Pressure. Water at 300 °C requires a system operating at roughly 85 bar, with the pressure vessel, certification, water treatment and blowdown that implies. A thermal fluid reaches the same temperature at close to atmospheric pressure, so the plant is simpler, lighter and less heavily regulated, and there is no condensate system to maintain. It also gives precise temperature control and avoids freezing problems in cold climates.

By the maximum bulk and film temperature the system will reach. Mineral thermal oils are economical and cover duty up to around 320 °C. Synthetic aromatic fluids carry higher ceilings, to roughly 400 °C, at greater cost. Choosing a mineral oil for a system that exceeds its film temperature rating guarantees rapid degradation and coking, so compare against the heater's film temperature rather than the process set point.

Chiefly thermal cracking at the heater tube surface, where film temperature is highest, and oxidation wherever hot fluid meets air, typically in a poorly managed expansion tank. Cracking produces light ends that lower the flash point and heavy ends that form carbon. Overfiring, low flow, fouled tubes and running above the rated film temperature all accelerate it, and each of those conditions tends to worsen once degradation has begun.

Because it means the fluid now contains volatile light ends produced by cracking, and the temperature at which its vapour can ignite has dropped, potentially toward or below the system's operating temperature. Combined with the fact that thermal fluid systems are hot and often have lagged pipework, this is a genuine fire risk. A falling flash point in analysis is a signal to investigate the cause promptly, not a routine trend to note.

It insulates. A carbon layer on the inside of a heater tube reduces heat transfer, so the tube wall runs hotter to pass the same duty, which raises film temperature and accelerates further cracking and deposition. The process compounds, producing localised hot spots that can distort or rupture the tube. Maintaining flow, avoiding overfiring and acting on analysis results are what keep the loop out of that cycle.

At least annually on a stable system and more often where duty is severe or a problem is suspected, testing flash point, viscosity, acid number, carbon residue or insolubles, and a distillation range or low-boiler content. Sample from a live, circulating point at operating temperature so the sample represents the fluid in the loop. Trend the results; the direction of travel gives more warning than any single figure compared with a limit.

No. Mixing fluid types can produce incompatibility, unpredictable degradation behaviour, an unknown flash point and a system that no supplier will support or analyse meaningfully. When changing fluid, drain the system fully, flush with the appropriate flushing fluid and dispose of the old charge correctly. Where a top-up is needed, use the same product, and keep records of what is in each loop so this can be verified.