Reapter

contact us
Location: Home >Q & A

Commissioning a Plate Heat Exchanger: A Field Checklist

Return list Reffer:Reapter
Commissioning a Plate Heat Exchanger: A Field Checklist
Click number:- Release time:2026-10-18

A correctly selected exchanger can still fail in its first week

Most plate heat exchanger problems that show up in the first month of operation are not selection errors. The unit was sized correctly, the plates and gaskets arrived as specified, and the duty is what the drawing says. What goes wrong is commissioning: debris left in the pipework, a plate pack closed without a torque sequence, air trapped in a circuit, a control valve opened fully on a cold unit, and no clean baseline recorded so that nobody can tell a fouling trend from a design margin.

Commissioning is short and cheap. The checks below are the ones that cannot be repeated later: once the pack is closed and the unit insulated, the same inspection costs a shutdown — and on an overseas project, it may mean a technician arriving on a plane.

Checks that must happen before the pack is closed

Everything in this section is invisible after the frame is tightened. It should be signed off while the plates are still on the floor.

  • Plate count and order against the arrangement drawing. Plates look identical in a stack but are not interchangeable: pass partitions, blind plates and the end plates are specific items, and the number of channels per pass must match the drawing. A single plate added or omitted at reassembly changes the duty of every channel.
  • Plate orientation. Consecutive plates are turned over relative to each other so that the corrugations cross and form the contact points that support the pack. Plates stacked the same way up close into each other and produce almost no turbulence.
  • Gasket condition and seating. New gaskets, correct compound for the circuit, seated fully in the groove, and the retention system the plate was designed for. A gasket that is partly out of its groove will be cut on closing.
  • Cleanliness of plate and port surfaces. Grit on a gasket face creates a leak path that no torque will close.
  • Measured plate pack dimension. Record the pack dimension across the tightened stack, against the value on the drawing. This number is the reference for every future re-closure; without it, "is the pack tight enough" becomes an opinion.
  • Pulling space. Confirm the space needed to open the pack and withdraw plates is actually free — it is often taken by pipework installed after the exchanger was set.
  • Instrumentation on both circuits. Pressure gauges and thermometers on inlet and outlet of each side, not one gauge shared between them. Without four temperatures and two pressure drops, commissioning data cannot be interpreted later.

Uniform compression is a sequence, not a single torque figure

Uniform compression is a sequence, not a single torque figure. The closing force comes from the tie bolts and has to be distributed evenly. Tightening one bolt to full torque before its neighbours have been brought up tilts the pressure plate: over-compressed at one corner, under-compressed at the other, weeping at the loose side while gasket life is already spent on the tight side.

The standard method is a diagonal sequence in two or three stages — bring all bolts to roughly a third of the specified torque, then two thirds, then the full value, always working across the frame rather than around it. The torque figure itself is unit-specific: it depends on bolt size, plate size, gasket compound and frame design, and it comes from the manufacturer's drawing for that unit. A torque value copied from another exchanger is not a specification, it is a guess.

Two failure modes are worth naming because both are common and both are self-inflicted. Over-tightening to stop a leak crushes the gasket into permanent set and, on some frame designs, distorts the pressure plate or damages the port liner — the leak gets worse, not better. Under-tightening produces an external weep at the plate edge that looks like a gasket failure but is a clamping failure. Neither is fixed by more torque.

Plate heat exchanger tie bolt tightening sequence: numbered diagonal pattern around the frame in three torque stages, with the measured plate pack dimension recorded against the arrangement drawing before and after closing

Flushing, strainers and start-up filters

A plate heat exchanger has channels a few millimetres wide. It is not a strainer, and treating it as one during construction is the single most common way to block a new unit. Welding slag, mill scale, jointing compound and swarf all arrive with the first flow.

  • Bypass the exchanger for the initial system flush. Flush the pipework on the bypass until the water runs clear, then bring the unit into service.
  • Fit a strainer upstream of each circuit — typically a perforated basket in the 0.5 to 1.0 mm range for water duties — and clean it daily through the commissioning period. On open cooling water, river water or seawater systems, this is not optional.
  • Flush in the normal flow direction, and backflush where the piping allows. Backflushing lifts debris off the contact points instead of driving it deeper into the channel.
  • Keep the strainer in service after handover for the first weeks, then review whether it is still needed. A strainer that is never cleaned is worse than none — it eventually blocks or collapses.

Pressure testing: one side at a time, and watch the other side

Pressure testing proves two things — that the unit contains pressure, and that the circuits are still separate. The second matters most: an internal leak between circuits is a product-contamination event, not a maintenance issue.

