Pressure Sensors for Steam and Aggressive Media: Syphons, Diaphragms, Temperature

Steam and aggressive media: where the choice really starts

Short answer. A pressure sensor for steam is chosen by the temperature at the tapping point, not by the pressure. A standard industrial transmitter lives up to roughly 80…100 °C: the Wika A-10 base limits are 0…80 °C, and the standard row for the S-11 and S-20 gives -30…+100 °C. Steam is hotter than that. From there you have exactly two legitimate routes, both described in the manufacturer's datasheets: put a syphon between the process and the instrument so a condensate pocket forms, or order a version with a cooling element. For chemistry the question is different — not "stainless or not", but the specific steel grade, the sealing material, and the pressure limit that seal allows.

Every number below comes from WIKA datasheets, because these are the families we actually stock: S-20 (PE 81.61), S-11 (PE 81.02), SA-11 (PE 81.80), IS-3 (PE 81.58), A-10 (PE 81.60), the 910.15 syphon (AC 09.06) and the 910.32 cooling element (AC 09.21). Page references for every claim are at the end. We do not convert steam pressure into saturation temperature here: take that from your own process data and we will supply the instrument limits.

The first limit is medium temperature, not pressure

This is the line that kills most sensors. Look at how widely it varies even within one manufacturer.

Two practical conclusions. First, an A-10 on a steam line without a syphon or a cooling element sits outside its datasheet, and that is not "safety margin", it is growing measurement error and a shortened service life. Second, the ambient temperature is rated too, and the A-10 datasheet adds a condition people routinely skip: the stated ambient limit applies only to a housed mounting location that is protected from condensation, and the restrictions and derating of that ambient limit themselves depend on the medium temperature.

The syphon: a DIN 16282 water pocket, not a decorative loop

Datasheet AC 09.06 classes the syphon as a cooling element in its own right: under "Applications" it is listed as a cooling element for protection from excessive heating of liquids, gases and vapours. The mechanism, given further down under "Description", is nothing like a radiator: condensate collects inside the syphon, which prevents the ingress of hot media into the pressure measuring instrument. Hence the key practical rule from the same datasheet, ignored on half the installations we see: it is recommended to fill the syphon with a cooling barrier liquid before commissioning. An empty syphon protects nothing during the first minutes of operation.

The geometries are standardised in DIN 16282, and the choice is not arbitrary. The U-form is intended for a horizontal pressure tapping, the trumpet form for a vertical one. The compact form is shorter: it minimises the influence of vibration and protects the instrument from pressure spikes and pulsations. Forms A and C have a threaded connection, forms B and D are weld-in.

The syphon has limits of its own, and they must be tracked separately from the sensor's: permissible temperatures to 550 °C, nominal pressures to 420 bar. But those are family limits. A specific version can be far more modest: form A with a G ½ connection (LH-RH adjusting nut per DIN 16283) has a maximum operating pressure of 160 bar. Material matters too — the datasheet prints separate pressure-temperature diagrams, and they do not all carry the same curves: the U-form and trumpet form, like the connecting pipes, get two curves each — stainless and carbon steel — while the compact form is plotted for stainless steel only.

The cooling element: 150 or 200 °C, and what it costs

A cooling element is a different mechanism. Per AC 09.21 it lowers the medium temperature through air circulation and heat radiation, and the datasheet adds something useful: doing so also reduces any possible temperature error. There is a side effect few people know about — the same element can be used to warm a very cold medium, provided the ambient temperature is higher than the medium's.

There are two versions, differing in the number of fins: 150 °C with three fins, 200 °C with five. Both connections are G ½ B per EN 837 and the material is 1.4571. The 600 bar maximum load carries two conditions at once in the datasheet: suitable sealing in accordance with EN 837 and a temperature below 100 °C at the measuring instrument. If the instrument runs hotter than that, 600 bar is not available to you.

Now the cost. On an S-20 with an integrated cooling element the medium limit rises to -40…+200 °C, but two restrictions arrive with it: the maximum range of that version is 400 bar, and the cooling element itself is only available for the G¼A, G¼B, G½B, ½NPT, R¼ and PT¼ connections. The ambient is then capped at -40…+125 °C with derating, and the datasheet gives formulas rather than approximations. They read like this: with a medium below 125 °C the ambient may go up to 125 °C; from 125 °C upwards the permissible ambient is -0.62 × T_med + 202 °C. The other way round: with an ambient below 80 °C the medium may reach 200 °C, and from 80 °C upwards the medium limit is -1.61 × T_amb + 326 °C. Substitute the numbers: at a 200 °C medium the ambient must not exceed 78 °C, and at a 100 °C ambient the medium is capped at 165 °C. So the cooling element does not remove the problem, it moves it to cabinet ventilation or to shading the instrument.

