If you look on the attached pictures, one is a basic sketch of system with the vessel on the right being formaldehyde reactor that uses an HTF, therminol with a SG of .845 at 293 degree C, which is just below the boiling point. The smaller vessel on the left is the level gauge which uses a magnetic float and a guided wave tube. The other drawing is a graphical representation of the level in the gauge at 2 seperate startup events and the temperature of the fluid. During steady state the therminol boils but only on the surface of the coils. The current issue is that while cooled down, the level indicates 7 inches in the site glass (both the visible float and the guided tube electronic display). Upon heatup the detector has a surge of 7-8 inches, which is considered normal for this plant due to the mass aeration of the process. The level then lowers slightly and steadies off, as is to be expected, but then there is a sudden and unexplained drop in level, followed by a slow and steady increase. My current theory on the increase is that the vapor side is condensing at the top of the level gauge and causing a small drop in pressure. Based on the specific gravity of the HTF being about .824 I calculated that a DP of only .36 would change the level indicated in the gauge by 12 inches. This condensed vapor is then being drawn back into the reactor due to its relatively high temperature and causes natural circulation in the device. What I don't understand is the sudden drop in level, as well as how this is not a common enough occurance that the company selling us the gauge wouldn't have known it was not suited for our application. Please critique my theory and offer an explination for the sudden drop. Thanks everyone.