When Should O&M Manuals Start? Not In The Last Three Weeks

When should O&M manuals start?

At the start of the project, not the end. Agree the template with the client team in the first weeks, put the obligation into every subcontract order before it goes out, and request draft content from subcontractors around the midpoint of the programme, while the people who hold the information are still on site.

That is the short answer. Almost nobody disagrees with it in principle, and most projects still do not do it. The reason is not ignorance. It is that none of those steps appear on the programme, so none of them have a date, an owner, or any money attached, and work without those three things does not happen on a construction site.

Why they turn into a scramble

Most O&M manuals become a panic in the last three weeks of a job. It is almost never because the document is difficult. Compiling one is administratively tedious and technically straightforward.

The problem is that nobody owned the information while the job was running.

By the time somebody sits down to assemble the manual, the specialist who commissioned the air conditioning has been off site for four months, the electrical subcontractor has moved two supervisors onto another project, and whoever is compiling it is reconstructing what was installed from drawings that may or may not reflect what actually went in.

Which produces the half of this that people underestimate. A manual assembled in the final fortnight is not only late. It is frequently wrong, because it was written by somebody guessing.

The obligation itself is not in doubt. Standard building contracts require the contractor to hand over as-built information and what is needed to operate and maintain the works before practical completion. It is not optional and it is not new. What varies is whether anybody planned for it.

What is actually in one

Worth stating plainly, because the scramble is partly a scoping problem. A reader who has never had to build one is often surprised how much of it is other people's information rather than their own.

• •

  As-built drawings

• •

  Equipment schedules and asset data

• •

  Manufacturers' literature and operating instructions

• •

  Commissioning records and test certificates

• •

  Statutory certificates, such as electrical installation and gas safety

• •

  Warranties and guarantees, with their start dates

• •

  Planned maintenance requirements for each system

• •

  Spares and supplier contacts

• •

  The record of who installed what

Almost none of that originates with the main contractor. It arrives from subcontractors, suppliers and commissioning engineers, which is exactly why it has to be requested while those people still have a reason to answer you.

The three answers that work on other projects, and not on fit-out

Raise this anywhere and you get the same three responses, all of them correct, and none of them available on a typical fit-out job.

"A good document controller handles this." They do, and on a project that has one it is exactly how it should work. Most fit-out projects do not have one. On a one million pound Cat B job there is no document controller, no information manager and no dedicated closeout resource. It lands on the project manager alongside everything else he is carrying, and it lands late.

"Start collecting six months before completion." Sound advice on a long programme. A great many fit-out jobs run sixteen to twenty weeks start to finish, so six months before completion is a date before the job existed. Any trigger expressed as a fixed number of months out is useless here. It has to be expressed as a proportion of the programme, which is why the request goes at the midpoint.

"Put one person on it from day zero." The gold standard, and it works. The reason you rarely see it is that nobody wants to fund the role. It reads as overhead right up until the point that it is not, and the cost of not having it shows up as three weeks of arguments that never appear on anyone's budget line.

None of those three is wrong. They are answers built for projects with more programme and more people than a fit-out has. The fit-out version has to work with a project manager who is already fully committed, and that changes what the fix can be.

Stop calling it the O&M and call it the closeout plan

The naming matters more than it sounds.

An O&M manual sounds like a document, and documents get produced by somebody at the end. A closeout plan sounds like a plan, and plans get programmed, resourced and reviewed. The same is true of the health and safety file, the as-builts, the certificates, the asset data and the training records. They are all the same job, they all fail for the same reason, and treating them as one deliverable with a sequence behind it is the difference between managing it and discovering it.

The sequence

Ordinary programme activities with ordinary durations.

1. 1.

   Template agreed and approved with the client team early, while there is still time for them to say what they actually want in it. A template signed off in month two costs an hour. A template argued about in the final month costs a rebuild of everything already collated against the wrong structure.

2. 2.

   Draft content requested from subcontractors at the midpoint. At the point where the people who hold the information are still on site and still have an interest in your goodwill.

3. 3.

   Two weeks for the principal contractor to review and comment back. Somebody has to actually read it. If nobody reads the drafts, the gaps surface when they can no longer be fixed.

4. 4.

   Two weeks for subcontractors to amend. A real duration, not a chase email.

5. 5.

   One week to collate.

6. 6.

   Out to the client team, two weeks to comment, draft approved. The step most people leave out entirely, and the one that decides whether the final version gets accepted or sent back.

7. 7.

   Final collation as the as-builts and certificates arrive, one week for client sign off. Ready at handover rather than chased after it.

Roughly eight weeks of activity, most of it waiting rather than working, spread across the second half of the programme. It fits if it is planned. It never fits if it is discovered.

The test of whether it is working is simple. The draft should be signed off around a month before handover. If you are still collating in the final fortnight, it has already gone wrong.

What month one actually buys you

The standard advice is to start compiling on day one. On a fit-out that is not quite right, and the distinction matters.

At month one on most fit-out jobs the majority of trade packages are not let. You cannot request draft content from a subcontractor who has not been appointed. Anybody advising you to start collecting content on day zero has either never run a job where procurement ran alongside construction, or is describing a different kind of project.

