Preventive maintenance analysis of high temperature heat transfer fluids can play a major role in maintaining production rates, minimizing unscheduled fluid-related downtime, predicting and defining mechanical component failure, and reducing overall heat transfer fluid costs. High temperature heat transfer fluids are defined as having a useful bulk fluid operating temperature of approximately 150 to 750ºF (66 to 399ºC). These fluids include aromatics, petroleum-based fluids, polyglycols and silicones.

Proper preventive maintenance analysis starts by taking a fluid sample at predetermined time intervals. This time interval may be determined based on past experience with the system. However, in the case of no experience or a new system startup, samples should be taken quarterly. The sampling time interval then can be adjusted based on an interpretation of the data with the first sample serving as a baseline. Samples should be taken from the discharge side of the main heat transfer system pump or from any main line flowing in the system to ensure a sample representative of the fluid in the system. If not already present, provisions should be made for safe sampling, including adequate cooling of the sample. The best arrangement is a sampling bomb that allows sample flow through, sample isolation and cooling. This procedure prevents accumulation of solids in stagnant sample lines.

Once a sample is taken, it generally is sent to the heat transfer fluid supplier for analysis. Many suppliers offer this analysis free of charge with a turnaround time of one to six weeks. The data is then transmitted to the user. In some cases, the interpretation is left to the user. However, often the heat transfer fluid supplier will interpret the sample results. In cases where the sample analysis is outside the used fluid limit determined by experience, the heat transfer fluid supplier will provide data interpretation and heat transfer system recommendations. In other cases where the fluid sample results are within the used fluid limits, the fluid is satisfactory for continued use. System problems occurring after successful startup, when the heat transfer fluid is within the used fluid limits, generally are operational in nature and not related to the heat transfer fluid.

What to Check?
Prior to discussing specific tests indicative of system performance, a review of tests commonly run on heat transfer fluids is useful. In addition, data interpretation guidelines for each test are provided. The commonly performed tests are:

Specific Gravity
This test provides a measure of fluid density relative to gravity and water. Each heat transfer fluid has a specific gravity range over a temperature use range. Any variance from this range indicates the presence of either high/low boilers (degradation products) or fluid contamination with a substance of different specific gravity.

Moisture 
This test detects the presence of water in the fluid. Water has very low solubility in most high temperature heat transfer fluids (with the exception of glycol). The presence of free water in a heat transfer system can cause volatility problems such as two-phase flow, pump cavitation and excessive pressurization, especially during system startups.

Total Acid Number (Neutralization Number) 
This test is an acid/base titration detects minute amounts of strong and weak acids in the fluid. Acids usually are formed in heat transfer fluid from contamination from material outside the system. Acids usually are formed in heat transfer fluid from contamination from material outside the system. Acid number increase usually is associated with open vent expansion tank operation, heat transfer fluid oxidation or heat transfer fluid degradation, producing weak acids. Process material containing oxidizing agents can contribute strong or weak acids.

Acids are harmful in two ways. First, acids tend to accelerate the molecular breakdown of the heat transfer fluid. Secondly, they tend to form insoluble solids that accelerate mechanical deterioration of seals, valves, pumps, etc. Most heat transfer fluids have an initial total acid number at 0.00 to 0.10. The maximum value in used fluid should not exceed 0.50

Insolubles 
This test measures the amount of inorganic (pipe slag, sand and other construction debris) and "hard" carbon (coke) carried by the fluid. High amounts of carbon indicate thermal degradation of the fluid and a probable coking/sludge problem on the heat exchanger/heater surfaces or fluid oxidation. Coke and sludge can adversely affect a system's heat transfer efficiency by heat transfer surface fouling. An insolubles level of 50mg solids /100ml fluids can sometimes indicate problems.

High and Low Boilers 
High and low boilers are formed when heat transfer fluids are heated to a high temperature and certain molecular bonds begin to break or thermally degrade. Some of the new materials that form have a lower molecular weight and typically a lower boiling point than the original fluid: These are low boilers. Other compounds resulting from thermal degradation will polymerize into higher molecular weight and higher boiling point molecules than the original fluid: These are high boilers. High and low boilers seen in components may not have the heat transfer efficiency and thermal stability of the original heat transfer fluid molecules.

This measure of the charge in chemical composition to determine the presence of high and low boilers can be determined by two tests performed  for cross reference. The first is an atmospheric or vacuum distillation; the second is a simulated distillation using gas chromatography (GC). The atmospheric or vacuum distillation test accurately detects high boilers, and in this test, the fluid is completely distilled. By comparison, the GC completes the run at a specific final temperature. GC analysis is extremely accurate detecting low boilers, especially aromatics, up to the initial boiling temperature of the fluid.

Viscosity 
This test measures certain fluid flow characteristics per unit time. It can be used as an indicator of thermal degradation since low boilers will tend to reduce viscosity while high boilers will increase it. Viscosity changes also affect the overall heat transfer capabilities of the fluid.

GC Scan 
This test gives a signature of the components of heat transfer fluids, degradation products (high\low boilers) and often can detect contamination. This information is cross-referenced with the atmospheric boiling range test to confirm levels of high and low boilers as well as the presence or absence of outside contaminants.

Flashpoint (Cleveland Open Cup) 
This test provides a means of determining the fire/flashpoint of a fluid. Low boilers created as a result of thermal degradation will lower of the flashpoint. Outside contaminants may have a similar effect. As such, a low flashpoint serves to indicated the presence of thermal degradation products/outside contaminants.

