A poly tank fails at its fittings far more often than at its wall, and almost never because the fitting was too small. It fails because the wrong route was chosen for the wall the fitting sits in: a bulkhead gasket clamped against a curved wall, a pipe thread cut straight into polypropylene, or a PVC adapter screwed into a PP socket and turned until the thread gave way. Choosing poly tank fittings is a compatibility decision before it is a sizing decision, and the compatibility question has three parts — how the fitting passes through the wall, what seals it, and what the wall can carry around the hole.
This page works through those three parts in the order they decide the outcome. It sets out the three routes through a tank wall and who each one suits, why the thread is the weakest element on a plastic tank, how to choose a gasket from published chemical-resistance data rather than from a catalogue heading, what to measure before ordering, and what a stand under a 5,000 L tank is really being asked to do. The searches that lead here are broad — plastic water tank fittings, tank connection, gasket for poly tank, water tank with stand — and they resolve into the same short set of questions.
The numbers come from sources a buyer can check: the EN 12573 series for welded thermoplastic tanks and their flanged joints, ASTM F2389 for polypropylene pressure piping, ASME B1.20.1 for the thread itself, and published material data from INEOS, IPEX and Trelleborg for the polymer and the elastomer. Where a statement is our shop practice rather than a published figure, the text says so. The fittings we supply are covered on the product pages; this page stays on selection.
Key Takeaways
- There are only three ways through a thermoplastic tank wall: a socket welded into the wall, a bulkhead fitting clamped through a drilled hole, and a flanged nozzle. The first must be planned before the tank is built, the second is the retrofit route, and the third is the only one with a dedicated calculation standard in the EN 12573 series.
- ASTM F2389-23 forbids molding or cutting a pipe thread directly into polypropylene and permits threads only as brass or stainless inserts molded into the fitting. That rule governs pressure piping rather than a gravity tank outlet, but the reason behind it applies to both: polypropylene cold-flows, so a thread in it relaxes.
- NPT is a tapered thread that seals by interference along the flanks, with no gasket. In plastic it is a one-shot joint; re-making it removes material and the second make is worse than the first.
- Gasket choice is decided by the fluid, not by the temperature rating. In 50% caustic soda EPDM is rated fully resistant through 80 °C while FKM is rated no resistance at 20 °C; in 50% sulfuric acid the verdict reverses.
- Polypropylene expands 0.8–1.0 mm per metre for every 10 °C of temperature change. A two-metre run swinging 40 °C moves about 7 mm, and that movement has to be taken by the pipe layout or a flexible section, never by the gasket.
- The fitting needs a flat annular land wide enough for the gasket and a wall thick enough for the body to pass through with thread left for the locknut. Measure the wall at the hole; the nominal sheet thickness is not the number at the knuckle or the weld.
- Five thousand litres of water is about 5,000 kg, or 49 kN, before the tank and fittings are counted. On four legs that is 12.3 kN per leg as a static share, and the stand maker has to put a rated load per leg in writing.
What a Poly Tank Fitting Actually Has to Do
A fitting is a hole with three jobs. It has to hold the contents in, it has to leave the wall as strong around the hole as it was before the hole existed, and it has to be serviceable — because the day will come when a hose, a valve or a whole pipe run has to come off. Every route through a wall trades those three against each other, and the trade is what the rest of this page is about.
Three Loads Every Outlet Carries
The first load is the liquid head at the outlet. Hydrostatic pressure depends on depth, not on tank volume, which is why a shallow wide tank and a tall narrow one of the same capacity do not load their outlets the same way. Three metres of water above an outlet gives about 29 kPa at that outlet; the same outlet two metres from the top of a three-metre tank sees two-thirds of that. The gasket and the joint have to hold this continuously, at whatever temperature the contents are stored.
The second load is the pipe. A filled pipe run hanging off a side outlet is a lever arm, and the bending moment at the wall rises with the square of the unsupported length. A two-metre run of filled 50 mm pipe is a modest weight but a substantial moment, and it is applied to a plastic wall at the point where a hole has already removed material. This is the load that most often produces a crack at a tank outlet, and it is the load nobody measures.