The second part of the cost is mechanical. S-20 vibration resistance is 20 g as standard, and 40 g with the metal M12 × 1 circular connector — but it drops to 10 g for instruments with a cooling element. The S-11 is the same: 20 g without, 10 g with, and shock resistance falls from 1,000 g to 400 g. A long cantilever, no surprise there. On a compressor or vibrating pipework that becomes a separate question about where to mount it.

Diaphragm, body, seal: what actually touches the medium

"Stainless diaphragm" is not yet an answer about chemistry. At least three components are in contact, and the weakest is usually not the diaphragm.

Start with the diaphragm. The S-20 datasheet splits the wetted material by range: 316L up to and including 25 bar, 316L plus PH grade steel above 25 bar, ASTM 630 with PH grade steel above 1,000 bar. The case is 316Ti. On the S-11 the wetted parts are 316Ti for G½B and G1B flush, and 316L for the hygienic G1B. On the SA-11 it is 1.4435, with a rated surface texture: Ra no more than 0.76 µm per ASME BPE SF3, except for the weld seam.

Now the weak point. The seal limits both temperature and pressure, and the S-11 table shows that most clearly. Without a cooling element up to 100 °C: NBR and FKM hold up to 600 bar, EPDM only up to 200 bar. Up to 125 °C: NBR 600 bar, FKM already down to 400, EPDM the same 200. With a cooling element up to 150 °C: FKM up to 300 bar, EPDM up to 200. The same seal therefore gives a different pressure limit at a different temperature, and ordering "0…400 bar with EPDM at 120 °C" simply makes no sense. There is also an FFKM option, but only for the G½B flush connection: there it holds 600 bar at all medium temperatures. And one more thing: the minimum permissible temperature for FKM/FPM on the S-11 is -20 °C, so an outdoor winter install is already at the edge.

The third component is the fill. On the S-20 it is synthetic oil for ranges up to 25 bar and a dry measuring cell above 25 bar. On the S-11 it is synthetic oil or a food-compatible fluid per FDA 21 CFR 178.3750. On the SA-11 the fill is food-compatible by definition: KN 77 (FDA 21CFR178.3750) or Neobee M-20 (21CFR174.5). In food production that is not paperwork, it is what the auditor asks about.

Oxygen and hydrogen deserve a separate note, because "aggressive" there means something else entirely. The S-20 oxygen version is specified by residual hydrocarbons (under 500 mg/m² for ranges below 30 bar and under 200 mg/m² above), limited to -20…+60 °C, a maximum range of 400 bar, and a leakage rate below 1 × 10⁻⁶ mbar·l/s. The hydrogen version has 316L and Elgiloy wetted parts. These are ordered versions, not "the same sensor, degreased".

The same media limits apply to pressure switches, they simply show up differently: the copper bellows of a PSM-520 or the brass connection of a Dwyer DX do not survive steam or chemistry, and the failure reads not as an error band but as a switch that has stopped switching. The symptoms and the check order are in our pressure switch fault diagnosis.

Flush diaphragm: viscous, particle-laden, crystallising

When the medium is not chemically aggressive but simply will not flow, the problem is the pressure port itself. A narrow channel clogs, sets or packs with particles, and the sensor reports a pressure that no longer exists. The answer is a flush connection, and that is what the S-11 is for: its datasheet names the purpose directly — viscous and particle-laden media, plus vacuum tightness.

There is a constraint here that breaks half the enquiries. The S-11 range depends on the connection: G½B flush covers 0…2.5 to 0…600 bar, while G1B flush covers only 0…0.1 to 0…1.6 bar. The hygienic G1B in accordance with 3-A Sanitary Standards gives 0…0.1 to 0…25 bar. So asking for "a flush G1 diaphragm at 100 bar" will not work, and that is not a stock question.

If steam, washdown and crystallisation come on top of viscosity, that is the SA-11. Its datasheet lists the media both broadly and specifically: gases, compressed air, vapour; liquid, pasty, powdery and crystallising media; pure-steam systems for SIP. The aseptic process connections are rated for process temperatures up to 150 °C, the diaphragm seal parts are fully welded, the instrument is suitable for SIP and CIP, and ingress protection goes to IP68. Vibration resistance is 15 g, shock resistance 500 g.

