Redesigning experimental equipment for determining peak pressure in a simulated tank car transfer line

Abstract

When liquids are transported from storage tanks to tank cars, improper order of valve openings can cause pressure surges in the transfer line. To model this phenomenon and predict the peak pressures in such a transfer line, a laboratory setup consisting of a pressurized water storage tank connected to different segments of pipe by ball valves was constructed. By varying parameters including water height within the tank, transfer line length, and applied driving pressure, the most critical variable was determined to be driving pressure. The hydrostatic pressure from the difference in water height was negligible and this fact was evident without the need for experimental verification. This setup therefore allowed for even fewer parameters to be tested. Due to the poor condition of the experiment because of age and corrosion along with the few insights the setup provided, the experiment needed to be updated. The newer version is designed to allow students to have more choice in what parameters they wish to test, with pipe segments of different length as well as different diameter with various impedances. To address spatial and practical considerations, the new design was assembled in PVC piping. This mockup proved useful in discovering inadequacies in the design that had not been considered.

(cont.) While the mockup proved that the design was safe to use at the operating pressures within the 2.672 laboratory for which the experiment is intended, it also proved that there were important factors influencing the aesthetics of the experiment that had been considered secondary to the safety. To add complexity to the problem, the design included clear segments of pipe near the ends in which the water hammer would oscillate so that digital imaging analysis could later be implemented. However, the increase in pipe length to hide the pressure tank below the table also caused the air pressure required to drive the oscillations in the clear section of pipe to be much higher than operating pressure. As this build was considered as a mockup, these problems have been noted so future designs for the final experiment to be used in the 2.672 classroom can address these problems.