The third load is movement. The tank expands when it is filled warm and contracts when it cools; the pipe run does the same, at a different rate and in a different direction. Polypropylene’s coefficient of linear thermal expansion is 8–10×10−5 cm/(cm·°C) for the homopolymer and 6–9×10−5 for the copolymer, which works out to 0.8–1.0 mm of movement per metre of run for each 10 °C. A run that swings 40 °C moves about 7 mm over two metres. Somewhere between the tank and the pipe that movement has to be absorbed, and if it is not, the gasket absorbs it by being worked loose.
Why “Tank Fitting” Means Two Different Products
The trade uses one phrase for two things. A tank fitting in the first sense is part of the vessel: a socket or a spigot welded or molded into the head, side or bottom, whose only threaded feature is the end that something attaches to. A tank fitting in the second sense is an accessory that clamps through a hole drilled in a finished wall — a bulkhead — and brings its own seal because it cannot rely on the wall’s material.
The difference matters because the responsibility differs. A welded socket is proved with the tank, in the same pressure or water test, and its quality is the fabricator’s. A bulkhead is installed on site by whoever drilled the hole, and its quality is the installer’s, the wall’s thickness, and the flatness of a surface that may be curved. Both are sold as poly tank fittings, which is why a search for the phrase returns a mixture of vessel parts and irrigation hardware.
What the Wall Is Doing While the Fitting Works
A hole is a geometric discontinuity in a stressed membrane. The wall of a filled tank is carrying hoop stress and, at the bottom, bending into the base; a hole interrupts both, and the stress that would have crossed the removed material has to go around the edge instead. Left unreinforced, that is where a crack starts, and a crack that starts at a hole edge does not stop at the gasket line.
Every one of the three routes is an answer to that problem. A welded socket puts a collar of material back around the opening and joins it by fusion. A bulkhead spreads the load through a wide gasket and a locknut on a flat land, and on a thin or curved wall it needs a doubler pad to give it that land. A flange carries the load into a ring of bolts that spreads it over an annulus far wider than the bore. None of them simply drill a hole and hope; the route that does is the one that fails.
How the wall itself was made sets what is available at the outlet. A welded tank can have a socket, a pad or a flanged stub designed into it at drawing stage — the process is described in plastic tank fabrication. A molded tank cannot, which is why the bulkhead exists at all.
Three Routes Through a Wall: Socket, Bulkhead and Flange
The route is decided by who owns the wall and when. If the tank is still a drawing, the fitting can be part of the vessel. If the tank is already standing full of liquid, the fitting has to be added without heat, which narrows the choice to one. Between those two situations sits the flanged connection, which can be designed in and can also be added to a tank that was built with a flange face already on it.
The Welded Socket: The Default on a Fabricated PP Tank
A socket is a short length of tube in the same material as the wall, welded to the wall around its full circumference, usually with a collar or a fillet of filler rod at the root. In a fabricated polypropylene tank it is the default side outlet, and it is the outlet with the fewest parts: no gasket in the wall, no second face, no locknut to work loose. The leak path through the wall is a weld, and a weld can be inspected and tested.
Its constraint is timing. A socket needs hot work and a qualified procedure, so it belongs to fabrication, in the same sequence and under the same inspection as the shell seams — see plastic tank fabrication. Adding one to a tank in service means draining, cleaning, inerting and welding on site, with the same test afterwards; the arithmetic rarely favors it over a bulkhead. Its second constraint is concentration: the weld that joins the socket to the wall sits at exactly the point where the pipe’s bending moment arrives, so the socket’s weld is a better weld than the tank’s average weld or the joint is a slow leak.
The Bulkhead Fitting: The Retrofit Route
A bulkhead passes a body through a drilled hole. A gasket sits on the wet face, the body goes through, and a locknut on the dry face pulls the assembly up until the gasket is compressed between the body’s flange and the wall. Some designs use a nut on both faces; the principle is the same. Nothing is welded, so nothing has to be drained and inerted, and the fitting can be moved later.
