated heated area but highest radiant fractions should be used with greatest numbers on the upwind on the borders. side. Never use less than one heater per (b) Border heaters should be distributed two trees on any outside edge. FOR OUTSIDE AND PERHAPS FIRST ROW IN ON ALL SIDES EXCEPT UPWIND SIDE, USE TWICE AS MANY HEATERS AS USED WITHIN THE PLOT ALONG UPWIND EDGE, USE 4 TIMES AS MANY HEATERS AS USED WITHIN THE PLOT (BUT NEVER MORE THAN 2 HEATERS PER TREE) PREVAILING AIR DRIFT < $<3a (*> rtfa dot c$o> c$&>® IN THESE TWO ROWS, USE TWICE AS MANY HEATERS AS USED WITHIN THE PLOT Fig. 19. Typical example of border heater spacing, when using one heater per two trees within the orchard. Small crosses indicate heaters. [27] The actual border heating require- ments will vary considerably, depending upon the topography, the wind velocity, the intensity of the heating (gallons per hour per acre), the temperature inver- sion, and other factors. However, for the usual conditions with reasonably flat terrain and low wind velocity, the follow- ing spacings are recommended: (a) Upwind borders. On the outside of the orchard, use four times as many heaters per tree as are used within the orchard. For the first two rows in from the edge, use twice as many as are used within the orchard. Never use more than two heaters per tree on the outside nor less than one heater per two trees on the outside and the first row in. (b) Downwind and side borders. On the outside and perhaps for the first row in, use twice as many heaters per tree as are used within the orchard. Never use less than one per two trees on the outside. Examples: (1) Suppose the required response is small so that only one heater per 8 trees is needed within the orchard. Then all borders should have the minimum of one heater per two trees on the outside. In addition, the first row in from the upwind side should have the minimum of one per two trees. (2) Assume that one heater per two trees is needed within the orchard. Then on the upwind side use two heaters per tree on the outside and one heater per tree for the first two rows in. On all other sides, use one heater per tree on the outside and perhaps for the first row in. (See fig. 19.) Use of heaters with wind ma- chines. Recent tests and field experience indicate that the combination of wind machines with uniformly distributed heaters gives a response greater than the sum of the normal response from the heaters alone plus the response from the wind machine alone. In tests during two nights at the Citrus Experiment Station at Riverside during February, 1950, the combined response from a 90-horsepower wind machine, plus 15 heaters per acre, was 20 to 30 per cent greater than the sum of the indi- vidual responses.* Within the zone of disturbance, the air mixing caused by the wind machines tends to make the convective heat from the heaters more useful than when heaters alone are used. However, beyond this zone of influence, such as in the corners of a square orchard, the heaters must do the entire job, and the radiant output is fully as important as for border heaters in an orchard heated in the conventional manner. The findings and recommendations apply only indirectly to large-scale heating operations In the so-called mass heating of large areas, the fuel requirements within the heated district will probably be a little lower than indicated by figure 18. As previously discussed, these large-area fuel requirements are determined pri- marily by the heat input required to counteract the normal radiation losses, plus the additional losses resulting from the higher temperatures of the trees, air, and soil surface in the heated orchard. For large-scale heating, the border effects are confined mainly to the or- chards on the outer fringes of the heated district. However, if an orchard well * Frost Protection by Combining Wind Ma- chine with Distributed Heaters, by F. A. Brooks, D. G. Rhoades and H. B. Schultz, California Agriculture, Vol. 4, No. 9, Sept. 1950. [28] within a heated district does not have other heated- areas reasonably close, it will probably need some extra border heaters. As with borders and isolated orchards, the use of heaters with high radiant energy outputs will give somewhat better response for mass heating, particularly when the inversion is small and heating requirements are severe. And here is a summary of the findings that resulted from the orchard heating tests The principal factors which affect the amount of fuel required for a given tem- perature response are: 1. Magnitude of temperature inversion. Heating is most difficult on nights with small inversions or so-called high ceiling. 2. Inflow of cold air, due either to natural drift or to the combined stack action created by the heaters in an isolated orchard. In the heating of large areas, the effect of inflow diminishes, except for border orchards. 