Heat-Transfer Oil Degradation: The Four Stages You Can Read

Heat-transfer oil is the quiet fluid in a plant. It moves hundreds of degrees through a system and asks for little. Then one day the filter clogs and the tube wall runs hot.

Glacier Coolant reads degradation as a sequence, not a single event. There are four stages, and each one leaves a signal.

The standard that draws the line

Parameter	Retirement limit
Kinematic viscosity at 40°C	50.00 mm²/s
Acid number	1.50 mgKOH/g
Water content	500 mg/kg
Closed flash point	Below 60.0°C
Carbon residue	1.50%

These are the numbers to track. They turn “the oil looks bad” into a measured decision.

Stage one: induction

Degradation starts before anyone sees it. Scale on the heater tube, low flow, or a poor layout creates a local hot spot. The oil molecule breaks at that hot point and makes light, low-boiling fractions.

The flash point is the early tell. Fresh oil often starts above 180°C and can fall toward the 60°C retirement line.

Stage two: growth

Heat and trapped air now act together. Oxidation and polymerisation run side by side. The molecule both breaks and cross-links.

Large gums and asphaltenes form. Carbon residue rises. At first they stay dissolved or finely dispersed, so nothing looks wrong.

Stage three: deposition

The gums and asphaltenes condense and drop out of the oil as carbide particles. The corrosion products on the hot metal wall hand them a nucleation site.

Published degradation studies have reported sediment iron as high as 878 mg/kg, with carbon making up 85 to 89 percent of the particle mass. Black irregular grains, high carbon and clear metal-corrosion clues confirm each other.

Stage	What forms	Where it stays
Induction	Light fractions	In solution
Growth	Gums, asphaltenes	Dissolved, hidden
Deposition	Carbide grains	On walls and in flow
Manifest	Black deposits	Visible to the eye

Stage four: manifest

Only now does the eye see it. The oil darkens. Dark brown to black deposits appear through the system. By then the cost is already set.

This is the stage most managers react to. It is also the latest one that still helps.

Stop the first push at the fluid

The easiest win is to slow stage one. Glacier Coolant offers LM-15 and LM-16, built from advanced alkane modification for water-free, wide-temperature duty.

They resist molecular break and cross-link under heat. Their low viscosity supports flow and helps lower local wall temperature. Their rust control on carbon steel, stainless and copper cuts the nucleation sites that stage three needs.

Read it before it costs you

Sample on a schedule. Watch viscosity, acid number, water and flash point against GB/T 24747. When two of the five move, you are past induction.

Changing the oil earlier is cheaper than cleaning a fouled heater.

Stage one: light oxidation

Every heat transfer oil breathes a little. At high temperature, the fluid meets trace oxygen and begins to oxidize. In stage one, the change is small. The oil still flows and still transfers heat close to design.

The signs are subtle: a slight darkening, a small rise in acid value, a hint of sludge precursor. Most plants never notice stage one because nothing fails yet. That is exactly why it matters. Catching oil in stage one is cheap. Catching it in stage four is not.

Good system design slows stage one. A closed expansion tank with a blanketing gas keeps oxygen out. Lower film temperature keeps the reaction rate down. The oil’s own oxidation resistance sets the floor.

Stage two: acid build and sludge

Left in service, light oxidation deepens. Acid value climbs. Insoluble matter starts to form. The oil begins to lay down a thin film on hot surfaces.

This is the stage where maintenance notices. Filters load faster. Heaters show small efficiency loss. The acid begins to act on seals and on the system metal. None of it is dramatic yet, but the slope is set.

Sampling tells the truth here. A rising acid number and a falling clarity are the two early flags worth watching every time.

Stage three: viscosity climb and film

In stage three, the oil thickens. Polymerization products raise the viscosity. The fluid moves more slowly and transfers heat less readily. The system compensates by running hotter, which speeds the damage.

A heavier film builds on heater surfaces. That film insulates the heater from the oil, so the metal runs hotter still. The loop enters a self-worsening cycle that no top-up can fix.

Stage	Main change	Early flag
One	Light oxidation	Slight darkening
Two	Acid and sludge	Filter load up
Three	Viscosity climb	Heater loss
Four	Coking	Hot spots

Stage four: coking

Stage four is the one everyone fears. The fluid breaks down into hard carbon on the hottest surfaces. Flow drops. Hot spots appear. In the worst case, a heater burns out or a line blocks.

Coking is rarely sudden. It is stage three left to run. The lesson is blunt: watch the early stages, because by the time coke forms, the cost is already large.

Monitoring without a lab on site

Not every plant keeps a lab next to the boiler. Good news: the early stages show themselves to a careful eye and a simple kit long before a full test is needed.

Color is the first signal. A clear oil that darkens is telling you stage one is underway. A sample that turns cloudy is hinting at stage two. Neither needs a spectrometer to notice.

A basic acid strip and a viscosity check catch stage two and three at the bench. The point is to watch direction, not to chase precision. A rising trend in a simple check is worth more than one perfect number taken too late.

Pair the simple checks with a yearly send-out to a proper lab. The lab confirms the trend; the eye raises the alarm. Together they keep the oil from reaching stage four unnoticed.

A simple logbook that catches the stages

The best monitoring tool is a logbook everyone actually uses. Write the date, the running hours and the three simple readings: color, acid strip, viscosity feel. You do not need a lab to start the record.

Review the log each month. The eye catches a trend long before a single test would. When two readings move the same way, that is your signal to send a sample to the lab for confirmation.

A logbook also protects the next fill. When oil is changed, the old log shows how far it got and why. The new oil starts with a known history instead of a guess. Small discipline, repeated, is what keeps heat transfer oil out of stage four.

FAQ

What are the stages of thermal oil degradation?

Induction, growth, deposition and manifest, from hidden molecular change to visible black deposits.

When should heat transfer oil be replaced?

When it crosses GB/T 24747 retirement limits for viscosity, acid, water, flash point or carbon residue.

What does GB/T 24747 specify?

It sets in-service limits for organic heat carriers, including a 60°C closed flash point retirement line.

Why does flash point fall first?

Local overheating breaks molecules into light fractions, so the flash point drops early in stage one.

What is stage three deposition?

Gums condense into carbide grains that nucleate on wall corrosion products.

How do LM-15 and LM-16 help?

They resist break and cross-link under heat and control rust, slowing the path to deposition.

How often should I sample?

On a fixed schedule, and always after a period of high or uneven load.

Can fouling be reversed?

Often not cheaply. Early oil change and pre-cleaning beat a fouled heater every time.

Can stage one be reversed?

Not by chemistry, but it can be slowed hard by sealing the system and blanketing the tank. The goal is to keep oil in stage one as long as possible.

What is the cheapest stage to act on?

Stage one. A small change in operation or a timely partial change prevents the costly later stages.

Does coking mean the system is ruined?

Not always, but it means a clean-out and a fluid change at minimum. Prevention is far cheaper than recovery.

Can I judge the stage by color alone?

Color is a good early flag, not proof. Use it to decide when to run a proper test, not as the final verdict.

What sampling interval makes sense?

Monthly simple checks for hot loops, with a full lab test at least yearly. Busier loops deserve more frequent attention.

What is the minimum I should log?

Date, running hours and the three simple readings: color, acid strip and viscosity. That is enough to see a trend.

Who should own the logbook?

The operator who runs the loop. If it lives with them, it gets filled; if it lives in an office, it gets forgotten.