Test each circuit separately, with the other side at atmospheric and its ports open and visible. Pressure is normally held at the value required by the applicable code — commonly 1.3 times design pressure — for the code's hold period, with the reading recorded. A weep or drip at the open port of the unpressurised side is an internal leak, and it should be found now, on an accessible unit, rather than after insulation, cladding and handover.

For food, pharmaceutical and other hygienic duties, use test water that will not leave a residue, and drain and dry both circuits afterwards. Water left standing in a stainless channel is the start of a chloride attack story.

The first heat-up: air, rate and control

Air is the most under-estimated problem in a new installation. Trapped in the upper part of a circuit it removes heat transfer area, makes the unit look under-sized, and causes the control loop to hunt as pockets migrate. Vent at the high points of both circuits with the unit running, and repeat once the system has reached temperature.

Bring the unit up to temperature gradually. Opening a control valve fully on a cold exchanger subjects the pack to a step change: the plates respond in seconds, the frame and gasket do not, and the differential expansion goes into the gasket line. Ramp the valve, hold at intermediate conditions, and watch the plate edges as the unit warms — a weep that appears hot and disappears cold is information, not a nuisance.

After the first thermal cycle, follow the manufacturer's instruction on re-checking the pack. Many designs call for a torque or pack-dimension verification once the unit has been hot, because the gasket takes its initial set and the pack relaxes slightly.

Record the clean baseline — it is the reference for every decision afterwards

At stable operating conditions, with the unit clean and the strainers clear, record: flow rate on both circuits, all four temperatures, pressure drop on both circuits, and the valve position. From those, calculate the approach temperature — the difference between the hot-side outlet and the cold-side inlet — and file it with the arrangement drawing.

That baseline turns future conversations from argument into diagnosis. Rising pressure drop at constant flow and temperature is fouling or debris. Falling pressure drop may be a bypass or a changed flow. Unchanged pressure drop with a rising approach temperature points to air, a scaled plate surface, or a changed secondary flow.

Diagnosing the first week of plate heat exchanger operation: pressure drop and approach temperature trends over time, separating normal stabilisation from fouling, trapped air and a bypassing circuit

Common questions

What torque should the tie bolts be tightened to?
The value on the manufacturer's drawing for that unit — it depends on bolt size, plate size, gasket compound and frame design, so there is no universal figure. What generalises is the method: diagonal sequence, two or three stages, and a measured pack dimension to confirm the result.

The unit weeps at the plate edge on start-up. Should we tighten further?
Check the pack dimension and the tightening sequence first. If the pack is at the specified dimension and the weep persists, the cause is usually a gasket out of its groove, a damaged plate edge, or a pack assembled in the wrong order. Adding torque beyond the specification protects nothing and consumes gasket life.

Can the exchanger be in service during the initial system flush?
No. Bypass it, flush the pipework to clear water, then bring it online with strainers fitted. Recovering a channel pack blocked with construction debris means opening the unit — on a new installation, that is a self-inflicted outage.

Why is the pressure drop higher than the design value on day one?
In order of likelihood: air in the circuit, a partially closed or wrongly sized strainer, a valve not fully open, or flow higher than design. Compare measured flow against design before concluding anything about the exchanger itself.

How soon should spare gaskets be ordered?
At the same time as the unit. Gaskets are the maintenance item, they are inexpensive relative to a shutdown, and on an overseas site the lead time — not the price — is what determines how long the unit stays down.

How Reapter supports commissioning, wherever the unit is installed

Every Reapter unit is issued with the documentation commissioning actually needs: general arrangement drawing, plate arrangement chart showing pass configuration and plate order, tightening specification and pack dimension, material certificates, pressure test record and an operation and maintenance manual. Our engineering team also reviews start-up data — four temperatures, both pressure drops and the recorded pack dimension — and returns a written assessment, so a query on site does not wait for a flight.

Where a project requires attendance on site, commissioning support can be arranged. For units already in service — including plate packs originally supplied by other manufacturers — spare plates and gaskets are produced to the installed plate pattern and shipped against the maintenance window, so that a mixed fleet can be supported from one source.

Send us the unit model, the duty and the commissioning date, and we will confirm the documentation package, the spares list and the support arrangement that fits the project. Contact us to request a quotation.

Reapter Engineering Team · Qingdao Ruipute Heat Exchange Equipment Manufacturing Co., Ltd.

Relevant article

Latest information