EMS-Trade
Cooling
Loop consumables

Document ECT-01 · REV C · ISSUED 2026-09 Six lines · 48 sheets

Note · NB-02 · seal reliability

Eight ways a loop seal degrades before it drips

A loop seal almost never fails as a burst. It is lost slowly, through eight mechanisms that each leave a different mark on the seat and each open in a different window. Seven of the eight are invisible from outside the joint, which is why a seal that has been weeping for a month still reads as healthy.

Service range −40 ~ +99 °C Working pressure 1.38 MPa Burst ≥ 5.5 MPa Cleanliness ISO 4406 17/15/12

Request this note as a sheet Section drawing, eight mechanisms, one schedule
A metal seal face photographed close in at an angle, a black elastomer ring lying in its groove with the contact band polished bright along one side
FIG 1 ·

One seat, one ring, one bright band. The band is where the joint is actually sealing, and it is the only part of the assembly that has to be right.

01

A seal is lost long before it drips

Claim · one sentence, and how to check it

The claim is narrow enough to be argued with. A joint that drips has been degrading since the first thermal cycle, and by the time there is a drop under it the mechanism has been readable on the seat for months. The drop is not the failure. It is the last of eight mechanisms finishing its work.

The consequence is that a seal is not maintained by watching for drips. It is maintained by reading the seat at a window that was going to happen anyway: the one in which a coupling is opened, a rack is moved, or a filter is changed. Nothing here needs new access or a longer outage.

Check it this way: open one joint that has been in service across a full season and photograph the contact band. If the band is bright and continuous, the joint was healthy and the photograph is your baseline.

A black elastomer ring under a bench loupe, the ring held flat on a white card with the glass over one side of its circumference
FIG 2 ·

At bench magnification the contact band reads as a flat, slightly glossy track.

A stainless coupling in a pipe run wrapped with a dry paper band tied round the joint, the paper clean and unmarked under a raking light
FIG 3 ·

A dry paper band is still the cheapest moisture witness on a loop, and it dates the onset for free.

A machined seal seat lit from one side so the contact track shows as a ring of fine scratches against the turned finish of the metal
FIG 4 ·

The turned finish is the datum. Anything polished across it is the seal doing its work.

A shallow stainless tray under a row of valve bodies, one dried ring mark on the tray floor where a joint above it has wept and stopped
FIG 5 ·

A dried ring mark with clean metal around it is a joint that sealed itself again. That is the mechanism that hid the problem.

02

What is happening inside the groove

Mechanism · the elastomer has one job and eight ways to stop doing it

1 2 3 4 5 Fluid side Extrusion gap Atmosphere side Body, upper half Body, lower half

Key

  1. 1 ·The groove. Its depth sets the squeeze the ring is given, and it does not change over the life of the joint.
  2. 2 ·The elastomer cross-section. Every mechanism here acts on the same small volume of material.
  3. 3 ·The contact band on the metal face. Bright and unbroken is healthy; matt or fretted is not.
  4. 4 ·The bore side. Coolant reaches the ring here, so chemical attack shows on this face first.
  5. 5 ·The extrusion gap. Working pressure is 1.38 MPa and burst is above 5.5 MPa; the gap is what the elastomer bridges at both.
A gland cut through the middle on a bench with the two cut faces set together, showing the rectangular groove and the ring sitting in it at working squeeze
FIG 6 ·

Cut lengthways, the groove and the ring are one shape. None of it is visible from outside the joint.

A seat face under raking light with a single unbroken bright band running right round it, the rest of the face still showing its turned finish
FIG 7 ·

A healthy band is narrow, unbroken and slightly glossy. A band that has gone matt and wide has stopped concentrating load.

The drawing is a section, not a scale model: proportions are set so the groove, the ring and the gap read together at page width. The dimensions a purchase turns on are in the schedule below.

