• ByRouging Solutions Editorial Team  ·  3 September 2026

Glass-Lined Reactor Cleaning: How to Protect the Lining

A glass-lined reactor is really two jobs pretending to be one vessel. The face that touches the batch is fused glass, on the wall and on most of the fittings. Stainless steel carries the valves and pipework around it. Glass and stainless want opposite things from a cleaning chemical.

That is the whole problem. A hot caustic cleaner that shifts soil fast can eat into the lining. An acid mild enough for the lining will not rebuild the passive layer on the stainless. Glass-lined reactor cleaning has to get both right at once.

The good news is that the limits are written down. Standards set how much enamel loss is allowed in acid and in caustic, and how big a temperature jump the lining has to survive. Those numbers point at one kind of chemistry.

What Makes Glass-Lined Reactor Cleaning Different?

You are cleaning two different surfaces on the same vessel, inside the same shutdown window. Nothing else in a plant works quite like that.

The lining is fused glass on carbon steel. It is there because it shrugs off acids that would eat the steel underneath. That is why these vessels turn up wherever the chemistry is harsh: pharmaceutical API work, dyes and pigments, and agrochemicals.

Most of the parts that touch the batch are enamelled too. The agitator, the baffles and the thermowells are usually glass-lined rather than bare metal. That surprises people who expect a steel paddle. The stainless steel sits further out, in the valves, the pipework and the instruments.

A multi-product plant adds another wrinkle. Duties change between campaigns, so the same vessel meets a different chemical mix every few weeks.

Our cleaning teams treat the two as separate jobs on the same vessel, because the method that suits one can damage the other.

What Does the Lining Resist, and What Attacks It?

Acid, mostly, and the lining handles it well. ISO sets limits for chemical enamels in two directions, and the two numbers sit far apart.

The 2015 edition of ISO 28721-2 set two corrosion limits, each with its own test. Resistance to "condensing hydrochloric acid vapour", measured to ISO 28706-2, had to be 0.08 mm per year or less. In sodium hydroxide solution the limit was 0.40 mm per year or less, measured to ISO 28706-4. That test sets the ratio of solution volume to exposed enamel area at 3.5 cubic centimetres per square centimetre (ISO 28721-2:2015).

Those are two different tests, so the rates are not a like-for-like race. What the gap does tell you is how much more loss the standard is willing to accept on the alkali side. It is five times higher, and ISO calls both figures "the minimum requirements a chemical enamel is expected to meet". A lining can pass the standard and still return 0.40 mm per year on that alkali test.

Those limits come with a grade. A chemical enamel is ordered against a spec, and the grade on your vessel documents is the one that governs. A general rule about glass does not. Two reactors on the same floor can carry different grades, bought years apart.

Why Is Caustic the Bigger Risk?

Because that is the side where the 2015 standard allowed more loss, and because caustic cleaning is usually hot. Heat and alkali work on glass together, and the loss is permanent. Enamel does not grow back the way a passive layer on steel does.

The damage rarely comes from one dramatic event. It builds from routine. A caustic boil-out at every changeover, a hot alkaline CIP cycle on a shared skid, a soak left running over a weekend. Each one takes a little thickness, and thickness is the whole protection.

That is the awkward part. None of those steps looks like damage on the day. The vessel comes back clean, the batch runs, and nobody writes anything down. The loss only shows up years later as a thin spot, by which point the cause is a hundred cycles back.

It is also why glass-lined reactor cleaning is worth writing down rather than inheriting. A recipe that came with the vessel in 2014 may have been built for a duty the plant no longer runs.

This is why the acid side is the safer direction for these vessels. Citric chemistry does the work without the alkali load, and the vessel's own lining spec still governs. Our comparison of citric and nitric acid shows why the acid route is gentler on equipment.

How Much Thermal Shock Can a Lining Take?

Less than most operators assume, and no single number covers your vessel. ISO 28721-2 sends you to the thermal shock and heating diagrams in ISO 28721-3 for a built apparatus. What matters is the size of the temperature jump rather than the level you finish at.

The same 2015 edition set a crack formation temperature of 190 °C or more for chemical enamels, tested to ISO 13807. For enamels used on accessories such as agitators, baffles and thermometer wells, it said at least 170 °C may be permissible.

