Vol.3 No.1 2010

62/110

Research paper : Development of primary standard for hydrocarbon flow and traceability system of measurement in Japan (T. Shimada et al.)−59−Synthesiology - English edition Vol.3 No.1 (2010) is 3~300 m3/h for both lines. The facilities for kerosene and light oil are completely independent of each other, but since they have common temperature adjustment device for the test liquid, the two facilities cannot be operated simultaneously.This facility employs gravimetric method with flying start and stop using diverters. The test liquid (kerosene or light oil) that passes the flowmeter to be calibrated is run for a certain time from the diverter nozzle to the weighing tank installed above the weighing tank, the standard mass flow rate is obtained by dividing the inflow mass measured by the weighing scale by duration time, and the figure is converted to standard volume flow rate by dividing the mass flow rate by the density of the test liquid. The calibration is done by comparing these standard flow rates and the readings of the flowmeter to be calibrated.As mentioned in the previous chapter, the characteristic of the calibration method using diverters is that there is little change in flow rate during the measurement compared to switching the flow using a valve. A newly developed diverter was used[8]. This diverter shifts the diverting wing in the same direction and at the same speed as the free jet flow at the beginning of the measurement when the flow is diverted to the weighing tank and at the end of measurement when the flow is diverted to the bypass. It has been employed in the national standard (water flow) of the United States and France, as well as calibration labs (water flow) of Japan, and is becoming a world standard in the liquid calibration labs. Although the diverter could not be used directly in the hydrocarbon flow calibration facility due to safety concerns, the generation of static electricity was successfully controlled while increasing the jet flow speed, by controlling the outlet surface area of the free jet flow inside the diverter, and this in turn led to successful decrease of the uncertainty of collection time in the calibration uncertainty of the flowmeter.To prevent the vapor or droplet of the test liquid generated in the free jet flow from flowing into the measurement room, adjustments were made so the interior of the diverter would be slightly lower in pressure than the atmospheric pressure, the vapor was forcefully evacuated from the room, and the oil vapor and droplets were condensed and collected as waste oil. Due to the forced evacuation of the vapor and droplets, the calibration uncertainty in the kerosene line deteriorated, and it was found that this was a dominating source in the kerosene line for certain flow range. Compared to the kerosene line, the effects of vapor and droplets were smaller in the light oil line. To remove the large amount of bubbles caused by the diverter, multistage screen mesh was installed in the 43 m3 storage tank and the buffer tank, and the bubbles could be sufficiently removed.The flowmeter to be calibrated was set in the test line with a diameter of 50~150 mm. To create an ideal flow, a straight pipe with a diameter 100 times larger (15 m) was installed upstream of the flowmeter to be calibrated. To reduce the pulsations caused by the pump, three centrifugal pumps with equal performances were operated in parallel. We also developed a method for reducing the flow fluctuation during supplying into the weighing tank[5].In the weighing room where the weighing scales were installed, the temperature was controlled within room temperature of 20 ± 5 ºC and the humidity at 30 % or above, as countermeasures against statics throughout the year using an explosion-proof air conditioning device. These also contribute to the reduction of the uncertainty of the weighing system. By calibrating the weighing scale before the measurement by loading ten 1000 kg standard dead weights for the 10 t scale and five 200 kg standard dead weights on the 1 t scale, the effect of reproducibility of the weighing scale was minimized[7]. Also, since several sources of vibrations such as the pump were installed in the same building, sufficient anti-vibration measures were taken by devising the pile foundation, to improve the uncertainty of the weighing scale that is sensitively affected by very small vibrations.To reduce the uncertainty due to the temperature expansion Fig. 4 Diagram of the hydrocarbon high-flow calibration facility.PumpTest lineHeat exchangerHeat exchangerDiverter, weighing tank, weighing scaleStorage tank (Kerosene: 43 m3)Storage tank (Light oil: 43 m3)SVPServo PD flowmeterLight oil lineKerosene lineTable 2 Sources of uncertainty in the national standard (kerosene test line). 0.008 ~ 0.016 % (Simplified to 0.02 %)Calibration uncertainty of mass flow (relative):1)+2) + 4) + 6)0.016 ~ 0.022 % (Simplified to 0.03 %)Calibration uncertainty of volume flow (relative):1)+2) +3) + 4) + 5) + 6)0.0032 ~ 0.0042 %6) Collection time to the weighing tank0.0124 ~ 0.0146 %5) Density measurement of test liquid in flowmeter0.0054 ~ 0.0154 %(0.0030 ~ 0.0146 %)4) Mass measurement of test liquid (the effect of vapor and droplets)0.0002 %3) Effect of fluctuation in flow or density0.0008 %2) Mass change in the connecting pipe0.0028 %1) Flowmeter pulse counting timeRelative uncertaintySource of uncertaintyNote) Relative uncertainty: The components that are considered to be the cause of the sources of each uncertainty are given, among the relative amount (calibration uncertainty) obtained by dividing the uncertainty of flow rate indicated by the flowmeter by the flow value.

※このページを正しく表示するにはFlashPlayer9以上が必要です