A thermic fluid system's health cannot be seen from the control room: the fluid ages chemically, the symptoms lag the causes by months, and by the time performance complains, the sludge is already in the coils. Running TF well is therefore a laboratory discipline stapled to an operating routine — this lesson is both, plus the commissioning procedure that decides how a system's whole life begins.
Commissioning: boiling the water out
Every new charge — and every opened system — carries water: residual flushing, atmospheric moisture, damp insulation of tank walls. Water in hot oil is violent (lesson C5's bubble physics with 280 °C ambitions), so first heat-up is a slow, deliberate dry-out: hold below 100 °C while circulation proves itself; then creep through the 100–120 °C band in small steps, watching the expansion tank — crackling, foaming and pump-pressure flutter are water announcing itself — and holding at each step until the noise dies; then climb to operating temperature in stages, venting freely throughout (lesson E4's tank path earning its design). Rushed dry-outs knock pumps out on vapour, spit oil from vents, and in bad cases hammer coils. The patient version costs a shift; the impatient one costs the commissioning engineer's reputation and sometimes the pump seals.
The annual blood test
Once or twice a year (and after any incident), a sample goes to the laboratory — drawn hot from a circulating point, not cold from the expansion tank where light ends hide. The panel and what each number diagnoses (lesson E3's chemistry made into columns):
| Test | Watching for | A move means |
|---|---|---|
| Flash point | thermal cracking | falling = film overheating: check flow, firing, fouling (E3/E4) |
| Viscosity | both directions | down = cracking; up = oxidation/polymerisation |
| Acid number (TAN) | oxidation | rising = air ingress — the expansion tank is running hot or breathing badly |
| Carbon / insolubles | sludge in the making | rising = fouling ahead: plan a look at the coils |
| Moisture | water ingress | present = leaky cooler/user or a breathing tank pulling damp air |
Trend, don't spot-read — one result is a fact, three years of results are a diagnosis (the same logbook philosophy as lesson B7). Most fluid suppliers run the panel inexpensively; the discipline is taking the sample the same way, at the same point, on the calendar, and filing the answer where next year can find it.
The operating rhythm
Daily: heater ΔT (a drifting temperature-rise at constant firing is flow or fouling talking), pump pressure and any seal weep, expansion-tank level and temperature (E4's touch test), burner behaviour. Weekly/monthly: walk the circuit for staining and oily lagging (E4's absolute rule), verify vents clear, exercise seldom-used valves. At every service: the flow-trip and high-temperature-trip witnessed live — the two interlocks that guard the film — plus strainer inspection (a TF strainer's catch tells the fluid's story as surely as lesson C10's dirt census). When fluid finally ages out: partial replacement — bleeding a fraction of degraded fluid and topping with fresh — often restores the trend at a fraction of a full change; full replacement plus a system flush is the answer once carbon and TAN say the inventory is spent. Either way, the laboratory decides, not the calendar alone.
- Dry-out discipline on every new charge or opened system: below 100 °C, prove circulation; creep through the water band; hold until quiet; then climb. Write it as a procedure, not folklore.
- Sample hot, circulating, same point, same method, on the calendar. A cold tank sample flatters the fluid.
- Put the five test numbers on one trend sheet by the heater. The plot, not the latest report, is the instrument.
- Witness the two trips at every service. The film has no other friends.
- Commissioning is a dry-out: patient steps through 100–120 °C until the water is gone. Rushing it costs seals and nerve.
- The annual panel — flash, viscosity, TAN, carbon, moisture — names the sin: cracking vs oxidation vs ingress.
- Trends diagnose; spot readings only reassure.
- Partial fluid replacement, guided by the lab, beats both neglect and reflex full changes.
Condition monitoring by oil analysis came to industry from the railways and the air force: post-war engine programmes learnt that a teaspoon of used lubricant, trended over time, predicted failures no inspection could see, and "spectrometric oil analysis" became standard for locomotives and jets by the 1960s. Thermic fluid inherited the philosophy wholesale — the fluid is the machine's bloodstream, and the laboratory reads it — making the TF sample bottle one of the many places where plant maintenance quietly practises aviation discipline.