Heat Exchanger Delivery Delays: Root Causes & Prevention (EPC Playbook)

Reader Promise: What You'll Get in This Playbook

If you've run industrial EPC procurement long enough, you've seen the same movie: the heat exchanger is labeled "simple," but it still slips — quietly, then suddenly. The punchline is almost never one big disaster; it's a chain of small misses that turn into a schedule wall.

This playbook is written for EPC buyers who want a repeatable method: freeze inputs early, design gates that match real-world inspection and documents, and stop the "work done but not shippable" trap.

You'll get an EPC-style control system you can copy into your kickoff notes: milestone gates, evidence definitions, and buyer checklists that align engineering, QA/QC, and logistics.

And because we're focused on Shell & Tube, we anchor the lifecycle to the same categories API 660 is built around: design → materials → fabrication → inspection/testing → preparation for shipment. (API)

Why "Simple Exchangers" Still Slip in Industrial EPC Projects

Picture a typical industrial package: the exchanger sits on the critical path — not because it's complex, but because it touches everything. Process defines duty and allowable ΔP, mechanical defines construction, QA defines hold points, and logistics defines what "ready to ship" really means.

When those interfaces don't close in the right order, the project doesn't fail loudly. It fails quietly: a late drawing approval, a missed witness booking, a hydrotest report that arrives after the crate is already built.

Here's the uncomfortable truth: many "delivery delays" happen after fabrication is mostly done. The shop may feel progress. The EPC schedule still slips because the actual release depends on evidence — not effort.

This is why the most effective EPC teams manage exchangers like a controlled pipeline: inputs frozen → approvals closed → hold points booked → test completed → reports issued → packing evidence captured → ship release.

The most common "delay math" (what the schedule feels like)

• The shop says: "We're 90% done."
• EPC reality says: "We're 60% release-ready, because the last 40% is interfaces + evidence."

To remove the surprise, we need a model that treats documentation and hold points as production steps — not admin work.

Quick Answer: What Works Fastest

Prevent heat exchanger delivery delays by doing three things early and ruthlessly:

1. Freeze datasheet inputs and the unit system (no drifting SI ↔ US customary midstream). TEMA explicitly highlights that a single unit system should be used for all aspects of design/fabrication, and notes updates to its specification sheet to better define design requirements — exactly the kind of detail that avoids redesign loops. (TEMA)
2. Plan ITP hold/witness points at kickoff (book them like you book cranes — availability is real).
3. Tie hydrotest + report issuance + packing/preservation evidence to one release-to-ship gate (ship release is a decision, not a feeling).

This playbook assumes an industrial EPC environment where schedule reliability comes from gates and evidence, not "percent complete."

The "release-to-ship" gate definition (a practical EPC sentence)

A heat exchanger is not ready when it is "fabricated." It is ready when it is tested, documented, preserved, packed, and traceable — and the release package is issued.

The 6 Delay Patterns EPC Sees Repeatedly (Shell & Tube Focus)

Below are the six patterns that show up across industries — refinery, chemical, energy, general industrial. You can think of them as repeatable failure modes that appear when procurement treats exchangers as "routine."

Each pattern includes: how it starts → how it becomes a delay → early warning signals → prevention move.

Pattern 1 — Datasheet "TBD" Inputs That Never Truly Close

It starts harmlessly: duty "TBD," fouling "TBD," corrosion allowance "TBD," metallurgy "TBD." Everyone assumes the TBDs will be closed "soon," so the vendor progresses on assumptions.

Then a late change arrives — often a process refinement or a site condition — and suddenly you're not changing one line; you're changing nozzle sizes, bundle layout, thickness, sometimes even material procurement.

Early warnings

• Vendor asks "confirm later" on key thermal inputs
• Too many assumptions in the calculation note
• Long gaps between "TBD list" updates

Prevention move

• Set a datasheet freeze date and treat it as Gate #1 (more in Section 6 later).

Pattern 2 — Unit System Mismatch (SI vs US Customary Drift)

This one is sneaky because it rarely announces itself. Someone enters allowable ΔP in kPa, someone else reads it as psi. Someone uses °C, someone else assumes °F. Over time, the project becomes bilingual in the worst way.

