Where and how you take a thermal fluid sample can make all the difference in what the test results reveal.
Where a sample should be taken is simple – any location where there is flow and the temperature is above 180°F. A blowdown valve on the pump suction strainer housing is a good bet since that’s where you’ll find the lowest pressure and temperature in most systems. Piping drain valves will work as long as you purge several containers worth of fluid before taking the sample. Expansion tank or thermal buffer tank drain valves are tempting as a sample location because they are (usually) cool and (mostly) accessible. Don’t do it. For a long list of reasons, it’s almost the worst place to take a sample, just above scooping it off the floor near the pump.
Taking the Hot Oil Sample
How to take a sample is not quite as simple. Why? Because improper sampling practices can actually alter the physical characteristics of the sample that will be measured.
Ideally, a sample should be taken directly into a glass sample jar so any contamination or carbon in the fluid is easy to measure. The problem with glass is that it can shatter if the sample is taken too hot (above 250°F). So if the next heater shutdown isn’t scheduled until the Phillies win the pennant, install 18-24” of ¼” copper tubing on the sample port and bend a loop or two through a bucket of water. This will knock the sample temperature down the couple hundred degrees needed to keep the glass from breaking. Or take the sample in a clean metal can with a screw top and send that in (just remember to label it with the system name and date). Do not take the hot sample in a metal “cooling” bucket and then transfer it to the sample container.
Thermal fluids usually don’t telegraph that they are about to fail. Generally the only hint of impending doom is some fall off in temperature control in the heat user which requires an increase in the heater temperature to compensate. Once the power is shut off and the fluid cools however, it’s a very different story. If the fluid turns to molasses when it cools, it makes startup very difficult. Not to mention that pulling apart piping is a very time consuming way to replace the fluid.
If you operate a thermal fluid system, it’s easy to understand why thermal fluids fail. Unlike other central energy sources (like steam boilers) thermal fluid heaters don’t require (key word REQUIRE) any routine maintenance – no chemicals to add, no blowdown tanks, no condensate traps to mess with. Once you get beyond the initial startup the heater just runs.
This absence of required maintenance activities frees up operators for the 10,000 other items that do require attention. Which is why periodic sampling of the fluid should become a required maintenance activity. Even if the sample isn’t sent out for testing, simply examining the cooled fluid will give you a hint about the condition – i.e. if it doesn’t pour out of the sample container when you tip it over, you’re in trouble. (Here’s another hint: If this happens, don’t shut off the system. Give us a call.)
“If it doesn’t pour out of the sample container when you tip it over…”
The next tip will review why where and how you sample is important.
How well (or how poorly) does lab testing reflect real-world manufacturing conditions?
In the real world, oxidation of heat transfer fluids occurs in a vented reservoir or expansion tank that for whatever reason is hot (>70°C).
The acids formed in the expansion tank subsequently circulate through the system, decomposing in the heater and producing carbon sludge.
To determine which of several tests is more representative of the real world, several brands of fluid that contain additive packages were tested with the following methods:
A modified D-2440 test running at 200°C with 15 liters/hour oxygen for 24 hours – essentially an IP-48 (Institute of Petroleum Standards) test.
The standard ASTM D-2440 test running at 100°C with 1 liter/hour oxygen for 164 hours.
Fluid Sample Prepared for Oxidation Testing
The IP-48 test trashed all of the fluids. Acid Numbers ranged from 1.9 to 3.9 mg KOH/g sample (normal upper limit is 0.4). Sludge ranged from 14 to 18 weight % (any sludge is a problem).
Prepared Sample In Heating/Oxygen Apparatus
The results of the D2440 test were more representative of what is expected from additized heat transfer fluids. Acid Numbers were 0.01 to 0.03 mgKOH/g sample and Sludge was less than 0.1 weight %. We also tested fluids that contained no additive packages using the D-2440. Acid Numbers were at least 30 mg KOH/g sample and Sludge was at least 1.5 weight %.
There are a number of accelerated aging laboratory tests that are designed to determine oxidation-inhibitor performance and longevity. Most involve bubbling pure oxygen through a heated sample that has an oxidation catalyst (usually copper wire) submerged in it. The effectiveness of the additive is determined by measuring the byproducts of degradation — sludge formation, acid-number increase and viscosity increase — at the end of the test.
Oxidation Stability Apparatus (Photo courtesy of Koehler Instrument Co., Inc.)
Older tests (such as IP 48) that utilized a high oxygen flow-rate for a short period of time have been superseded by longer duration but lower oxygen flow-rate tests (such as ASTM D2440) that have proven to be more representative of real-life oxidation conditions in lubricating oils. While the D2440 test is not completely applicable to heat transfer fluids (which are exposed to even less oxygen than lubricating oils in service,) it must suffice because there exists no specific oxidation test method for heat transfer fluids. ASTM D2440 and other newer methods are also more accurate than the older tests.
Next post: Discussion; How well (or poorly) does lab testing reflect real world conditions?
While oxidation is the #1 reason that heat transfer fluids need to be replaced, it doesn’t always follow that using a fluid with an oxidation inhibitor will prevent oxidative sludging.
Oxidation inhibitors are chemical additives designed to prevent the sludge formation, acidification, and viscosity increases that result when air and hot heat-transfer-fluid molecules meet. These additives, which can prevent such viscosity-related symptoms as slow startups and even fluid solidification at room temperature, don’t last forever though. They are sacrificial in nature and are used up steadily as the fluid is exposed to air.
Figure 1 — Oxidation can influence fluid color and consistency; but darkened color doesn’t always indicate oxidative deterioration
The rate of inhibitor depletion is increased when there is greater exposure to air (by circulating the fluid in the expansion tank for example) and higher temperature (depletion rate doubles with every 10°C increase in temperature). So problems can hit hard when these additives are completely depleted leaving the fluid unprotected and ready to begin its new life as an uninhibited fluid. And sometimes, the problems that the additives have been holding off, such as the viscosity issues mentioned above, can crop up quickly and unexpectedly.
Insulation is nothing more than a large number of air pockets that are held in place by some type of material. For high-temperature systems pumping combustible liquids, these materials may consist of mineral fibers, compressed particles (calcium silicate or perlite) or cellular glass.
When designing or maintaining a hot-oil system, one important aspect is thermal insulation; adequate insulation is a necessary evil on these systems.
Even if energy costs were zero, there would still be the need to protect company personnel (especially overzealous young process engineers and nosy visitors from headquarters) from exposure to hot pipes. Not to mention the huge ventilation fans that would be required in the heater room to keep the control panel from melting.
Steel pipes insulated for plain old hot water
Hydrocarbon-based heat transfer fluids present a unique insulating problem because fluid that leaks into the insulation can become a fire hazard. The continued exposure to high temperature inside the insulation and the limited fresh air supply combine to partially oxidize the fluid into very different material. Autoignition occurs when either—
The molecular rearrangement produces a compound that ignites at the existing temperature and oxygen level.
A sudden increase in oxygen allows ignition as is.
Either way, you’ve got a problem. This phenomenon is similar to the pile of oily rags that spontaneously ignites in the garage.
The next couple of blog postings will focus on how to minimize the fire hazard from insulation. READ PART II