3. Heater characteristics, particularly in regard to radiant energy. With a small inversion the convective heat is largely wasted above the tree tops and radiant heat thus becomes the principal means of protection. In the one-acre heating studies, when the inversion was large, there was little difference between the various kinds of heaters in regard to effectiveness. When the inversion was small, however, the effectiveness was directly related to the radiant fractions of the heaters. The several kinds of heaters which were tested may be rated in the following order in regard to heating effectiveness or re- sponse for a given fuel energy input: 1. Experimental coke 2, 3. Jumbo Cone and 7-inch Exchange stack (no difference) 4. 5. Return-Stack and Kittle (no differ- ence) 6, 7. Hy-Lo 230A and Riverside Junior Louvre (no difference) 8. Fugit (one heater per two trees) 9. Fugit (one heater per three trees) Five test runs were made in isolated orchards ranging in size from 7 to 28 acres with inversions of from 5 to 14° F and with fuel inputs from 6 to 34 gallons per acre per hour. Fuel requirements for a given tree-center response within the center portions of the areas were con- sistent for all of these tests, and indicate a definite increase as the inversion be- comes smaller. In none of the tests was there any measurable response at heights greater than 100 feet above the ground. With fairly large inversions, the heating effect was confined mostly to the lower 50 feet, while with small inversions there was considerable heating between 50 and 100 feet (practically useless to the orchard). A definite orchard stack action, created by the rising of the heated air and gases from all the heaters, was observed and is indicated by the test results. Cold air apparently is drawn in from all sides of an isolated heated area, with the depth of inflow probably being as great as the tree height, even on the downwind side. The combined effect of the natural air drift and the inflow of cold air caused by this orchard stack action is to reduce the response for as much as 10 to 15 rows in from the edges. The degree of reduction of border response is greatest for high rates of fuel input per acre because of the increased orchard stack action and in- draft. Thus it is important to operate the heaters, especially those within the plot, at the lowest rates which will give ade- quate protection. [29] Border heaters should be distributed over the first two or three rows in, and not be all concentrated on the outside of the orchard. Heaters with high radiant output are especially advantageous in border heating. Border heaters are needed on all sides of an isolated orchard but with greatest numbers on the upwind side. When heaters are used in combination with wind machines, extra response is obtained because the convective heat from the heaters is made more useful by the mixing action of the wind machine. However, heaters beyond the zone of dis- turbance of the machine must do the entire job, and the radiant output is fully as important as for border heaters in an orchard heated in the conventional manner. ACKNOWLEDGMENTS A project as extensive as the one discussed in this publication involves the com- bined efforts of a considerable number of persons. * Staff members in addition to the author who were actively engaged in the field work on orchard heating during the period from 1937-1942 included the following: C. E. Barbee, L. M. K. Boelter, F. A. Brooks, C. Lorenzen, F. L. Rhinehart, and H. B. Walker. The co-operation of the Citrus Experiment Station, particularly the Purchasing Department and the Division of Cultivations, was an invaluable aid during these studies. [30] 10m-5,'51(4665)W.P. [31] test tube farming pays off for you t all of the agricultural research done by the University of California is field work. Much useful knowledge comes to light through work done under controlled laboratory conditions. This information, after thorough checking and application to field problems, becomes available to all California farmers. Distribution of this knowledge is made through: LITERATURE: Circulars, bulletins, lithoprints, and leaflets by specialists are available free. These publications cover many subjects re- lating to agriculture in the state. For a catalog of this litera- ture write to the Office of Agricultural Publications, 22 Giannini Hall, University of California, Berkeley 4. COUNTY FARM ADVISORS: Farm Advisors are agricultural specialists with a background of practical knowledge. They serve 52 counties throughout the state and their mission is to help farmers work out their problems. Get to know your Farm Advisor — take advantage of his services. MAIL INQUIRIES: If you prefer to put your questions in a letter, mail them to the Public Service Office of the College of Agriculture, Uni- versity of CaJifornia, either at Berkeley or at Davis. Your problem will be referred to the person or department best j^ able to give you the exact information you need. THE COLLEGE OF AGRICULTURE UNIVERSITY OF CALIFORNIA