The seal is a compression seal, and it depends on two things the buyer often does not control: the wall thickness and the flatness of the wall around the hole. Internal pressure helps, because the liquid presses the gasket harder onto its seat, which is why the gasket belongs on the liquid side. What internal pressure cannot fix is a curved wall, because the gasket’s outer edge lifts off the surface and the compression is concentrated on a narrow arc. On a cylindrical wall that means a flat pad, or a machined spot face, or a bulkhead with a saddle.
There is a third weakness that is easy to miss. A bulkhead is a friction joint, and polypropylene creeps under sustained load. The compression that was right at installation relaxes over weeks and months, and the leak appears after a season rather than on the day. Re-checking the nut after the first thermal cycle is part of commissioning, not a sign that something was done wrong.
The Flanged Nozzle: The Route EN 12573-4 Designs
A flanged connection is a stub with a flat face, a gasket, and a ring of bolts — or a PP stub with a steel backing ring, which is how most thermoplastic flanges are made. It is the only one of the three with its own part in the EN 12573 series: Part 1 is general principles, Part 2 covers the calculation of vertical cylindrical tanks, and Part 4 is “Design and calculation of flanged joints” for welded static non-pressurised thermoplastic tanks. The existence of a separate part is the point. A flanged joint is designed, with a calculation, rather than selected from a shelf.
What a flange buys is demountability and diameter. It can be broken and remade without destroying anything, the gasket is a consumable that can be replaced, and it scales to sizes where a thread stops being sensible. What it costs is flatness and bolt load. Two faces have to be parallel and clean, the bolts have to be tightened in a sequence that loads the gasket evenly, and because the plastic creeps under the bolt load, the torque is not permanent either.
| Route | How the seal is made | Can be added to a built tank | Best suited to | Characteristic failure |
|---|---|---|---|---|
| Welded socket | Fusion weld at the wall; no elastomer in the wall | No — needs hot work on site | New fabricated PP tanks, side and bottom outlets | Weld crack at the root under pipe bending |
| Bulkhead | Gasket compressed between body flange and wall, locknut on the dry face | Yes — drill and clamp | Molded tanks, retrofits, small to medium bores | Relaxation as the wall creeps; lifting on a curved wall |
| Flanged nozzle | Gasket compressed between two parallel faces by a bolt ring | Only where a flange face exists | Larger bores, lines that must be broken and remade | Uneven bolt load; face not parallel |
The Thread Is the Weakest Link on a Plastic Tank
Whatever route is chosen, the end that something attaches to is usually threaded, and the thread is where a plastic tank leaks for reasons that have nothing to do with the gasket. The cause is a mismatch between how a taper thread is meant to seal and what polypropylene does under load.
NPT Is a Tapered Thread, and That Changes the Seal
ASME B1.20.1-2013, reaffirmed in 2018, is the standard for general-purpose inch pipe threads, and its scope covers NPT, NPSC, NPTR, NPSM and NPSL. NPT is the tapered one: the crests and roots are cut on a cone, and when a male and female thread are engaged the taper wedges them together until the flanks interfere and the spiral gap closes. There is no gasket. The seal is the interference itself, all along the engaged length.
That works in metal because the metal is stiff enough to hold the interference without yielding much. In plastic, the same wedge is pushing into a material with less than one percent of steel’s stiffness — a Young’s modulus of 1,300 MPa for PP homopolymer against about 200 GPa for steel — so the material yields a little and the interference relaxes. Turn it harder and the thread deforms further; turn it less and a spiral leak path remains. Between those two there is a narrow band that works, and it is narrower than the equivalent joint in brass.
Two practical consequences follow. A plastic NPT joint is best made once: every re-make removes material from the flanks and the second joint is looser than the first. And the straight-thread variants matter — NPSM is a straight mechanical thread rather than a tapered sealing thread, so a fitting threaded NPSM will not seal on its own and needs a gasket face to bear against. Mixing the two is a common reason a fitting that worked on one tank will not seal on the next.