What month one genuinely buys you is two things, and neither of them is content.

The first is the structure. The template, the index, the file naming, the format, and who owns each section. Settle it before there is anything to put in it, because that is the only point at which anybody has the attention span for the conversation.

The second is the contractual hook, and that is the one that does the work.

Put the obligation in the subcontract order

Write the requirement into every subcontract order as a condition of payment. Name the format, name the deadline, and tie a proportion of an interim payment to receiving it.

The obvious objection is that you still hold retention. You do. Five per cent is common, half released at practical completion and the balance at the end of the rectification period. Nobody is arguing that the contractor has no leverage at all.

The difficulty is what that leverage reaches. Retention is a hold over the business. It is not access to the individual who knows what was installed. The commercial team want the money released and will chase it hard. The engineer who commissioned the equipment left months earlier and is on another project. So you get a subcontractor who is willing to comply and slow to do it, because the information now has to be reconstructed rather than simply sent.

It costs the main contractor too. A final account that stays open while a manual is argued about is a final account nobody is closing.

Which is why the paragraph in the order should not point at retention. Tie the draft content to an interim valuation, at the point where the package is substantially complete and the people who did the work are still on site. Name the format and the date. You are then collecting information while it still exists in somebody's head, rather than negotiating for it a year later.

Retention is what you fall back on when the sequence has already failed. It should not be the sequence.

The paperwork at the end is rarely a paperwork problem. It is a procurement decision somebody made months earlier without noticing they were making it.

Asset data has the same problem, and worse

Asset registers and equipment data usually sit in the employer's requirements, get quietly assumed to be part of the M&E subcontractor's commissioning package, and are almost never a dated activity with a name against them. So the data arrives at the end as a spreadsheet nobody can verify against what is actually in the ceiling.

Captured progressively as systems go in, with the installer standing next to the unit, it is a few minutes per asset. Done retrospectively it stops being data capture and becomes a survey, and somebody has to pay for that survey.

Does software fix it?

Not on its own, and the question is worth answering because it is the first thing people reach for.

A structured folder with dividers, populated from day one, beats any platform used badly. The tool has never been what fixes this.

What a live system gives you that the folder does not is visibility while there is still time to act. In month four you can see that forty per cent of the content is missing and go and get it. With files you tend to find out when you open them. That is a real advantage, but it is an advantage in seeing the problem, not in solving it. The sequence and the contractual hook still have to exist underneath.

And then it has to survive handover

Getting the manual produced accurately and on time is only half the job. The other half is whether anybody can find it in two years.

They rarely go missing during construction. They go missing afterwards, inside the client's own organisation, and very often at the first change of FM provider. The receiving end has no process for it either, so a document that was compiled properly ends up in a folder nobody inherits, on a server nobody searches, or in a boiler room.

That is a genuine argument for a live system over a set of files and it deserves to be taken seriously. It comes with a caveat that matters more than the format debate, though. A live system populated with guesses is worse than a lever arch full of accurate information, because people trust a screen in a way they never trusted a binder. Nobody double checks a database.

Which lands back in the same place. Whatever the destination format turns out to be, the discipline that fills it has to come first.

What the client can do before appointing anyone

If you are commissioning rather than delivering, there is a thirty second check worth doing.

Ask to see the O&M and closeout activities in the tender programme, with durations, before you appoint.

Most tender programmes show a single bar labelled O&M in the final month, if they show it at all. A programme showing template approval, the midpoint request, the review and amendment cycles and the client comment period is telling you that handover has been planned. One bar is telling you it was added at the end because somebody noticed it was missing.

It costs nothing to look, and it is one of the few things in a tender assessment that a good bid writer cannot retrofit.

Who actually pays when it fails

Both sides, on the same failure, which is probably why it survives.

The client carries it at the worst possible moment. If the information is not ready, practical completion gets qualified, and you end up arguing about whether the building is complete while people are already working in it. The operational risk of running a building without accurate maintenance information does not sit with the contractor once the doors are open.

The contractor carries it too. Retention does not release cleanly against an incomplete handover, the defects period starts with a dispute already in it, and the final account conversation begins on the back foot.

Nobody wins from the scramble. It persists because the cost of preventing it lands in month one on somebody's programme, and the cost of not preventing it lands in the final fortnight on somebody else's.

One clarification worth making

An O&M manual and a CDM health and safety file are not the same document. They are frequently produced by the same people at the same time, and the same discipline fixes both, but they answer to different obligations and contain different things. The health and safety file exists so that whoever works on the building next knows what will hurt them. The O&M manual exists so that whoever operates it knows how it runs.

Confusing the two is common and it produces a document that does neither job properly. That is a subject for its own article.

The short version

The document is rarely the problem. The sequence is.

Agree the structure in month one, put the obligation in the subcontract orders, request the content at the midpoint, and give the review cycles real durations on the programme with a name against each one. If it is not on the programme it does not happen, and by the time anybody notices, half the people who held the information have demobilised.

Related reading