When the liner goes, it usually goes quietly. The reading drifts a little, then gets noisy, then one day the meter sits at zero while the pump is clearly running. By the time somebody pulls the sensor, the liner has a bulge the size of a thumb at the flange joint and the process has been off-spec for three weeks.
We get more questions about liner and electrode combinations than about any other part of a flow meter, and the reason is simple. The electronics either work or they do not, and you find out in a week. Materials fail over months, in a medium you may not have fully characterised, at a temperature that moves with the season.
Two failure modes, both slow
Liner failure is chemical attack, thermal cycling, or permeation. PTFE in hot concentrated caustic blisters. Rubber in a solvent swells and delaminates. A liner under vacuum collapses into the flow path, and the meter still reads — badly.
Electrode failure is corrosion, passivation, or fouling. Hastelloy C in hydrochloric acid above 5% at 60°C corrodes. Titanium in dry chlorine ignites. A platinum electrode in a sulphide stream fouls and the signal drifts.
Both produce plausible-looking wrong numbers before they produce an obvious fault. That is what makes them expensive.
The pairing table
| Medium | Liner | Electrode | Watch out for |
|---|---|---|---|
| Sulphuric acid 98% | PTFE | Hastelloy C or tantalum | Dilution at the seal face causes local attack |
| Sulphuric acid 10–30% | PTFE or PFA | Tantalum | Hastelloy is not enough at this concentration |
| Hydrochloric acid | PTFE or PFA | Tantalum | Hastelloy C only to about 5% at ambient |
| Hydrofluoric acid | PTFE | Platinum or tantalum | HF attacks glass and most alloys; nothing here is cheap |
| Nitric acid | PTFE | Titanium or Hastelloy C | Titanium passivates well; avoid in reducing conditions |
| Sodium hydroxide, hot | PFA or F46 | Hastelloy C or titanium | PTFE blisters above 100°C in strong caustic |
| Sodium hypochlorite | PTFE or PFA | Titanium or Hastelloy C | Off-gassing creates bubbles on the electrode |
| Organic solvents | PTFE or PFA | Hastelloy C | Rubber liners swell; check explosion protection |
| Slurries with solids | Hard rubber or polyurethane | 316L or tungsten carbide | Abrasion dominates; chemistry is secondary |
| CIP caustic and acid | PFA | Hastelloy C | Thermal cycling, not concentration, is the killer |
Three combinations account for most of the failures we are asked to diagnose: Hastelloy C in hydrochloric acid above 5%, titanium in dry chlorine, and PTFE in hot concentrated caustic. All three are standard catalogue pairings that happen to be wrong at the edges.
Liner materials, cheapest first
Hard rubber. Cheap, abrasion-resistant, good to about 80°C. Not for solvents or oxidising acids. The right answer for slurries and anything where wear matters more than chemistry.
Neoprene. Similar niche, slightly wider chemical range, lower temperature ceiling. Common on water and mild chemical duty.
Polyurethane. Best abrasion resistance of the group, and the right choice for slurries with sharp particles — mining tailings, ash slurry, sand-laden water. Limited to around 60°C.
PTFE. The general-purpose chemical liner. Broad resistance, 180°C ceiling, but it is a relatively thin sleeve and it permeates. On hot aggressive service, particularly under vacuum, it deforms.
PFA and F46. Thicker, tougher, better permeation resistance. The right choice for hot caustic, high-purity duty and any application with vacuum. Most chemical plant specifications end up here after the first PTFE liner fails.
Temperature is the variable people underestimate
A pairing that works at 40°C may fail at 110°C. Chemical resistance tables are almost always quoted at ambient, and two things change with temperature: the reaction rate of the attack, and the mechanical properties of the liner.
Above 100°C we work to a simple rule: drop one step in chemical aggressiveness from the ambient recommendation. Hastelloy marginal at ambient becomes tantalum at temperature. PTFE marginal becomes PFA.
Thermal cycling is the second variable. A meter that sees 25°C on Monday and 130°C on Tuesday will fail at the liner-to-flange joint long before it fails in the bore. If the duty cycles, specify a liner with mechanical retention rather than a simple press fit.
Conductivity: the quiet disqualifier
Electromagnetic meters need conductivity, and the figure usually quoted is 5 µS/cm. That is optimistic. Below about 20 µS/cm the signal-to-noise ratio gets difficult, particularly on large meters at low velocity.
Duties that sit near the limit: demineralised water at 0.1 µS/cm (no), condensate at 1–10 µS/cm (marginal), distilled solvents (no), dilute organic acids (usually fine). Below 20 µS/cm, either specify a low-conductivity build with capacitive electrodes, or change technology.
Installation in a chemical plant
Full pipe, always. A vertical rising line is best; horizontal with electrodes at three or nine o’clock is acceptable. Never at the top of a horizontal line, where gas accumulates over the electrodes.
Earthing. Chemical plants carry more stray current than any other environment. Separate earthing boss, copper braid to the pipe, never through the flange bolts.
Explosion protection. Solvent lines usually need Ex d or Ex ia. Specify at order stage — an Ex enclosure cannot be retrofitted to a standard meter.
Straight run. 5D up and 3D down as a floor. Behind a control valve or a pump, 10D and 5D.
Commissioning checks worth the hour
- Fill the line, confirm zero flow reads zero with the downstream valve closed.
- Test the empty-pipe alarm by draining the line.
- Verify conductivity with a handheld meter at operating temperature.
- Record initial electrode resistance — that is the baseline for future fouling diagnosis.
- Log the liner batch number against the meter serial. When a liner fails in year three, that record makes the warranty conversation short.
Frequently asked questions
One meter for both acid and caustic in the same plant?
Only if you accept the more aggressive pairing for both. Tantalum handles most acids but is attacked by strong alkali above pH 10. In a plant running both, two meters is usually cheaper than one compromise.
How do I know the liner has failed?
The reading drifts, then gets noisy, then goes to zero or full scale. Physically you will often see a bulge at the flange. By the time the reading is obviously wrong, the process has usually been off-spec for weeks.
Is Coriolis better for chemical duty?
For small lines, high viscosity, or where density is also needed, yes. Above DN50 the cost gap is large and electromagnetic is the practical answer.
What about two-phase flow?
Electromagnetic meters handle gas slugs badly — the reading spikes. If the process can carry gas, install where the line is guaranteed liquid-full, or move to Coriolis with two-phase capability.
The same electrode logic appears in pulp mills, where white liquor is hot caustic with sulphide and the failure mode is identical: flow meters in pulp and paper mills. And when the chemical is being metered into a food process rather than a reactor, sanitary flow meters for food and beverage plants covers the hygienic constraints that come on top.