What ASTM F2389 Says About Threads in Polypropylene
ASTM F2389-23 is the specification for pressure-rated polypropylene piping systems, covering pipe, fittings, valves and manifolds in metric and IPS Schedule 80 sizes, with a minimum pressure rating of 160 psi at 23 °C and 100 psi at 82 °C. Its treatment of threads is blunt. Threads are permitted only by way of metal inserts molded into the fitting, the inserts are to be brass or stainless steel, and threads shall not be molded or fabricated directly into the polypropylene plastic.
The boundary has to be stated clearly, because it is easy to overreach. F2389 governs pressure piping that carries water service at a rated pressure; a gravity outlet on a static tank is not pressure piping and is not designed to this specification. What the rule is worth to a tank buyer is its reason rather than its jurisdiction. If a standards committee will not accept a thread cut into polypropylene on a pressure-rated fitting, and instead requires a metal insert, then the same material behaviour applies to a thread on a tank outlet — it just has less margin behind it. On a tank, the thread is a means of attachment. The seal is the gasket or the weld.
Thread Engagement, Wall Thickness and the Locknut
Three lengths have to agree: the length of thread on the fitting body, the thickness of the wall it passes through, and the depth of the locknut. If the wall is too thick, the nut lands on unthreaded body and cannot load the gasket. If the wall is too thin, the thread runs out before the nut is tight, and the installer reaches for a washer or a second gasket to take up the gap — which is where a great many leaks begin.
The thickness to measure is the thickness at the hole, not the nominal sheet. A formed or welded wall is thinner where it has been stretched at a knuckle and thicker where a weld collar or a doubler was added, and the difference can be the whole margin. A caliper at the drilled hole, on the actual tank, is the only measurement that counts; the drawing is a starting point.
The Metal Insert Question
If the thread is an insert, its material is a second chemistry problem sitting inside the first. Brass is a copper alloy and copper alloys are the wrong answer for some of the duties polypropylene tanks are bought for; stainless steel is the wrong answer for others, depending on chloride content and on whether the duty is aerated. There is no insert that is right for every fluid, so the insert has to be chosen against the same chemical-resistance data as the gasket, and the buyer should ask which metal is in the fitting before it is ordered rather than after it is installed.
Gaskets: The Decision That Decides Service Life
A gasket is the cheapest part on the tank and the one most likely to end a campaign. The mistake is to treat “chemical resistant” as a property a gasket has, when it is a relation between one elastomer, one fluid at one concentration, and one temperature. The published tables are four-dimensional for a reason, and reading a single cell out of them is how a gasket gets chosen that fails four degrees into the duty.
EPDM, FKM and What Actually Separates Them
Two elastomers cover most of the tank work. EPDM — ethylene propylene diene rubber — is the default for water, hot water, steam, alkalis and hypochlorite; it is poor in oils, fuels and aromatic solvents, which is exactly what one number shows: in IRM-OIL 903 over 70 hours, EPDM swells 136% by volume while FKM swells 4%. FKM — fluorocarbon rubber, sold under names including Viton — is the reverse: excellent in acids, hydrocarbons, oils and solvents, poor in caustic, ammonia, amines and low-molecular-weight esters.
Temperature separates them less than people expect. Trelleborg’s published service range is −45 °C to +150 °C for EPDM and −35 °C to +200 °C for FKM, with special grades reaching higher for short periods. A polypropylene tank above 100 °C is unusual, so both elastomers sit inside their range for most duties, and the choice is essentially chemical.
Caustic Soda: Where EPDM Wins Outright
Sodium hydroxide is the cleanest case in the tables. IPEX rates EPDM as fully resistant to 10%, 30% and 50% solutions at 20, 40 and 60 °C, dropping to limited resistance only at 80 °C. FKM at the same conditions is rated limited resistance and then no resistance at 10% from 40 °C, and no resistance at 20 °C at 30% or 50%. An FKM gasket in a 50% caustic tank is not a marginal choice; it is the wrong one at room temperature.
