From ITP to Data Book

Every fabrication contract contains the same implicit question: how will the buyer know the item was built and tested as specified? The Inspection and Test Plan is the industry's answer. It lists, in production sequence, every quality intervention on the scope — from material receipt to final release — and pins down four things per activity: the governing procedure, the acceptance criteria, the record produced, and the involvement of each party.

The ITP is agreed and signed before fabrication starts. That timing is the point: it converts quality from an end-of-job argument into a step-by-step contract that all parties executed together.

A typical line reads like this: activity "Hydrostatic test" · procedure "HT-PRO-004 per ASME VIII UG-99" · acceptance criteria "no leakage, no visible deformation" · record "hydrotest certificate" · manufacturer H, client W, notified body H. One row, and everyone knows what happens at that step, what paper it produces, and who has to be in the room.

Multiply by thirty to eighty lines — material receipt, fit-up checks, welding qualification verification, in-process NDT, PWHT, dimensional control, coating, final inspection — and you have the complete quality story of the item, in advance.

The codes are assigned per party, per line. A hydrotest may be a hold point for the notified body but only a witness point for the client; splitting the difference correctly is what makes an ITP negotiation. The golden rule on the shop floor: nothing irreversible happens on a line carrying an unreleased H.

Most ITPs carry three signature columns: the manufacturer (who drafts the plan and performs the work), the client or EPC (who marks its intervention points and accepts the plan), and where applicable the third party / notified body — mandatory under PED conformity assessment modules with production surveillance, common under client-appointed TPI in oil & gas. Each executed line collects initials and dates as the job proceeds; the fully signed-off ITP is itself a record that enters the data book.

Walk the approved ITP top to bottom and note the "record" column: mill certificates reviewed at receipt (EN 10204 3.1/3.2), fit-up reports, weld register entries, NDT reports, PWHT charts, dimensional reports, test certificates, coating records, release notes. That list, in that order, is 80% of your MRB index.

Experienced document controllers therefore build the data book structure from the ITP on day one — every ITP line becomes a placeholder in the book, and compilation becomes a matter of filling slots as records are produced, instead of reverse-engineering a pile of paper at the end. It is the single highest-leverage habit in creating an MDB.

The ITP fails in execution, not in drafting. Drift looks like this: a witness point waived verbally but never documented; an NDT extent increased by an NCR but the ITP never revised; a sub-supplier working to the previous revision. Each gap surfaces months later as a mismatch between what the ITP promises and what the data book contains.

Three habits prevent nearly all of it: treat the ITP as a controlled document (every change is a numbered revision), document every waiver in writing the day it happens, and keep the record-collection status visible per line — so "which records are still missing?" is a lookup, not an investigation. The same discipline applies to collecting supplier documents against their own sub-ITPs.

• Copy-paste ITPs — plans inherited from a previous job whose lines don't match this scope, spotted immediately by good reviewers.
• Notice periods ignored — inviting the TPI two days before a hold point when the contract says ten working days.
• Records that don't match the line — the ITP promises a "dimensional report per ISO 13920" and the book contains an unreferenced sketch.
• Sub-supplier scopes without ITP coverage — purchased items arriving with no agreed inspection level and no records.
• The unsigned ITP — executing against a draft; at review, an unapproved plan makes every downstream record contestable.

Because the ITP maps one-to-one onto the data book, it is the natural backbone for a digital workflow: import the approved ITP, generate the book structure from its lines, attach each record to its line as it is produced, and let completion status roll up in real time. At any moment, the answer to "are we ready for handover?" is the percentage of ITP lines with accepted records — visible to the manufacturer, the client and the inspector alike. That is exactly the workflow a purpose-built data book platform automates.