ISO 13807 gets there by heating a specimen and quenching it into water. The figure is the size of the drop the enamel survives.

Both are procurement figures. ISO says the result "represents a characteristic of the specimens tested" and "shall not, therefore, be applied directly to an apparatus or accessories". They tell you what to accept when the enamel is supplied. They do not tell you which part of your reactor cracks first, and they are not permission to shock your vessel by 190 degrees.

In practice, charging cold water into a hot reactor is the classic way to ruin a lining. It happens most often when a shift is running behind and somebody wants the vessel cool.

The fix is dull and it works. Ramp the jacket. Bring wash water in near the wall temperature, and leave time between a hot step and a cold one. None of that needs new equipment. It is the difference between a lining that lasts and one that crazes early.

How Do You Check a Lining for Damage?

With voltage rather than with your eyes. A pinhole too small to see will still expose bare carbon steel to the process. Once the steel is exposed, the damage spreads.

The corrosion is only half of it. US GMP requires that product-contact surfaces "shall not be reactive, additive, or absorptive" in a way that alters the drug (21 CFR 211.65). Bare steel in a reactor is reactive, and it can add iron to the batch.

ISO 28721-1 sets the quality rules for glass-lined equipment and the tests a maker runs. Its test list includes a visual check, a high-voltage test, crack testing and coating thickness (ISO 28721-1:2019).

That list is what a maker checks before delivery. The high-voltage test is also called the spark test. After a cleaning job, it is the one we would repeat, because it finds the defects a visual walkdown misses. Passivation testing plays the same role on stainless. ISO 28721-1 also separates defects that can be repaired from those that cannot. That is the decision a plant actually faces when a spark test fails.

Run the spark test at the same points every time and write down where you tested. A single pass tells you those points are sound today. A run of passes tells you whether the lining is holding. That is the thing worth knowing before you plan a campaign.

There is a citric acid spot test for porcelain enamels, ASTM C282. It is worth knowing that its own scope rules it out here: it does not apply to finishes on chemical ware (ASTM C282-24). Reaching for it on a reactor would be using the wrong test.

Where Does Stainless Steel Come Into It?

In everything bolted to the reactor. The vessel and its fittings may be enamelled. The valves, the transfer lines, the instrument tappings and the receiver downstream usually are not.

Those parts behave like any other stainless steel in a hot process. They grow rouge. They lose passivity. They need derouging where required, then passivation to rebuild the chromium oxide film. Passivation is its own cycle, run at a different strength and dwell from a cleaning pass, even where both use citric acid.

This is the half that gets forgotten, because the lining is the expensive part and it takes the attention. A reactor can pass its spark test while the transfer line feeding it sheds iron oxide into every batch. Our guide to what rouging is explains how that film forms and why it returns.

Glass and steel pull in opposite directions, and one recipe cannot cover everything on the vessel. Here is what a single reactor puts in front of you:

Choosing a Glass-Lined Reactor Cleaning Method

Work from the surfaces you actually have, in this order.

1. Find the lining spec in the vessel documents. The enamel grade tells you what the maker certified it to take.
2. Rule out hot caustic unless the lining spec and the vessel maker both say otherwise.
3. Set a ramp, so nothing enters the vessel far above or below the wall temperature.
4. Treat the stainless pipework separately. What is safe for glass will not rebuild a passive layer.
5. Spark test after the work as well as before it, so you know the condition you are handing back.
6. Record the lot, the chemical and the readings. An auditor will ask, and the vessel cannot answer.

Plants across the Gujarat chemical belt, from Vapi and Ankleshwar to Dahej, run these vessels hard on short campaigns. The plants we see keeping linings longest are the ones that treat the ramp rate as seriously as the chemistry.

Before You Book the Work

It comes down to one thing. A glass lining and the steel around it want different care, and a single visit has to serve both.

Before the next shutdown, pull the papers for your worst-behaved reactor. The grade on them is what your maker stands behind. Compare it against what your cleaning cycle really does.

If the two do not match, that is worth sorting out before the vessel is opened. Talk to our team and bring those papers with you.

Frequently Asked Questions