Acid and Chlorine: Where the Verdict Reverses
Sulfuric acid flips it. At 50% concentration EPDM is rated limited resistance from 20 to 80 °C — usable with a shorter service life and a larger safety factor — while FKM is fully resistant from 20 through 100 °C and limited only at 120 °C. Sodium hypochlorite flips it again, and more sharply than most buyers expect: at 3% concentration EPDM goes from fully resistant at 20 °C to no resistance at 60 °C, while FKM stays at fully resistant through 40 °C and limited at 60 °C. A disinfectant duty that is fine in winter is not fine in summer.
| Fluid | EPDM at 20 / 40 / 60 / 80 °C | FKM at 20 / 40 / 60 / 80 °C | Practical reading |
|---|---|---|---|
| Sodium hydroxide 10% | 1 / 1 / 1 / 2 | 2 / 3 / 3 / — | EPDM, decisively |
| Sodium hydroxide 50% | 1 / 1 / 1 / 2 | 3 / — / — / — | EPDM; FKM not usable |
| Sulfuric acid 10% | 1 / 1 / 1 / 1 | 1 / 1 / 1 / 1 | Either |
| Sulfuric acid 50% | 2 / 2 / 2 / 2 | 1 / 1 / 1 / 1 | FKM, for service life |
| Sodium hypochlorite 3% | 1 / 2 / 3 / — | 1 / 1 / 2 / — | FKM above ambient |
| Sodium hypochlorite 12.5% | 2 / 2 / — / — | 1 / 1 / 2 / — | FKM |
Note on the figures — The ratings are IPEX’s compound compatibility classes from the EPDM & FKM Chemical Resistance Guide, second edition (2020), whose data come from FIP S.p.A.: 1 is high resistance, 2 is limited resistance with a shorter service life and a larger recommended safety factor, 3 is no resistance and a dash means no data for that temperature. They describe the seals IPEX uses in its own valves and fittings, and are quoted here as published elastomer data, not as a guarantee of any particular gasket in any particular tank. The service-temperature ranges are Trelleborg’s; special grades extend them, and a supplier’s own datasheet overrides both.
Compression, Creep and Re-torque
A gasket works by being compressed, and compression is a strain held in the elastomer. Under a sustained load the elastomer relaxes, and the plastic flange face behind it creeps as well, so the two effects add. The joint that was leak-free at commissioning can be wet three months later without anything having been done wrong in between; it simply lost the preload it started with.
The answer is a schedule rather than a material. Check the fasteners after the first thermal cycle, and again after the first month, and record what was found. Where a duty runs hot, a thicker gasket does not help — it gives the elastomer more material to relax — and a stiffer one only helps if the flanges can take the extra bolt load without bowing. Re-torquing is the intervention that works, and it needs access, which is a drawing-stage decision.
Sizing the Fitting to the Tank, Not to the Pipe
Most wrong fittings are the right size for the pipe. The pipe is what the buyer is thinking about, and the pipe’s nominal size is the number on the fitting’s box. What decides whether the joint works is the wall the fitting goes through, and that has to be measured on the tank. A set of poly tank fittings is sized to the tank first and to the pipe second, and the two answers are not always the same number.
Measure These Five Things Before Ordering
Wall thickness at the hole. Measured with a caliper at the actual opening, not read from a drawing, because formed and welded walls vary.
The flat land available around the hole. How wide an annulus of flat surface exists for the gasket to sit on, and whether the wall curves away inside that annulus. On a small-diameter tank the curve is severe; on a large flat panel it is negligible.
The widest diameter of the fitting body. Not the thread size. A body with a shoulder, an O-ring groove or a hex section has to pass through the hole at its widest point, and the hole is sized to that.
The length of thread beyond the wall. Enough for the locknut plus the gasket compression, and enough that the nut does not bottom on the shoulder before the gasket is loaded.
