PTFE seal

PTFE is not rubber and does not behave like it. Eastern Seals describes PTFE as hampered by very poor elastic memory at room and low temperatures, which is what makes an O-ring moulded from solid PTFE awkward to fit and unreliable once fitted. It also cold flows. Under sustained load the material creeps and the sealing force bleeds away, which is why PTFE gaskets are usually compounded with fillers to control it.

So a PTFE seal on a pump is a composite. PTFE provides the chemical barrier and something else provides the load. In a bellows or a wedge seal that job goes to a metal spring. In an encapsulated O-ring it goes to a rubber core sitting inside a fluoropolymer jacket. Vulcan Seals describes its encapsulated ring as an unbroken FEP or PFA jacket enclosing an elastomeric core, and says plainly that the core is there because solid PTFE rings have cold flow problems.

This is the difference that matters when you compare PTFE with EPDM or FKM. Those two are elastomers and you specify them as a material. PTFE is a construction.

A PTFE seal reaches a pump in one of three constructions: a PTFE bellows seal, a PTFE wedge seal, or an encapsulated O-ring. They are not interchangeable, and the one you need depends on whether you are sealing a rotating shaft or a static joint.

The bellows is the one worth understanding. John Crane positions the metal components, including the springs and clamp rings, outside the pump seal chamber so that they are isolated from direct contact with the process fluid. On a strongly corrosive duty that is the whole argument: there is no metal spring standing in the acid waiting to fail, and the wetted path is PTFE and the seal faces.

Encapsulated O-rings get mistaken for PTFE O-rings, and they are not the same thing. Vulcan Seals rates an FEP jacket for continuous use to 205°C and short excursions to 260°C, and a PFA jacket to 260°C continuous and 300°C for short periods. The core inside that jacket is silicone or FKM, so a nicked jacket puts a quite different material in contact with your medium.

The wedge is the least exotic of the three and it is often enough. John Crane describes it as a tough thermoplastic sealing element in place of the usual O-ring, inert to most corrosive chemicals, solvents and caustics, and non-leaching. It suits chemical processing and pharmaceutical duties where the medium rules out rubber but does not need a full bellows.

Chemical inertness is not mechanical toughness, and the two get conflated. A PTFE seal is less forgiving than a rubber one in three ways, and all three show up in service rather than on the datasheet.

• It creeps. A bolted PTFE gasket loses load over time. Filled grades are the usual answer, and a re-check of the joint after the plant has been hot is worth the trip.
• It has no give. Rubber accommodates a shaft that runs out slightly and a face that sits fractionally off. PTFE does not, so alignment and fitting matter more.
• It is not a temperature or pressure exemption. The polymer will survive the chemistry. The bellows, the faces and the housing still have their own limits, and those are the numbers that bound the duty.

There is also a specification trap in the word "PTFE". A filled grade is no longer pure PTFE, and the filler has its own chemical behaviour. Check the grade, not the name.

The seal is the weak point in almost any pump, and the medium is only one of the things that kills it. Dry running is the other, and it is the one we see most often on the phone. A mechanical seal needs liquid to lubricate it. Take the liquid away and the seal goes, whatever the sealing element is made from.

The causes are unglamorous. Suction conditions that were never right, a line that gets pumped empty, or somebody shutting a valve and forgetting it is shut. If you are on your third seal in a year, the material is probably not the problem. Look at the suction side first, then at what the pump is being asked to do when nobody is watching.

PTFE is the widest chemical window available, and specifying it for a duty that does not need it buys complexity you will pay for at every rebuild.

On water, hot water and cleaning in place, EPDM is the standard fit and it is the right one. Alfa Laval quotes EPDM as the standard product-wetted elastomer on the DuraCirc, SolidC, SX and OptiLobe alike, and offers FPM as an option on each. PTFE appears in that world as a gasket or an encapsulated O-ring rather than as the shaft seal. On hydrocarbons, fuels and mineral oils, FKM normally covers it.

PTFE earns its place when the medium sits outside both windows, or when the consequence of a leak is bad enough that you want no elastomer in the wetted path at all. That is the reasoning behind most of the sealing decisions on a corrosive duty.

The medium, its concentration and its temperature. Those three fix the seal, and they are the three that enquiries usually arrive without.

Then be careful with the compatibility chart you were sent. Trelleborg Sealing Solutions makes the point that charts are useful for the generic susceptibility of a class of polymer, that most of them list results from room-temperature exposure only, and that chemical attack is a time-dependent process. A material that shrugs off a chemical for an hour can fail on a continuous duty. Trelleborg's own recommendation is to work with an experienced seal supplier rather than rely on the chart, and that is our practice too. Where a seal choice is genuinely difficult, the seal manufacturer's advice is the one to take.