One word decides who pays for the rock.

Three books, one job

Most international construction runs on one of three standard forms, and they exist because they allocate risk differently. That is the whole point of having three.

The Red Book is for building work designed by the employer. He gives you drawings and a bill, you build what is on them, and he carries the consequences of his own design.

The Yellow Book is for plant and design-build. You design it and you build it, so you carry the design.

The Silver Book is for turnkey work where the employer wants one number, one date and no surprises — typically because he is borrowing against them. The price of that certainty is that a great deal more risk sits with the contractor, and the form says so.

Somebody chose one of those for this job before any of us arrived. Nobody on this register has asked which.

The same number, a different noun

Go to sub-clause 4.12 in each of the three.

In the Red Book it is headed Unforeseeable Physical Conditions. In the Yellow Book it is headed Unforeseeable Physical Conditions. The two forms handle the subject the same way: if you meet physical conditions on site that could not reasonably have been foreseen, you give notice, and you can be entitled to extra time and extra money.

In the Silver Book, sub-clause 4.12 is headed Unforeseeable Difficulties.

One noun changes, and the clause changes with it. It is not a version of the same relief with tighter conditions. It runs the other way entirely: the contractor is treated as having satisfied himself about everything affecting the works before he priced them, and the price is taken to cover the lot. There is no route back for ground that turns out worse than expected, because the form does not contain one.

Same number. Same position in the book. Opposite effect.

What “unforeseeable” actually tests

Even under the Red and Yellow forms, the word does a lot of work, and it is not a synonym for surprising.

The definition sets a standard: conditions that an experienced contractor could not reasonably have foreseen by the base date, given the information available to him at the time. Two parts of that matter enormously and are routinely missed.

It is a fixed date. The test looks at what could have been foreseen when the tender was priced, not what you know now. Everything learned since is irrelevant to whether the condition was foreseeable.

It is an objective standard. Not what your estimator actually thought. What a competent, experienced contractor in that position should have made of the material in front of him.

Now hold that against Week 5. There were five boreholes. Two found rock. All five sat in the southern third of the site, along the strip where the drilling rig could reach without building a temporary road.

An experienced contractor reading that report would foresee rock in the south. It is right there, twice. Whether he could reasonably foresee the conditions under the northern half — where there is no data at all, because the investigation did not go there — is a genuinely different question, and it is the question the money turns on.

That is not a lawyer's point. It is a project controls point, because the answer depends entirely on what was in the tender documents on a particular date, and somebody in your office has that file.

4.12 does not stand alone

Here is the part that gets missed by everybody who reads the clause on its own, and it is why the branch was empty in Week 3.

Sub-clause 4.12 is the third of three consecutive provisions, and the two above it are doing work.

4.10 deals with site data. The employer hands over what he holds, and the contractor is taken to have been responsible for interpreting it.

4.11 deals with the sufficiency of the price. Your tendered figure is taken to cover everything necessary to build the works properly.

Only then does 4.12 open a door, and only for conditions genuinely outside what an experienced contractor could have foreseen.

Read in order, the structure is obvious. Two clauses close the door and the third opens it a crack. A contractor who has read only the third one thinks the ground is shared. A contractor who has read all three knows the default is that it is his, unless he can get through a fairly narrow gap.

And in the Silver Book, 4.11 does the same job and 4.12 does not open anything at all.

What this is worth on our job

The rock is $48,450. The margin on this project is $48,163.

Under a Red or Yellow contract, if the northern conditions are held to be unforeseeable, that money moves to the employer and the job keeps its profit. If they are not, it stays with us and the profit is gone.

Under a Silver contract there is no argument to have. It is ours.

One line on a register — form of contract — is worth the entire margin on this job, and it was never on the register at all.

Now go back to the two boreholes from Week 13. They were worth $5,033 at P80 on the construction cost alone, and $456 worse on expected value. There is a third reason to drill them that has nothing to do with either number.

Because the foreseeability test is fixed at the base date, investigating now cannot make the condition retrospectively foreseeable. What it does is produce evidence, on your own initiative, before the rig arrives — a record of what is actually down there, dated, alongside the tender report that did not cover it. Whatever the contract says about who carries the ground, that file is worth having and it costs $6,800.

Transfer is not a discount

One thing to be clear about, because it is the most common mistake made with this whole subject.

Allocating a risk to somebody else does not reduce it. The rock does not care whose name is on the contract. Transferring it changes who pays, and it always has a price — either one you can see, or one you cannot.

The visible version is a Silver Book job, where a contractor who prices the ground risk properly bids higher than one who does not, and the employer pays for the certainty he asked for. The invisible version is the same job where the contractor does not price it, wins, meets the rock, and then either absorbs it or spends two years arguing about it.

The second version is more common. That is what Week 7's winner's curse looks like when a contract form is involved.

Practical insight

Find out which form your job is on, and which edition. It takes one email and most planners on a project cannot answer it.

Then read three sub-clauses, in this order: 4.10, 4.11, 4.12. Twenty minutes.

Then open the particular conditions and check whether any of the three has been amended, because they very often have, and an amendment to 4.12 is the single most valuable thing in the document for anybody carrying ground risk.

Finally, put one line on your risk register that reads ground conditions worse than the investigation indicates, and in the owner column write the name of whichever party the contract actually gives it to. If that is you, the number goes in your contingency. If it is the employer, the number is still worth writing down, because it is the size of the notice you will one day have to serve.

Key takeaways

✔ Three standard forms exist to allocate risk differently. Which one you are on is a risk register entry.
✔ Sub-clause 4.12 is Unforeseeable Physical Conditions in the Red and Yellow books, and both give a route to time and money.
✔ In the Silver book the same number is Unforeseeable Difficulties, and it removes the relief rather than granting it.
✔ Unforeseeable is tested against an experienced contractor at the base date, not against what your estimator happened to think.
✔ 4.12 sits after 4.10 and 4.11. Two clauses close the door before the third opens it a crack.
✔ Five boreholes in the southern third make rock in the south foreseeable. The north is a different question, and that question is worth $48,450.
✔ Transfer does not reduce risk. It moves who pays, and it always has a price — visible in the tender, or invisible in the final account.

What's coming next

Ground is the risk the contract talks about most explicitly. It is not the only one it allocates, and the others are less obvious because they are spread across different mechanisms that look unrelated.

Some risks buy you time but no money. Some buy money but no time. Some buy neither, and a few buy both. Meanwhile there is a bond sitting against this contract that somebody is paying for every month, and insurance that covers a specific and surprisingly narrow list of things.

Next week we take the remaining thirteen risks and ask, for each one, what the contract actually gives us when it happens — and find that on this job the answer for most of them is nothing at all.