What will hang from it. The pipe run, its support and its filled weight, because that is the moment the wall sees, and it may argue for a larger fitting with a wider flange rather than a longer spanner.
Hole Size, Flat Land and the Curved Wall
The hole must clear the body, which means it is usually larger than the thread’s nominal diameter, and that larger hole removes more wall. The gasket then has to cover the difference and still find flat material outside the hole to seal against. On a curved wall those two requirements pull apart: a hole big enough for the body leaves a land that is already falling away, and the gasket seals on a crescent instead of a ring.
The remedy is material, not torque. A doubler pad welded or bonded around the opening gives a flat face and puts back the section the hole removed; on a fabricated tank it is designed in, and on a molded tank it is the reason some bulkheads come with a saddle. Bolting harder does not flatten a curved wall. It bends the fitting and concentrates the load at the two points where the fitting meets the curve.
What Breaks When the Hole Is Oversized
An oversized hole leaves the locknut bearing on unsupported wall. The wall bends locally under the nut, the bend starts cracks that radiate from the hole edge, and the crack grows with every fill and empty cycle. This is not a fitting failure and it will not be fixed by a new gasket, because the leak path is through the wall beside the fitting rather than past it.
The second failure of an oversized hole is gasket extrusion. A soft elastomer under compression with nowhere to go squeezes into the gap between the body and the hole edge, and once it has extruded it has lost the section that was doing the sealing. Both failures are visible on inspection — a star pattern of cracks, or a gasket with a lip pushed into the bore — and both trace back to a hole drilled to the thread size instead of to the body.
PVC Into PP: The Adapter Problem
PVC tank connectors and PVC bulkheads are widely sold, and on a PVC tank they are the natural choice because PVC pipe and fittings are joined by solvent cement into what is effectively one material. Polypropylene cannot be solvent cemented. It is joined by heat, which is why a PP tank’s outlets are welded sockets or gasketed bulkheads, and why a PP fitting cannot be glued to a PVC pipe.
So the connection between a PP tank and a PVC line has to be mechanical, and that means a gasketed face rather than a bare thread. A PVC male adapter screwed into a PP female thread is two different polymers with different stiffness and different expansion rates, held together by the one feature that is least able to cope with either: a plastic taper thread. It may hold for a season. It is not a joint to design a process around. The arrangement that lasts is a flanged or bulkhead transition with a gasket of the right elastomer, and a support under the pipe so the joint is not carrying the run.
What Carries the Tank: Pallet, Plinth or Steel Stand
Three different structures get called a base, and only one of them is normally designed for the job. The searches that bring buyers here — water tank pallet, tank stand, tank stanchion, and the long phrase 5000 litre plastic water tank steel stand specifications — are all asking the same question in different words: what does the tank sit on, and who says it is adequate.
The Load the Base Has to Carry
Start with the arithmetic, because it removes the argument. A litre of water is about a kilogram, so 5,000 litres is about 5,000 kg, and multiplying by gravitational acceleration gives roughly 49 kN. That is the payload alone. The tank shell, any fittings and the contents of the pipework add to it, and the base has to carry the lot whether or not anyone sized it for that.
The same calculation at 1,000 litres gives about 9.8 kN, and at 10,000 litres about 98 kN. Nothing about the number is subtle. What is subtle is where it goes: through the tank’s own bearing surface into whatever is underneath, and if the tank stands on legs, through a small number of points rather than evenly across the base.
A Water Tank on a Pallet Is Not a Stand
A pallet exists to move a load, not to hold one for years. It is a handling device, made of timber or plastic, with no rated legs and no stated load path; its bearers concentrate the tank’s weight into a few lines on the floor, and its joints are designed for a forklift rather than for a decade of static load. A water tank on a pallet is fine for transport and for a tank that is emptied and moved; it is not a permanent foundation.
Where a pallet is genuinely wanted — for a portable tank, or for a tank that has to be shifted between positions — the sensible arrangement is a molded tank base or a steel skid with a stated rating, not a warehouse pallet under 5 tonnes of water. The distinction to keep is between a base that carries the load continuously and a base that carries it for the length of a journey.
