[pullquote] A monitoring program for heat transfer fluid (HTF) often is essential, as I pointed up in a recent column, “Don’t Hobble Your Heat Transfer Fluid.” So, let’s now discuss what an HTF monitoring program might include.
First, you must take the samples from the right part of the system. You need samples from the flowing fluid in the system to get a true representation of its condition. Don’t take samples from the expansion tank; in many systems, HTF in the expansion tank is nearly stagnant.
You are looking for three sources of problems: oxidation, thermal cracking and contamination. Specific tests look at each area. The most important oxidation test is for total acid number (TAN). Thermal cracking tests include distillation range, viscosity and flash point. Contamination testing includes the Karl Fischer water test for checking for leaks into the system as well as metals tests. Particulate tests indicate both oxidation and thermal cracking.
If necessary, get an outside laboratory or the HTF supplier to run the tests. A vendor often comes back with recommendations, not just results. Nevertheless, you still must understand the tests and what they may mean. So, let’s delve into some them.
Total acid number. This is a measure of acids present (commonly determined via ASTM D664 and D974); these normally result from oxidative degradation forming carboxylic acids. Carboxylic acids are the reaction products of oxygen with the HTF at high temperature. As TAN number increases, corrosion and fouling rates rise.
A fresh HTF typically should have a TAN of 0.01–0.05. The HTF supplier should indicate the maximum TAN number before the fluid needs replacement. If you don’t have any value specified, consider at least partial fluid replacement when the TAN exceeds 1 and total fluid replacement when the TAN goes beyond 3.
Distillation range. Thermal cracking creates smaller molecules. These molecules have lower boiling points. The ASTM D2887 GC method for boiling range distribution by gas chromatography generates the equivalent of a batch distillation curve for 15 stages of distillation with a 5:1 reflux ratio. This curve will show increased content of both low-boiling material (from thermal cracking) and high-boiling material (from polymerization). Too much low-boiling material can cause vapor pressure problems and require venting from the expansion tank. High-boiling material eventually will polymerize and create deposits.
Viscosity and flash point. These are much simpler tests than the distillation one. Decreased viscosity and lower flash point both indicate low-molecular-weight (cracked) material. However, they aren’t as sensitive as the ASTM D2887 distillation test. Increases in viscosity result from polymer formation and usually correlate with high fouling rates.
Water content. The ASTM has defined more than 20 test techniques for water content based on Karl Fisher methods, including ASTM D1744 and D6304, which commonly are used. A high water level is a symptom of leaks into the system or arises due to condensation of water from the expansion tank vapor space during shutdowns. Water leaks increase corrosion rates and create a potential safety hazard due to flash vaporization during startups. Metals tests often will identify sources of water leaks. Water should be vented from the expansion tank during heating.
Metals. Metal content of the base oil used to formulate an HTF should be under 1 wppm. However, proprietary additives used by suppliers frequently contain significant amounts of metals. If corrosion occurs in the system, metals content will rise; levels under 50 wppm normally are considered acceptable.
Solids. You should dilute the sample with ASTM precipitation-grade naphtha and then filter it — typically with a 0.8-micron cellulose filter. Solids include both inorganics (from leaks and corrosion) and organic materials (from thermal cracking and polymerization). New HTF should have zero solids content. You can wash the solids gathered with various solvents to determine the breakdown between inorganic and organic.
Replacement criteria and troubleshooting indicators depend upon the specific HTF. However, you should sample and analyze often enough to establish trends for the HTF quality. More-frequent sampling also will help catch problems like leaks that can lead to rapid changes. In normal operation, check an organic-fluid HTF system annually. A problem system demands more-frequent analysis — often quarterly or even monthly.