EMS-Trade
Cooling
Loop consumables

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

Reference sheet · DG-04 · selection and its cost

Four judgements, and the figure each one turns on

Every guide below is written the same way: the rule for choosing a figure, and the second rule for what that choice costs you later. None of the four is about picking the best part. Each one is about picking the figure you can hold for the life of the loop.

Absolute rating 0.05 – 100 µm Interface 1.0 – 15.0 W/m·K Cleanliness ISO 4406 17/15/12 Water µS/cm · pH

Request the specification file All four guides, in one file
A working bench with the four subjects of this sheet on it: two filter cartridges of different rating, a pad and a graphite sheet, a sample bottle, and a conductivity probe in its case
FIG 01 ·

Four subjects, one bench. Each guide on this sheet ends on the figure that decides between two of them.

01

The four guides

Index · rule, then the figure it turns on

Two pleated cartridges side by side on a bench, the pleats of one visibly tighter than the other, with a rule laid along them
01 ·

Rating · absolute against nominal, and where the ladder starts.

A pad and a graphite sheet laid flat on a bench with a steel rule standing against the edge of each to show thickness
02 ·

Interface · bond line against pump-out.

A sample bottle held against a white card with a dark field behind it, the fluid lit through so anything suspended in it shows
03 ·

Cleanliness · the target, and where the sample comes from.

A conductivity probe and a pH probe lying beside a sample bottle on a bench, both with their caps off and their cables coiled
04 ·

Water · the pair read together, never one alone.

01

Rating: choose absolute, and choose it once

A nominal rating describes an average opening; an absolute rating describes the largest particle that gets through. Only one of them can be written into an acceptance condition. The ladder runs 0.05 µm to 100 µm in eight steps, and the loop is normally quoted on two: a coarse stage that takes the load and a fine stage that holds the figure, with 1 µm absolute as the usual fine end.

Turns on 1 µm absolute · eight steps from 0.05 µm to 100 µm · elements in 20 in and 40 in, with 10 in on a small branch

02

Interface: the thickness decides the second year

A quotation states a conductivity figure, and the figure that decides whether the joint still works in year two is the bond line under it. Grease gives the thinnest line and is the first to pump out; a gap pad tolerates the tolerance and pays for it in interface resistance; a phase change material stabilises once it reaches its window; graphite sits at the top of the range and asks for flatness in return.

Turns on 1.0 – 15.0 W/m·K across four classes · grease 1.0 – 6.0 · pad 1.2 – 7.5 · phase change 2.0 – 8.5 · graphite 3.0 – 15.0 W/m·K through plane

03

Cleanliness: one target, three classes, one sample

The figure is written for the moment before the charge, and it is read as three code numbers at the same sampling rather than as an average. Where the sample is drawn decides what it says: a sample from the tank describes fluid that has not been round the loop, and it reads clean earlier than the loop is clean.

Turns on ISO 4406 17/15/12 · all three classes at one sampling · sampled on the return, upstream of the charge connection

04

Water: two readings that only mean something together

Conductivity moves when the inhibitor package is spent or when top-up water has got in; pH moves when the fluid itself has aged. Either one alone sends you to the wrong conclusion, which is why they are read on the same sample as the particle count. Glycol concentration is the third reading and the one most often skipped, and an elastomer responds to it within weeks rather than hours.

Turns on µS/cm and pH read together · glycol held at 20% or under in the facility supply

02

What the first two guides look like on the bench

Guides 01 and 02 · the articles the figures are written about

The rating ladder is a physical thing: eight pleat densities, each one a different opening. Choosing a step is choosing how much surface area you are willing to pay for and how fast the element loads.

The change figure follows the choice: 0.5 – 0.7 bar across the housing, read from the clean-water zero the cart was run to before it shipped.

Eight cartridges of different rating standing in a row on a bench, coarse at one end and fine at the other, each in its own sleeve
FIG 02 ·

The ladder as stock. Two steps of it go into a loop; the other six exist so the two can be chosen rather than inherited.

A loaded cartridge cut lengthways on a bench, the pleats packed on the outer face and the inner face still open, with a clean one beside it
FIG 03 ·

What loading looks like from the inside. The outer face takes it, which is why the differential is read rather than looked at.