Hazardous areas: where the IS-3 belongs

If the medium is also explosive, an ordinary transmitter is out regardless of its temperature rating. The IS-3 is the intrinsically safe Ex ia version, and its datasheet claims suitability for medium temperatures up to +200 °C, with optional ranges of -40…+150 and -40…+200 °C.

Do not read that as "IS-3 = 200 °C in a gas zone". On page 8 the datasheet states an honest caveat: for ignition protection type Ex ia the permissible medium and ambient temperature limits depend on the option selected, the EPL, the temperature class, the chosen electrical connection and the chosen seal — there is no single universal number. For Ex ec and Ex tc the limits are fixed and short instead: T6 is -15…+55 °C, T4/T5 is -15…+70 °C. So an Ex selection always starts from the zone marking, never from the temperature you would like.

What we do not do

We do not put an A-10 or any other general-industrial transmitter on a steam line "with a margin" — the datasheet has a number, and it is 80 °C for the base version.

We do not mix up the two ways of dealing with temperature. The datasheet does class the syphon as a cooling element, but what it works by is the condensate plug — which is why an empty syphon protects nothing and why it is filled before commissioning. The finned 910.32 lowers the temperature by convection and radiation, and has its own ceiling: 150 °C with three fins, 200 °C with five. Different mechanisms, different conditions of use.

We do not convert steam pressure into temperature from tables — we are not citing that source here, so the medium temperature comes from your process data.

And we do not take on a steam or chemical line without input data. If the data is missing we will say so, rather than ship "something stainless".

What we carry

Wika pressure transmitters, by family: S-20 at 0…16 bar for harsh conditions, S-11 at 0…4 bar and S-11 at 0…10 bar with a flush diaphragm, SA-11 at 0…10 bar for steam, CIP and SIP, IS-3 at 0…10 bar for hazardous areas, A-10 at 0…10 bar for ordinary industrial duty and O-10 at 0…10 bar with an M12 connector. The whole section is pressure sensors, the brand as a whole is Wika, the brand itself and its other lines are on the Wika page, and the full sensor catalogue is here.

We do not list the 910.15 syphons or the 910.32 cooling elements on the site: they are accessories matched to a specific connection and a specific pressure-temperature pair, so a manager selects them together with the sensor on request. The SA-11 and the IS-3 are a similar story: these are configured versions, and the price for your particular configuration comes from a manager on request. The sensor section also holds Endress+Hauser items (including ceramic-diaphragm versions) and Siemens ones — we quote no parameters for those here, because this pass verified WIKA datasheets only; if that is specifically what you need, ask a manager. The current status of a specific part number is checked by a manager on request.

And a word on budget, because it is asked immediately. If the medium is neither hot nor aggressive and the duty is ordinary — water, air, hydraulics within 80 °C — there is nothing to pay extra for: take the A-10 at 0…10 bar, which is a series version rather than a made-to-order configuration. Steam and aseptic instruments cost more for exactly what is in them: a cooling element, a fully welded diaphragm seal, a rated surface texture, certificates. If you are not sure which side of that line your job falls on, send us the temperature at the tapping point and we will tell you whether an A-10 is enough.

If the signal goes to a controller, have a look at wiring sensors to a PLC and at 24 V power supplies: in a two-wire 4…20 mA loop the supply and the input are sized together.

Frequently asked questions

Can I mount a pressure transmitter directly on a steam line?

It depends on the temperature at the tapping point, not on the pressure. A standard S-20 is rated for -30…+100 °C; the version with an integrated cooling element raises the limit to -40…+200 °C, but with a capped ambient and a derating formula. If the temperature is higher than that, or unknown, a syphon goes between the process and the instrument.

Syphon or cooling element — which one?

They solve different problems. The syphon maintains a condensate pocket that keeps hot media out of the instrument, and the compact form additionally damps pressure spikes and pulsations. The cooling element lowers the temperature by convection and radiation, giving 150 °C with three fins or 200 °C with five. On steam we sometimes fit both — but that is our installation practice, not a recommendation found in the datasheets.

Why does vibration resistance drop with a cooling element?

Because the instrument is moved out onto an extended stem. On the S-20 the rated 20 g becomes 10 g for versions with a cooling element; on the S-11 it is likewise 20 g against 10 g, with shock resistance at 1,000 g against 400 g. If the sensor sits on a compressor or on vibrating pipework, that has to drive the choice of mounting point.