03

The eight mechanisms, one schedule

Schedule · the signature you can see, and the window it opens in

M1 ·

Compression set

The ring takes a flat where it was squeezed and stops pushing back. Re-torquing helps once, then stops helping, which is the point at which most joints are first noticed.

From the first thermal cycles

M2 ·

Thermal hardening

The cover loses elongation and crazes along the contact band. The ring reads stiff rather than springy between gloved fingers, and it no longer follows the metal as the joint cycles.

At the top of the range

M3 ·

Swell and shrinkage

Volume follows glycol concentration. A swollen ring grips the groove and will not come out clean; a shrunk ring has already lost its squeeze and cannot be re-used.

Within weeks of the change

M4 ·

Attack from a spent package

Surface crazing and a colour shift, always on the fluid side of the ring. By the time this shows, the inhibitor package holding the fluid back has already gone.

From the third year

M5 ·

Particle embedment

Solids press into the elastomer and sit proud of the contact band, lifting the sealing line off the metal in local spots. The count that matters is the one at the joint, not at the tank.

As the count rises

M6 ·

Extrusion into the gap

A nibbled edge on the low-pressure side, matching the clearance it was pushed into. The one mechanism driven by pressure rather than by time.

Under sustained pressure

M7 ·

Spiral twist

A helix in the ring that was not there when it was fitted, from an installation that rolled the ring in instead of laying it in. Fitting damage, not service damage.

Invisible until opened

M8 ·

Seat damage carried in

A nick, a tool mark or a scratch on the metal that prints itself onto every replacement ring. Replacing the ring without dressing the seat reproduces the leak on the same schedule.

Immediately, and permanent

Eight elastomer rings laid out in a row on a bench, each showing a different kind of damage: a flat, a crazed band, a swollen ring, a nibbled edge and a scratched one
FIG 8 ·

Eight rings, eight signatures. Read as a set, the schedule stops being a list of failure modes and becomes a way of sorting what came out of a joint.

04

The parts this note is written against

Supplier parts · photographed on the mount they arrive on

A single black elastomer seal ring standing on a white ground, its section uniformly round and unmarked
SP 1 · Seal ring, as supplied Round section · fresh · the datum for set and hardening
A group of red elastomer seal rings laid on a white ground, one ring resting across the others so its section is seen against the light
SP 2 · Ring, close group Same section, read against a second colour
Seven green elastomer seal rings of one size arranged in a ring on a grey cloth ground, each section catching the light along its top
SP 3 · Ring set, one size Batch consistency · checked against the schedule
A loose pile of black elastomer seal rings on a white ground, the rings overlapping so several sections are visible at once
SP 4 · Ring, loose stock Handled stock · the M7 risk starts here
A black rubber profile gasket on a white ground, a long straight extrusion with two circular eyes and moulded clip feet along one edge
SP 5 ·

A profile gasket works differently from a ring: it seals on a moulded lip, so compression set appears as lost lip height rather than as a flat.

A white silicone sealing frame with six rectangular apertures lying flat on a dark bench top, one corner of the frame slightly lifted
SP 6 ·

A frame seals along its whole perimeter at once, so one local flat opens a path. Its inspection is the perimeter, not four edges.

Six parts from the six lines on this site, held as the physical reference for the schedule above. They are the subjects of the figures, not a range: fittings and seals schedules what is held.