Tank Stand and Stanchion: What the Specification Has to State
The phrase “steel stand specifications” is doing real work, because a stand without a specification is a set of legs. A stand that can be checked states its rated load per leg, the points on the tank where it bears, the size and thickness of each leg, how the leg meets the base plate, how the base plate is anchored, what resists lateral movement, and what the finish is. Those six items are what makes the difference between a stand and a guess, and each of them has to come from the stand’s maker in writing.
For a 5,000 litre tank on four legs, the static share is 12.3 kN per leg before the tank, the fittings and any wind or seismic load are counted. That static share is a starting number, not a design value: a real leg sees more than its quarter, because fill is never perfectly even, because wind pushes sideways on an empty tank, and because a leg that settles takes less and throws its share onto its neighbours.
| Support | Payload share per support at 5,000 kg | What it does not tell you |
|---|---|---|
| Three legs | 16.4 kN | Any single leg carries the whole load if the other two settle |
| Four legs | 12.3 kN | Assumes even fill and level ground; neither is guaranteed |
| Six legs | 8.2 kN | Statically indeterminate — the real distribution depends on adjustment |
| Eight legs | 6.1 kN | Same, plus more base plates to level and anchor |
Note on the figures — The per-support values are simple divisions of the payload (5,000 kg × 9.81 m/s² = 49.05 kN) by the number of supports. They are arithmetic, not a design method, and they exclude the tank’s own mass, the pipework, wind and seismic effects, and any eccentricity of the contents. A stand for a 5,000 litre tank has to be rated by its fabricator for the actual imposed loads, in writing; the table exists to show why leg count changes the number a leg has to carry.
Why Tank Stand DIY Ends in an Engineer’s Drawing
Building a stand is straightforward in the sense that welding four legs to a frame is straightforward. What is not straightforward is knowing which of the failure modes is live in a particular installation, because they are not the ones a builder inspects. A slender leg fails by buckling, not by the weld at its top. A base plate fails by punching through a slab that was never rated for a point load. An empty tank fails by overturning in wind, which is the case a full tank does not have. And a tank standing on a mezzanine puts its load into a floor that was designed for people.
The two cases that catch people are the empty tank and the eccentric fill. Wind on an empty tank produces uplift at the windward legs and the anchor has to hold it; a tank filled from one side, or standing on ground that settles unevenly, throws a larger share onto one leg than the division above suggests. Neither is visible from a photograph of a finished stand, and neither is fixed by making the legs thicker without checking the base plate and the anchor.
None of this argues against a steel stand. It argues for buying one that is rated, or for having one drawn. The conversation starts with the tank, the contents, the ground and the wind, and it belongs with an engineer rather than with a welder working from a picture.
Specifying a Set of Poly Tank Fittings
The last step is the one that prevents most returns: writing down what the fitting is for. A purchase order that says “2 inch bulkhead, PP” leaves the four decisions that matter — body size, thread standard, gasket material and the wall it goes into — to whoever fills the order, and any of the four can be wrong without the order looking wrong.
Match the Material to the Wall, Not the Catalogue
The fitting’s material has to match the tank’s, because the two are joined by fusion or by compression, and neither works across a material boundary. A polypropylene tank takes polypropylene sockets and PP-bodied bulkheads; a PVC tank takes PVC. What does not exist is a weld between PP and PVC, and what does not exist either is a solvent cement for polypropylene, so the adapter between two materials is always a mechanical joint with a gasket and, ideally, a flange.
The same logic governs the gasket. It is chosen against the fluid, its concentration and its temperature, from a table that lists all three, as the section above sets out. A gasket described only as “chemical resistant” has not been chosen at all.
Buy the Second Gasket
A gasket is a consumable. It will be replaced at some point, and the replacement will be needed at the least convenient time, which is why the second one should be bought with the first rather than ordered during a shutdown. It should also be the same elastomer as the first: a spare in a different material is a spare that gets used by mistake.