TEMA explicitly addresses the unit-system issue: it provides both unit systems in its materials, but calls for a single system to be used across design and fabrication (except where permitted). (TEMA)

Early warnings

• Datasheets show mixed units without a controlling statement
• Vendor calculations reference a different unit basis than the datasheet
• Approval comments include "please reconfirm units" (a very expensive sentence)

Prevention move

• Add one line to every datasheet and kickoff minutes: "Controlling unit system: SI only (or US customary only). No mixed units." (TEMA)

Pattern 3 — Drawing Approval Loops (IFA/AFC Churn)

Exchanger drawings are interface drawings: nozzle orientation, saddle locations, bundle pulling clearances, maintenance access. In industrial EPC projects, these drawings attract comments from many parties.

If EPC doesn't enforce a rule for comment scope, you get "review creep": every revision becomes a chance to redesign. The vendor waits, fabrication waits, the schedule bleeds.

Early warnings

• More than 2–3 IFA cycles without a firm AFC date
• Late comments that are "nice-to-have," not safety or interface critical
• "Hold fabrication until…" language without a schedule impact decision

Prevention move

• Create an AFC rule: Interface-critical comments only after Rev 1; everything else goes to "next revision" unless it affects fit, safety, or code.

Pattern 4 — ITP Hold Points Not Booked (Witness Delay)

Your vendor may be ready for NDE or hydrotest — but your witness isn't. In industrial projects, third-party inspectors, client reps, and class/authority witnesses have real calendars.

So the exchanger sits ready, occupying space, waiting for the witness slot. The delay looks "small" (days), but it lands at the worst possible time: right before shipping.

Early warnings

• ITP exists, but witness dates are "TBD"
• Vendor asks for witness with short notice
• Multiple packages competing for the same inspectors

Prevention move

• At kickoff, convert hold points into a booking plan with windows and responsible owners.

Pattern 5 — Hydrotest Completed, But Report Issued Late

This is the classic EPC frustration: the shop says "hydrotest done," but EPC cannot release shipment because the evidence package isn't closed. The work is finished; the milestone isn't.

In API 660-style lifecycle thinking, inspection/testing is not complete until it is verified and closed out — your schedule must treat "report issued" as the event. (API)

Early warnings

• Reports lag behind shop activities by a week or more
• Document register doesn't show due dates for test records
• QA team is overloaded at the end of the job

Prevention move

• Define the gate as: Hydrotest complete + Hydrotest report issued + NCR closure (if any).

Pattern 6 — Packing/Preservation Missed (Receiving Quarantine + Rework)

The exchanger arrives — but the receiving team flags it: missing flange protectors, moisture ingress, rust spots, damaged nozzle caps. Now it's quarantined, cleaned, re-preserved, sometimes re-inspected.

This becomes an EPC delay because site cannot install what it cannot accept. And in many industrial sites, "receiving acceptance" is a formal process.

Early warnings

• No preservation checklist in the VDB
• No packing photos or records
• Export packing treated as "logistics only," not QA evidence

Prevention move

• Make packing/preservation evidence a ship-release condition (photos + checklist + materials used).

Standards Reality Check: What Standards Solve vs What EPC Still Must Control

Standards are powerful — but they don't manage your project for you. They define minimum technical expectations; EPC still owns the interfaces and decision gates.

API 660 (Shell & Tube) — The lifecycle backbone

API 660 specifies requirements and recommendations for mechanical design, material selection, fabrication, inspection, testing, and preparation for shipment of shell-and-tube exchangers. (API)

That scope is exactly why API 660 aligns well with an EPC gate model: the project naturally moves from design → fabrication → inspection/testing → shipment readiness. (API)

TEMA — "Small details" that prevent redesign loops

TEMA highlights two EPC-critical ideas in one place:

• It updated its heat exchanger specification sheet to better define design requirements, helping reduce ambiguity. (TEMA)
• It provides both unit systems but calls for a single unit system to be used across design/fabrication (with limited exceptions). (TEMA)

In practice, this is the difference between a stable datasheet and a moving target.