The four interface classes, on the same bench. Thickness is the property none of them is quoted on and all of them are judged by.

A grey thermal pad sheet and a red one overlapping on a white ground, both cut square with their edges showing the sheet thickness
Pad · 05Gap filler sheet, 1.2 – 7.5 W/m·K. Tolerates the gap, adds an interface.
Several dark graphite sheets fanned out on a white ground, the edges of the fan showing each sheet's thickness
Graphite · 02Through-plane 3.0 – 15.0 W/m·K. The top of the range, and the flattest surface.
A stack of thermal pads in several colours and thicknesses fanned on a white ground, from a dark grey sheet at the back to a pale blue one at the front
Colour code · 06One colour per compound, so a thickness cannot be mistaken for a rating.
Four pale grey thermal sheets stacked and fanned on a white ground, the top sheet casting a soft shadow on the one below
Sheet · tim-5Bare sheet before die-cutting. Thickness is set here and does not change later.

Every sheet in this strip is a photograph of the article, not a rendering. The dimensions that matter are the ones a rule can be laid against.

03

What the last two guides look like on the bench

Guides 03 and 04 · the sample and the two readings taken from it

A sampling valve on a return line with a bottle under it and a short tail, taken in a plant room with the pipe run behind out of focus
FIG 04 ·

Where the sample comes from. Change this and the code number changes without a thing being done to the loop.

A portable particle counter on a bench with a bottle docked into its top, the readout face angled away and no numbers legible
FIG 05 ·

Read on the spot, at three classes, once. A sample that travels has had time to settle and reads better than the loop is.

A conductivity probe standing in a beaker of clear fluid on a bench with its transmitter beside it and a pH probe laid next to its own beaker
FIG 06 ·

Two readings, one sample. Conductivity without pH tells you the fluid changed and not which way.

A refractometer and a hydrometer on a bench with two sample bottles of coolant beside them, one paler than the other
FIG 07 ·

Two bottles of the same coolant at two concentrations. The elastomer reads the difference even when the loop runs the same.

04

How each of the four is misused

Common misuse · the choice and the cost of making it

Done on the order as written

  • Two stages, coarse then fine, each on its own change rule.
  • Absolute rating written into the acceptance condition, not nominal.
  • Bond line stated with the conductivity figure, on the same line.
  • Cleanliness written as the target at the charge, with the sampling point named.
  • Conductivity and pH read on the same sample, with the glycol figure beside them.

Done short, and what it costs

  • One stage, run to a single rating: it takes the load and passes the fine particles while it does.
  • Nominal rating quoted: the acceptance condition cannot be checked at the bench.
  • Conductivity alone: a figure that satisfies the quotation and fails in the second year.
  • Sample drawn from the tank: clean earlier than the loop is.
  • Glycol concentration assumed from the drum label: the elastomer finds out first.
A single filter housing on a loop with the element half loaded, the differential gauge above it and no second housing anywhere on the run
FIG 08 ·

One stage. It is not wrong until the differential climbs, and then it is wrong for every hour after.

A cold plate lifted off its mounting with the interface material left behind on the plate face, the print of the die outlined where it has pumped out
FIG 09 ·

Pump-out. The figure on the quotation was right and the print of the die is where the material used to be.

A sample bottle filled at a reservoir sight glass rather than at the return, with the tank behind it and the return line running past out of focus
FIG 10 ·

Sampled in the wrong place. The bottle is clean and the loop is not, and both are recorded as passing.

DOCUMENT REQUEST

Request the four guides

The rating ladder, the interface comparison, the cleanliness target and the water chemistry pair, each on one page with the figures behind it.

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.

Two cartridges of different rating standing side by side on a bench with a tag tied to each, the tags blank of print
PL 01 ·

Two options, one decision. The guide is written to make the decision take one reading rather than a meeting.

A record sheet and a pen on a bench beside a sample bottle, the sheet blank of print and the bottle capped
PL 02 ·

The figure that gets written down is the one that can be checked later. Everything on this sheet is written to be that figure.