What to Put on the Purchase Order
Six lines make a fitting order that can be filled correctly. The material of body and gasket. The thread standard, named, since NPT and NPSM are not interchangeable. The body size against the measured wall thickness. The fluid, its concentration and its maximum temperature. The quantity, with spares. And the tank the fitting is going into, because that is the only way the supplier can check the wall and the land before shipping.
When the Answer Is a Different Tank, Not a Different Fitting
Some duties do not have a fitting answer. If the wall is too thin to carry the moment from the pipe run, the fix is a reinforced nozzle or a different tank, not a larger gasket. If the temperature is above what the elastomer will take in that chemical, no gasket material will rescue the joint. If the run needs a bore larger than the bulkhead range sensibly covers, the answer is a flange. Recognising those three cases early is worth more than any amount of tightening; it is also the point at which poly tank fittings stop being a catalogue choice.
Selection questions that the tables above do not answer can go to the engineering team with the five measurements from the sizing section. The vessels these fittings go into are described under PP chemical storage tanks, the sheet they are built from under PP plastic sheet, and the wall and support geometry that sets the land around an outlet in vertical poly tank design. Where the duty is aggressive rather than merely awkward, the material side is treated in acid storage tank selection and plastic chemical storage tanks.
Sources
- SIS, Swedish Institute for Standards — SS-EN 12573-4, Welded static non-pressurised thermoplastic tanks — Part 4: Design and calculation of flanged joints (edition 1, approved 19 May 2000, 16 pages, ICS 23.020.10; the same series lists Part 1, General principles, and Part 2, Calculation of vertical cylindrical tanks).
- ASTM International — ASTM F2389-23, Standard Specification for Pressure-rated Polypropylene (PP) Piping Systems (scope covers PP pipe, fittings, valves and manifolds in metric and IPS Schedule 80 sizes; minimum pressure rating 160 psi at 23 °C and 100 psi at 82 °C; section 5.7 permits threads only as metal inserts molded into the fitting, brass or stainless steel, and section 5.7.2 states that threads shall not be molded or fabricated directly into the polypropylene plastic; section 1.3 permits use for industrial process fluids subject to manufacturer confirmation of chemical compatibility).
- ASME — ASME B1.20.1-2013 (R2018), Pipe Threads, General Purpose, Inch (dimensions and gaging of NPT, NPSC, NPTR, NPSM and NPSL threads; reaffirmed 2018, 40 pages; the foundational standard for NPT, NPSC, NPSM and NPSL).
- INEOS Olefins & Polymers USA — Typical Engineering Properties of Polypropylene (density 0.904–0.908 g/cm³ for homopolymer and 0.898–0.900 for impact copolymer; Young’s modulus 1,300 MPa homopolymer and 1,100 MPa copolymer; Poisson’s ratio 0.42; DSC melting 160–165 °C; Vicat softening 152 °C; coefficient of linear thermal expansion 8–10×10−5 cm/(cm·°C) homopolymer and 6–9×10−5 copolymer; water absorption below 0.01%).
- IPEX — EPDM & FKM Chemical Resistance Guide, second edition (2020) (compound compatibility classes 1 high resistance, 2 limited resistance, 3 no resistance, no entry meaning insufficient data; columns at 20, 40, 60, 80, 100 and 120 °C; data obtained from FIP S.p.A. and applying to IPEX valves and fittings using EPDM and FKM seals; ratings quoted here for sodium hydroxide at 10%, 30% and 50%, sulfuric acid at 10% and 50%, and sodium hypochlorite at 3% and 12.5%).
- Trelleborg Sealing Solutions — Materials and Chemical Compatibility Guide (EPDM service range −45 °C to +150 °C with a maximum working temperature of 150 °C; FKM service range −35 °C to +200 °C, with formulated grades to +230 °C and special conditions to 225 °C; volume change after 70 h in IRM-OIL 903, 136% for EPDM against 4% for FKM).
