| | According to the technical literature the readings of flow meters are rounded
up to a liter or cubic meter. Withot quoting the criteria, the suggested time period of taking readings tretches from 5 minutes to a mounth, or even from 10 days to 18 mounths. In order to define relationship between the read error and the required minimum period of data collection, experimental rig consisting of a pipe system and tree flow meters having nominal diameter of 20 mm has been set up in the laboratory of the Faculty of Civil Engineering, Subotica. Flow meter No. 1 measured the inflow discharge while flow meters No. 2 and 3 were installed on the outlet. Using water balance method, the variation of error in water balance for Qmin=0,06 m3/h1<1,5 m3/h and Q2>Qmin, while Q3 has taken velues higher and lower than Qmin, have been investigated in four series with 30 measurements in each. For a period of flow meter reading equal to 60 s (rounding up the readings to 0,1 dl), discharges Q1, Q2 and Q3 were determined in each experiment. These walues have been fixed and then only the period of reading off has been varied. Following rounding up the values, the errors of water balancing have been calculated. The minimum and maximum errors of water balancing were compared to the acceptable errors (determined based on the declared accuracy of the flow meters), the minimum time periods corresponding to the acceptable errors have been determined, resulting 7 minutes in case of roundig up to a liter and 4 days in case of rounding up to a cubic meter. The results are applicable if the discharges are all the time above Qmin. The results confirm suggested values of the technical literature excluding cases of rounding up to a liter and the period of taking readings is less than 7 minutes. |
