DELAYS EXCOUKTERED BY TOLL OPERATORS 541 



Constructing engineering tables based on a preselected constant circuit 



delay or trunk speed of service. 

 Arrangements were made with several Associated Companies to furnish 

 data for this purpose which would show: 



1. The average overall sj^eed of service on different sizes of groups under 



various conditions of loading. 



2. The minimum average overall speed interval on these same groups at 



times when circuit {provision was not a factor, i.e., when NC condi- 

 tions were not encountered. 

 The speeds obtained in Item 2 were subtracted from those obtained in Item 

 1, the difference representing that portion of the overall speed which can be 

 attributed to circuit delay, or the trunk speed. 



In order to determine these trunk speeds it was necessary to obtain from 

 several sources as much data as possible of the following nature: 



Per cent, circuit usage, by hours, as derived from group busy timing 

 registers on selected groups of various sizes. Hours during which the 

 traffic over a group was handled subject to posted delay were dis- 

 regarded. 

 The number of originating terminal calls handled over the groups during 

 the hours corresponding to the usage data and the average speed of 

 servdce on these calls. The call and speed of service data were sum- 

 marized first to include all calls and then separately for calls not en- 

 countering NX. Correction was made for transfer of tickets to point- 

 to-point positions by subtracting from the speed shown on each such 

 ticket an inter\'al representing the average length of time required to 

 send a ticket to point-to-point positions in the office in which the data 

 were obtained, provided the transfer time was included in the overall 

 speed inter\'al. This interval of transfer time is not properly chargeable 

 as part of the trunk speed. 

 These data were obtained for trunk groups of various sizes ranging from 

 one up to eighteen trunks. To secure a comparable amount of data for the 

 smaller groups which handle fewer calls, it was necessary to include more 

 of the smaller groups or to continue the record for a longer period of time on 

 such groups. 



The data for all hours of the day or evening were useful because as the 

 volume of traffic recedes from the busy hour the data are typical of the busy 

 hour condition of other groups engineered on a more liberal basis. The 

 very light hours also show the minimum speed interv^al which can be ob- 

 tained when lack of an available circuit is not a factor. 



Five Associated Comi)anies obtained data at eleven toll offices on 112 

 intertoll groups having 561 trunks. Approximately 17,000 calls (occurring 

 during hours when the groups were at least 40% busy) were included. 



