GREKMIOL'SE 



When all of the pipes are under the benches t.r upon 

 the walls, a single large pipe may be used as a flow t.j 

 supply all (if the ethers in the eoil, or two or more of 

 the pipes of the same size, as the returns may be used as 

 flow pipes. These pipes can be so arranged that they 

 will each supply one or more returns, or they may con- 

 nect with a header from which all of the return pipes 

 start. Care should be taken to give all of the return 

 pipes a slight fall, and it will be best if this is only 

 enough to insure their being kept free from air. It wiil 

 be safest to give the smaller pipes a slope of one inch in 

 15 feet, but 2-inch pipes, if carefully graded and securely 

 supported at intervals of 10 feet, will give good results 

 if the fall is not more than 1 inch in";!l) feet. This is 

 often of considerable importance in long houses where 

 it is not possible to sink the heater so as to give the 

 returns a fall of 1 inch in 10 or 15 feet, as is often recom- 

 mended. It should be understood tlnit better circulation 

 can be secured when a return pipe has but a slight slope 

 if suflicient to keep it free from air, with a vertical drop 

 of the return pipe at the lower end. than when the coil 

 has a much greater fall in running from one end of the 

 house to the other, if this brings the lower end of the 

 coil down to about the level of the main return. The 

 circulation in a eoil fed by an under-bench flow will be 

 quite unsatisfactory when the lower end of the coil is 

 below the top of the heater, if it is connected at its own 

 level with the return pipes from other coils, that are 

 considerably higher, and especially if they are ted by 

 elevated flow pipes. When overhead flow pipes are used, 

 the slope of the returns will necessarily be toward the 

 heater, but when the pipes are all under the benches 

 the slope may be in either direction, and if connected at 

 the end nearest the heater it will be necessary to run a 

 return pipe of the same size as the su]i]dy pipe. b:i.-k 

 from the farther end of the house, unless there are a 

 number of houses in the range, when a main return pipe 

 can be run across the farther end of the houses, to which 

 coils can be connected. If a eoil is nunle up of two or 

 more pipes of the same size, a part of which are flows 

 and the others returns, it \vi\[ be advisable to run all of 

 these pipes down hill; although, if there are only one or 

 two flow pipes, and the lower end of the coil is con- 

 siderably above the he.ater, a good circulation can be 

 secured if the flow pipes run up hill to the farther end 

 and are brought back with a downward How. The down- 

 hill system, with a flow pipe running to the farther end 

 of the house, has two advantages, as it does away with 

 the necessity of air valves, or other openings for the es- 

 cape of air, except at one point, which should be the 

 highest in the system, and it provides for a somewhat 

 more even distribution of tlie heat, the farther end of 

 the houses being fully as warm as the end near- 

 est the boiler. Where there is a large range 

 of houses and overhead pipes are m»t de- 

 sired, the difference in temperature that ^N^ 

 can be secured at the two ends of c>^Y^^ 

 the houses will nt>t be marked if 

 the coils are connected with 

 the main ^^'i^ pipe at the 

 end nearest the boiler, 

 and are ,i<.^ined with a 

 main return pipe pass- "7 

 ing along the farther : 

 end of the houses, and 

 if the coils upon the 

 walls are carried along 

 the ends of the houses 

 to the doors. 



For all hot water 

 heating plants an ex- 

 pansion tank is neces- 

 sary (Pig. ilflO). This 

 may be made from heavy galvanized sheet-iron, or a 

 riveted boiler iron tank may be used. It should be con- 

 nected with the heating pipes, but the point of connec- 

 tion will make little difference, although when the 

 downhill system is used, if the \<\^'? leading to the ex- 

 pansion tank starts from the highest point of the sys- 

 tem it will make the tise of air valves tinnecessary. 

 The tank \w\\\ be located only slightly above the high- 

 est point of the system, but it will be best placed at 

 le.ast 10 to 15 feet higher, as the elevation of the taid-c 



GREKNlIOUSli 



693 



will lessen Ihc dailgiT uf the boiling over of the water 

 in the system, and make it possible to sc-cnre a higher 

 temperature in the water of the coils than when the 

 tank is not thus elevated. Trouble from the boiling of 

 the water in the heater is most likely to occur when 

 the flow or return pipes are too snuxll, and when the 

 Are surface in the boiler is composed of small, wrought - 

 iron pipes or drop tubes. When there is a proper ad- 

 justment between the size of the b lilcr and the raili- 

 ating surface, and tlie return connections are of snlli- 

 cient size, there will be little danger from it. 



Estimating Hot Water Hadialion. — Owin^ to the 

 great variations in temperature and the difl:erences in 

 the construction of Greenhouses ami in their exposures, 

 it is impossible to give an explicit rule regarding tin 

 amount of radiation to be required under all conditions; 

 but experience has shown that, in well-built houses, 

 any desired temperature can be secured, for various 

 minimum outside temperatures, when there is a certain 

 ratio between the amount of radiating surface and the 

 amount of exposed glass ami wall surface, supposing, of 

 course, that there is a prO[ier adjustment bi'tween the 

 size of the boiler and radiating surface, and that tbi- 

 system is so arranged as to give good results. Thus, 

 when a temperature of 40° is desired in sections 

 where the mercury does not drop below zero, it will be 

 possible to maintain a temperature of -45° inside tin- 

 Greenhouse when there is 1 square foot of radiating 

 surface to 4% square feet of glass. Under the same 

 conditions, 50° can be secured when there is 1 foot of 

 pipe to i of glass, and 55°, (iO°, 05° and 70° can be ob- 

 tained when there is, respectively, 1 square foot of pi]ie 

 to each 3'-., 3. 2U and 2 square feet of glass. For out- 

 side temperatures slightly under or above zero, there 

 sle.uld be a proiHtrtionate in<.-rease or decrease in the 

 amount of pipe used, and if the houses are poorly con- 

 structed, or in an exposed location, it will be desirable 

 to provide increased radiating surface. Under the best 

 conditions the temperatures mentioned could be ob- 

 tained with a sliglitly smaller amount of radiation, but 

 the greatest economy, so far as both coal consunipfi<ni 

 and labor are C(uicerned. will be secured when the 

 amount of radiation reciennn-niled is nsi.-d. In detei- 

 mining the amount of (-xposed glass surf;n-e, the nnm- 

 lier of square feet in the i-eef, ends and sides of tlie 

 houses should be added, and to this it is al\va\s well te 

 add one-flfth of the exposed wooden or other wall sur- 

 face, and if this sum is divided by the linmber which 

 exiiresses tlie ratio between the area 

 of glass ai}d tlie amount of radi- 

 ation, it will give the number 

 of stpiare feet of heating 

 pipe to be retpiired. The 

 unit of measurement 

 ' f wrought - iron 

 pijie is the in- 

 terior diain 



tool. Carnation house. 100 x 23 ft. 6 in., piped for hot water. 



eter. while its radiating surface is <leterniined by its out- 

 side circumference ; and, although it will vary slightly ac- 

 cording to the thickness of the pipe, it is customary to 

 estimate that 1-inch pipe will afford abotit .34-1 square 

 feet of radiating surface per linear foot, whib;- 1^4-, 1^-2-, 

 2-, 2^2- and 3-inch pipe will supply, respe^'ti^"ely, .434, 

 .497, .621, .759 and .91(3 sipiare feet of radiation for each 

 foot in length of pipe. The best results can be secured 

 only when the pipes are in straight runs. The use of 

 ells and tees should be aveitpHl whenever possible, but 