Putting the tests to use 
Preventive maintenance analysis of heat transfer fluids can aid in maintaining high level heat transfer system performance at minimum cost through proper interpretation of the sample data. Some of the common tests serve as primary indicators of fluid condition while others serve to confirm these primary indicators. The primary tests are:

1. Moisture [ASTM D1744 (Karl Fishner)]
2. Percentage of high/low boilers (ASTM D86 or ASTM D1160 and ASTM D2887)
3. Insolubles [ASTM 893 (Modifies)]

Each of these tests defines a major area of fluid quality and system performance. The secondary tests - specific gravity, acid number, viscosity and flashpoint - then can be used to confirm the interpretations reached with the data generated by the primary tests.

The moisture content, as defined by ASTM D1744 (Karl Fisher), generally is a primary indicator of the integrity of a water/fluid interface. Two examples of this interface would be water-cooled heat exchangers or reactor jackets in which water is a component of the product being processed. Although some moisture from humid air may be drawn into the fluid through improperly blanketed expansion tanks in liquid-phase systems or during the cooling phase of sample collection, moisture levels in excess of 500 ppm (0.05%) or the presence of free water in a sample generally indicate a failure at a water/fluid interface.

Initially, such failures usually are not dramatic. Generally, the leak is of the "pinhole" variety, and in fact, depending on the type of fluid in use, system pressures may not be noticeably affected at the onset. Sample frequency should be increased to confirm and monitor the situation, and plans should be made to examine or repair the appropriate heat exchange surface at the next scheduled shutdown. Should moisture levels in subsequent samples increase, or the system begin to excessively pressurize, more immediate action may be required.

High/low boiler content as defined by ASTM D86 or ASTM D1160 and ASTM D2887 serve as an indicator of heat transfer fluid thermal degradation or outside contamination. ASTM D86 and D1160 are initial- and final-boiling point determination distillations. ASTM D1160, a vacuum distillation, is used when the boiling points are too elevated to make atmospheric distillation practical (d86) or the product being tested thermally degrades at its atmospheric boiling point/range. ASTM D2887 is a gas chromatography-simulated distillation, which also defines the initial and final boiling points of products being tested. In each case, once the initial and final boiling points of virgin products are known, used fluid may be tested to determine if high/low boilers are present.

Heat transfer fluids gradually degrade during normal use with the rate of thermal degradation increasing as the fluid approaches the high end of its bulk fluid operating temperature. Experience indicates that under normal conditions, most fluids with 5% low boilers or 10% high boiler levels call for reprocessing or replacement to maintain heat transfer capabilities. In the case of a dramatic increase in high or low boiler levels from the previous sample, and with the fluid still operating within its maximum bulk operating temperatures, either outside contamination, heater malfunction or decreased fluid flow rate may be assumed to be the cause.

The presence of other outside contaminants can be determined by an examination of the ASTM D2887 GC trace of the sample and a comparison of that with the virgin product trace. One caution: With petroleum-based fluids, the contaminant, depending on molecular weight, may sometimes be masked within the characteristic bell-shaped petroleum CG trace. However, the contaminant usually will thermally degrade, and its components may be seen as high or low boilers. In either case, the product being processed by the system should be the first candidate as the outside contaminant. If the identity of the contaminant is confirmed to be product under process, a leaking product/heat transfer fluid interface should be suspected. The solution is similar to that of high moisture: Increase sample frequency to monitor the situation while making provision for repair. In some cases, the nature of the contaminant or the size of the leak may require immediate action.

A rapid increase in high/low boiler also may be indicative of heater malfunction. Either flame impingement on a tube and/or excessive tube carboning can cause abnormally high film temperatures, resulting in increased thermal degradation. A low fluid flow rate due to circulation pump wear or a restriction to flow in the system may contribute to the problem, especially in systems operating at maximum bulk fluid/film temperature. Again, increased sampling is warranted, the process pump flow rate and heater should be checked and provision made for heater tube wall examination during the next scheduled shutdown.

Insolubles, as defined by ASTM D893 (modified) primarily measures pipe slag, metal fines, hard carbon and other inorganic contaminants. Insoluble solids measurements in excess of 50mg solids/100ml fluids, and especially in excess of 250mg solids/100ml fluids of particles above 10pm, call for installation of slip-stream filtration in the system where a modest pressure drop is available to filter a low flow slip stream (less than 1% of the main flow rate). This will minimize wear on mechanical components such as pumps and valves. The presence of metal fines may indicate process pump, mechanical seal, or valve wear. Pipe slag, bits of welding beads, sand, etc. are commonly found in new systems after systems after startup or in older systems that have not been filtered.

The presence of hard carbon may confirm heater tube fouling, degraded process material for contamination and
often, insoluble fluid oxidation products. Of note, the presence of hard carbon also may indicate heat exchange surface fouling that will decrease production capabilities. A simple ashing test or solvent-washing and vacuum-drying solids collected from the system will determine if the solids are inorganic (mill scale, sand, etc.) or organic (fluid oxidation products or carbon from over heating) in origin. High ash content indicates the solids are inorganic in nature while a low ash content indicates the origin of the solids was from an organic material. These changes in solids content may lead to heat exchange surface fouling and result in increases in production times of only a few seconds per day. However, over the course of years, a significant reduction of production capability may occur almost unnoticed. As a result, if excessive hard carbon is present, heat exchange surfaces should be checked periodically to determine if fouling and its resultant production loss is present.

Preventive maintenance analysis of high temperature heat transfer fluids greatly reduces the unpredictability of heat transfer system performance. Proper interpretation of the data allows early identification of potential system problems. Planned maintenance based on the used fluid analysis reduces unscheduled system downtime caused by system upsets. In addition to these cost reductions, fluid costs are reduced since fluid changeouts or reprocessing can occur on a scheduled basis, maximizing fluid life and heat transfer capabilities.