IOGP JIP33 (Optional, but very useful in industrial EPC)

If your customer base includes large operators or global EPC frameworks, JIP33 is worth referencing as a mindset: standardizing procurement expectations through supplementary requirements, information requirements, and quality requirements.

For shell & tube exchangers, IOGP S-614 is explicitly positioned as supplementary requirements to API 660, and it is available through IOGP's library. (IOGP)

This matters because it formalizes what experienced buyers already do: define exactly what "done" means, including evidence.

The Delay-Chain Model: A Visual Your Team Can Reuse

Most EPC teams track fabrication progress. High-performing EPC teams track closure readiness.

Here's the simplest way to teach it internally: delays happen when the chain breaks at inputs, approvals, hold points, evidence issuance, or preservation — not just at welding.

Gate-to-Evidence Mapping (Paste into Kickoff Minutes)

Gate	What "Done" Means	Evidence (minimum)	Owner (EPC side)
Gate 1	Datasheet frozen + unit system locked	Signed datasheet + assumption list	Package engineer
Gate 2	AFC drawings issued	AFC GA/nozzle orientation	Lead engineer
Gate 3	ITP hold points agreed + witness windows booked	Approved ITP + witness plan	QA/QC + Expeditor
Gate 4	Hydrotest completed	Test record reference	Vendor QA
Gate 4b	Reports issued (release reality)	Hydrotest report + NCR closure	Doc control
Gate 5	Preservation + packing evidence complete	Photos + checklist + packing list	Logistics + QA
Gate 6	Release-to-ship approved	Release note + VDB index	EPC PM

Expediting Cadence That Actually Works (The EPC Project Pulse)

In an industrial EPC project, expediting can easily become a source of frustration when it is treated like a vague "get things done" task. Instead, effective expediting is about keeping track of gates and evidence, not just chasing % completion.

By linking progress tracking to the evidence-based system outlined in this playbook, you'll cut down on uncertainty and unnecessary firefighting, ensuring the project moves smoothly from gate to gate. This section is about how to set the expediting cadence that stops small delays from accumulating into major bottlenecks.

Weekly Expediting Agenda: Gate-Focused, Not Percent Complete

When you run an EPC project, your weekly meetings should revolve around confirming whether the gates are still on track. If you're simply looking at a "% complete" metric, you miss the risk of late-stage surprises. Instead, ask yourself:

• Are all milestone gates defined and actively being managed?
• Are hold points and evidence issuance being tracked to avoid last-minute rushes?

Core questions to ask during weekly expediting:

1. Gate 1 (datasheet freeze): Is the datasheet truly locked? Are there no open TBDs?
2. Gate 2 (AFC drawings): Have all interfaces been closed? No revision churn allowed.
3. Gate 3 (witness points booked): Are witness dates confirmed in writing?
4. Gate 4 (hydrotest completed + report issued): Is the hydrotest report issued or still missing?
5. Gate 5 (packing/preservation evidence): Is preservation documented with evidence?
6. Gate 6 (release-to-ship decision): Are all documents in place? Is everything ready for release?

One-Page Dashboard: The EPC Control Center

To make expediting effective, set up a dashboard to track each gate. The dashboard should give you a snapshot of where the project stands based on document readiness (not just physical progress). You'll rely on this to move the project forward efficiently.

A simple status dashboard for tracking progress against the gates (you can use Excel or any project management tool that supports task assignment and due dates).