Which sensor suits CIP/SIP and crystallising media?

The SA-11. Its datasheet explicitly lists vapour, pasty, powdery and crystallising media and pure-steam systems for SIP; the aseptic connections are rated for process temperatures up to 150 °C, the diaphragm seal parts are fully welded, and the wetted material is 1.4435 with a surface texture of Ra ≤ 0.76 µm per ASME BPE SF3.

Will a stainless diaphragm survive any chemistry?

No, and the diaphragm is usually not the first thing to go. The seal limits temperature and pressure at the same time: in the S-11 table EPDM holds 200 bar, FKM at 125 °C holds 400 bar instead of 600, and the minimum for FKM/FPM is -20 °C. Oxygen and hydrogen need dedicated versions altogether: the S-20 oxygen version is specified by residual hydrocarbons and limited to -20…+60 °C and 400 bar, while the hydrogen version uses 316L and Elgiloy.

When do you not need a syphon?

When the temperature at the tapping point already sits inside the instrument's rated limits: -30…+100 °C for a standard S-20, up to +150 °C for the SA-11, and up to +150 °C for an S-11 with a cooling element. The datasheet wants a syphon for protection against excessive heating, not as a matter of course, so on a medium that is not hot it is simply extra volume and one more joint. The second case is a viscous or crystallising medium: a narrow loop there will clog sooner than it protects, and the right answer is a flush diaphragm — S-11 or SA-11 — rather than a pipe.

What to do next

Send us six items and we will select the instrument together with whatever goes between it and the process: the medium (name, concentration, whether it carries particles), the temperature at the tapping point in normal operation and at peak, the working and maximum pressure, the process connection (thread or aseptic clamp), the output and supply you need, and finally the area — the Ex marking if there is one, plus the vibration at the mounting point. If nothing in our catalogue fits your conditions, we will say so plainly. Write to us via the contacts page.

Related reading: industrial pressure sensors: types and selection, pressure sensor for a booster set, differential pressure sensor and gauge, pressure switch fault diagnosis.

Sources. WIKA data sheet PE 81.61 ∙ 02/2026, "Pressure transmitter, model S-20" — p. 1 (stability, load cycles), pp. 5-6 (process connections, seals, overpressure limits), p. 11 (medium and ambient temperature limits, wetted materials, pressure transmission medium), p. 12 (derating curve and formulas for cooling elements, vibration and shock, service life), p. 13 (oxygen and hydrogen versions), p. 14 (approvals including UkrSEPRO, MTTF). WIKA data sheet PE 81.02 ∙ 04/2026, "Flush pressure transmitter, model S-11" — p. 1 (purpose, vacuum tightness), p. 4 (temperatures with and without a cooling element, vibration and shock), p. 5 (connections and ranges, sealing table with pressure limits, wetted materials, fill fluid). WIKA data sheet PE 81.80 ∙ 03/2026, "Pressure transmitter, model SA-11" — p. 1 (media, SIP/CIP, fully welded diaphragm seal, IP68), p. 6 (temperatures, vibration and shock, materials, surface texture, fill fluid). WIKA data sheet PE 81.58 ∙ 02/2026, "Pressure transmitter, Intrinsic safety Ex ia, model IS-3" — p. 1 (Ex ia, up to 200 °C), p. 8 (dependence of permissible temperatures on option, EPL, temperature class, connector and seal; limits for Ex ec and Ex tc; oxygen and hydrogen versions), p. 15 (optional medium temperature ranges). WIKA data sheet PE 81.60 ∙ 02/2026, "Pressure transmitter, model A-10" — p. 9 (standard and extended temperature ranges, the condensation-protected mounting condition, humidity). WIKA data sheet AC 09.06 ∙ 01/2025, "Syphons and connecting pipes, model 910.15" — p. 1 (the condensate-pocket principle, filling before commissioning, DIN 16282 geometries, 550 °C and 420 bar limits), p. 2 (pressure-temperature diagrams), p. 3 (form A: connections, LH-RH adjusting nut per DIN 16283, 160 bar, materials). WIKA data sheet AC 09.21 ∙ 01/2015, "Cooling element, model 910.32" — p. 1 (principle, reduction of temperature error, use for warming cold media), p. 2 (150 °C with three fins and 200 °C with five, G ½ B per EN 837, 600 bar, material 1.4571).