16 N. H. Agri. Experiment Station [Bui. 244 



erage of six sets of readings. Those from Fami No. 4 were not included 

 in these averages, because the refrigerator is not the conventional 

 household type. 



The curve for 1928 is valuable only as corroborative evidence. The 

 readings cover the starting period in the spring when the variations are 

 most erratic. In June the test meters were removed. 



Although the real refrigeration season does not start until March or 

 April, there was an increased consumption in February which declined 

 to the starting level in JNIarcJi. This repeated tendency may be ex- 

 plained by severe weather and poor roads — a condition which stimulates 

 buying and storing foods in larger quantities. Also, at this time homes 

 are often overheated to counteract the humid weather. 



The operation of two machines on the full year basis held the curve 

 away from zero at both the starting and closing points. The bulk of 

 the load came in a nine months period. The heaviest consumption oc- 

 curred between June and the middle of October. The peak load came 

 in July or August and varied between 60 and 75 kwhs. per month. The 

 average of 40 kwhs. a month .is taken from the records of the year 1927. 

 Since three machines were operated on a nine months basis, two for a 

 full year, and one for only five months, to find the average yearly con- 

 sumjition of current one must multiply the average per month by about 

 ten, which is the average number of months the machines were in oper- 

 ation, and not by twelve. 



Due to favorable weather conditions, the declining curve in the fall 

 tapers off smoothly and gradually in contrast to the irregular rising 

 spring curve. 



The current consumption curve on Farm No. 4 is beneath the curves 

 for the five other farms and shows the distinct saving which results 

 from the use of the combination method. The maximum consumption 

 of 21 k\\hs. for No. 4 was about equal to the minimum of the other 

 types. The maximum for the other types is 75 kwhs. 



A comparison of averages shows that this method can operate 

 throughout the year on 30 percent of the current required for the other 

 methods. The initial investment, too, is less. 



Farms Operating on Short Season 



Figure 9 shows the load cur^-e for the four farms operating for the 

 nine months between April and December. Farms No. 2 and No. 5 

 were the maximum and minimum power consumers respectively. The 

 peak load, averaged for the four farms, is approximately the same as in 

 Figure 8. Other tendencies are also much the same. 



Power Consumption for Full Season 



The curves in Figure 10 are for the two machines operating for the 

 full year and, as might be expected, they show a more constant amount 

 of current was consumed which is represented by a flatter, more even 

 curve. It is particularly interesting to note the difference in these two 

 machines. Both compressor and cabinet were located in the kitchen of 

 a brick dwelling on Farm No. 1. The room temperature remained con- 

 stant throughout the year which largely accounts for the flat load curve. 