EPC Gate Tracker Dashboard (Copy-Paste Template)

Gate	Milestone	Target Date	Owner	Status	Key Issues
G1	Datasheet Freeze	[Date]	Package Eng	Open	TBDs not finalized
G2	AFC Drawings	[Date]	Lead Eng	Closed	No issues
G3	ITP + Witness Booked	[Date]	QA/QC	Open	Witness dates TBD
G4	Hydrotest Complete	[Date]	Vendor QA	Open	Report not issued
G5	Packing Evidence	[Date]	Logistics	Closed	Preservation photos ready
G6	Release-to-Ship	[Date]	EPC PM	Open	Documentation not complete

Escalation Ladder: Turning Small Issues into Big Ones

Every EPC project has its bottlenecks. The trick is to get ahead of them before they derail your timeline. Here's how you can escalate issues systematically:

1. Vendor PM or Package Manager: First point of contact. Confirm if the work is getting done or if there are blockers.
   • Escalate to Expediting Team if they're not addressing issues.
   • Check that critical documents are closed at each gate.
2. QA/QC Lead: Often responsible for approving critical quality documentation, inspections, and hydrotest completion.
   • Escalate to Vendor QA if documentation or testing is lagging.
   • Ensure proper NDE inspections are scheduled ahead of time to avoid any delays at later stages.
3. EPC Project Manager (PM): If the vendor is stuck or if delays are related to documentation or scheduling, this is when you bring in senior management.
   • Escalate to Client/Owner's Representative if they have a direct impact on schedules.
   • Ensure any site logistics and receiving acceptance issues are identified early.

Key Performance Metrics for Expediting

An effective expediting system relies on metrics that help you measure the "health" of the project in terms of gates, documents, and key approvals. Track:

1. Document Due Dates: Number of documents overdue across each gate.
2. Witness Booking Lead Time: Number of days before the witness booking window closes.
3. Test Report Issuance Lag: Number of days between when tests are completed and when reports are issued.
4. Packing Readiness: Percent of items that have shipping documentation in place, with photos.

KPI Dashboard Snapshot (for internal team review)

KPI	Target	Actual	Variance	Action Needed
% overdue documents	0%	7%	7%	Expedite Gate 4 report
Witness date booked	100%	80%	-20%	Confirm availability ASAP
Test report lag	0 days	5 days	+5	Follow-up with vendor QA
Packing readiness	100%	90%	-10%	Check packing documentation and photos

Contract & PO Clauses That Reduce Delays (Aligning to Gates)

The final piece of the puzzle is ensuring your contractual framework aligns with the gates and evidence process. Standardizing your payment terms, witness costs, and responsibilities can provide both clarity and leverage for expediting and release.

Milestone-Based Payment Clauses (Aligned to Gates)

In industrial projects, payment milestones should be aligned to key gates — not arbitrary progress percentages. This keeps vendors focused on delivering the right evidence at each gate, which in turn reduces delays.

Example PO Clause for Milestone Payments:

• Gate 1: 10% of contract value upon datasheet freeze and unit system confirmation.
• Gate 2: 15% of contract value upon AFC drawings issued and all interfaces closed.
• Gate 3: 20% of contract value once ITP hold points are booked.
• Gate 4: 25% of contract value after hydrotest completed and report issued.
• Gate 5: 10% of contract value upon packing evidence provided.
• Gate 6: 20% of contract value upon release-to-ship approved.

"Ready for Shipment" Definition (Ship-Release Clauses)

Ensure that the contract clearly defines what constitutes "ready for shipment." A lack of clarity here can lead to vendor confusion and EPC schedule stress. For example:

"Ready for shipment" is conditional upon:

• Hydrotest complete and hydrotest report issued.
• Packing and preservation evidence fully documented.
• Vendor Data Book (VDB) index updated and closed with transmittal.

This creates a clear, actionable definition of what needs to be done, not just what's in progress.

Witness Cost Responsibility Clause (Avoid Delays in Scheduling)

In many EPC projects, witness costs or delays can become a burden. Make it clear who will bear the costs of re-tests or delayed inspections.

Clear Responsibilities for Rework and NCRs

You need a clear clause about rework responsibility. Rework can often be avoided by early detection of issues, so ensure that there's a clause in place that limits the need for rework, especially for critical items like NDE and hydrotest.

Optional: IOGP JIP33 S-614 Overlay (For Stronger Requirements)

For projects requiring stronger procurement discipline, JIP33 S-614 offers detailed procurement guidance for shell-and-tube exchangers, including specific standards and documentation. It is particularly useful for large operators and projects with high scrutiny on vendor documentation.

Field Resources for Your Team