05

The figures that decide it

Schedule · value, unit and the condition it holds under

Seal figures, by what they decide
Item Figure Value Unit Condition it holds under
01 Service temperatureWhole joint −40 ~ +99 °C The narrower of the two published ranges. The supplier range reaches higher; the loop figure is the one a rack sees.
02 Working pressureAt the joint 1.38 MPa About 200 psi. This is the figure that drives extrusion into the gap, mechanism M6.
03 Burst, minimumAssembly ≥ 5.5 MPa Tested as crimped, not as a bare hose length. It is the ceiling the seal is never meant to see.
04 Cleanliness at the jointLoop target ISO 4406 17/15/12 The particle figure behind M5. Counts taken at the tank flatter the number the seal actually sees.
05 Fine filtration stageElement 1 µm Absolute rating, on the loop side of the joint. The ladder runs to 100 µm at the coarse end.
06 Flame rating of the coverHose UL 94 V-0 On the outer cover of the hose the seal sits in, not on the elastomer itself.
07 Material certificateMetal parts EN 10204 3.1 Traceable to the heat the bar came from, which is what makes a seat replacement checkable.
08 Glycol in the facility supplyFluid 20 % At or under this figure for a water based loop. Concentration error is mechanism M3, and it shows within weeks.

Eight figures, each written with the condition it holds under. A value without its condition is not a specification.

A record card beside a disassembled coupling on a bench, the card ruled into rows and a ring lying on the card where a row would be filled in
FIG 9 ·

The eight mechanisms are only useful as a schedule if the joint they came out of is written down beside them.

06

What to do at the next window

Practice · five actions, all inside an outage you were taking anyway

STEP ONE Dress the seat first Replace the ring only after the metal is clean
STEP TWO Lay the ring in, never roll it Ends M7 before it starts
STEP THREE Tie a dry band on the joint Dates the onset for free
STEP FOUR Sample at the joint, not the tank The count that drives embedment
STEP FIVE Photograph the contact band Your baseline, on your own loop

Measure the groove, not the ring

A worn groove changes the squeeze every replacement ring is given. The ring is the part you can buy; the groove is the part that decides whether buying it helps.

Keep one joint as a reference

The same joint, opened every time, photographed the same way. One reference joint tells you more about a loop than eight random ones spread across it.

A hand tool laid along a seal seat on a bench with a clean lint-free cloth beside it, the seat wiped back to bare metal
FIG 10 ·

Dressing the seat is a two-minute action, and it decides whether the next ring lasts as long as the last one did.

A ring being lowered flat onto a seal groove on a bench, both thumbs holding it level so no part of it twists into the groove
FIG 11 ·

Level and even. A ring that goes in flat cannot carry a helix into service.

A coupling on a pipe run with a blank paper band tied round the joint and a pencil mark on the pipe beside it
FIG 12 ·

The band stays on until the joint is opened. Its condition when it comes off is the only free history a loop gives you.

07

What this note does not cover

Boundary · the three questions this schedule cannot answer

Not a lifetime figure
Nothing here predicts how long a seal lasts on your loop. The mechanisms and their windows are the judgement; the interval is set by your own reference joint, and it is usually shorter than the maintenance plan assumes.
Not an elastomer choice
The schedule does not say which compound to use. That is set by the fluid, the glycol concentration and the temperature the loop holds, and it is written on the material certificate rather than invented here.
Not an installation procedure
Installation and commissioning are carried out by the contractor holding the work. This note describes what to look at and what the marks mean; it is not an instruction to open a joint.
A bundle of ruled documents and a folded drawing on a bench with a single seal ring resting on top of the stack
FIG 13 ·

A supplier hands over the schedule, the drawing and the certificate. The loop supplies the history.

DOCUMENT REQUEST

Request this note as a sheet

The eight mechanisms, the seal section drawing, the eight-figure schedule and the material certificate that goes with the metal parts.

If the form cannot send, this address takes the request instead: sales@ems-cool.com
The only promise on this page

One business day, with the file attached.

Nothing else on this sheet is promised. Price, stock and lead time are quoted against a request; the mechanisms above are the position, not a warranty on somebody else's joint.

A row of seal rings of increasing size on a bench beside a folded section drawing, each ring sitting in its own shallow recess
PL 1 ·

Rings are quoted by section and bore, so the drawing is what makes a request answerable.

A bench cleared for seal work with a clean cloth, a loupe, a marking pencil and two labelled sample pots lined up along the back edge
PL 2 ·

Nothing on this bench is special. The point of the note is that the bench does not need to be.