Interim Valuations, Variations and the Cost of Delay

An interim valuation is a certificate, not an invoice

A payment during a construction contract is not a bill for work done. It is the value of everything completed TO DATE, assessed at contract rates, less the deductions the contract specifies, less everything already certified — so each valuation is a cumulative assessment with the previous one subtracted, rather than a statement of the period's work.

That structure matters when anything is corrected. An error in one valuation does not need a credit note; it is absorbed by the next cumulative assessment, because the arithmetic re-states the total every time.

RETENTION is the deduction that surprises people outside the industry. A percentage of the value is held back — conventionally a few per cent — and released in two stages: half at practical completion, half at the end of the defects period, which can be a year or more later. Until then the contractor has performed the work, paid for the materials and the labour, and is financing that held-back sum.

That financing cost is real and is rarely priced explicitly. A small contractor carrying retention across several projects can be profitable on paper and short of cash, which is the mechanism behind a great deal of construction insolvency — and it is why retention reform and project bank accounts keep appearing as policy proposals.

What counts as done is where valuations are argued

The arithmetic is simple and the inputs are contested. Three categories behave differently and each has its own conditions.

MEASURED WORK is work physically complete, valued at the rates in the contract. Partially complete work is valued as a proportion, and how that proportion is assessed — by measurement, by milestone, by the contractor's judgement — is set by the contract rather than by convention.

MATERIALS ON SITE are paid for before they are installed, because otherwise the contractor finances the whole supply chain. The contract usually attaches conditions: the materials must be properly stored and protected, they must be intended for the works, and title must have passed — which is not automatic, since a supplier's retention-of-title clause can mean the contractor does not own what has been paid for.

MATERIALS OFF SITE are the most restricted, because paying for something in a supplier's warehouse gives the employer nothing if the contractor fails. Where the contract admits them at all it usually requires them to be identified, insured, marked as the employer's property, and covered by a bond.

None of this is arithmetic. The calculator applies the contract's percentages to the values entered; whether a particular item belongs in the valuation at all is a question for the contract administrator.

A change costs more the later it arrives

The direct cost of a variation — the extra materials and labour for the new work — is frequently the smaller part of what it actually costs, and the gap widens sharply with time.

Early in a project a change is a drawing revision. Later it means undoing work already built, re-procuring items already ordered, re-sequencing trades who were about to start, and absorbing the delay to everything that follows. None of that appears in a measurement of the changed work.

The commercial handling reflects this. A variation is valued at contract rates where the work is similar and carried out in similar conditions; where it is NOT similar — because it is out of sequence, in a confined area, or at a different time of year — the contract generally allows a fair valuation instead, and that is precisely the provision the disruption argument runs through.

Which is why an omission is not simply a credit at the original rate. Removing work leaves the preliminaries largely intact, may make the remaining work less efficient, and can strand materials already bought. A client asking for a straight rate-for-rate credit on a late omission is asking for a number the cost structure does not support, for the same reason set out in the estimating paper.

Prolongation and disruption are different claims

These two are routinely treated as one thing, and the confusion is the single commonest reason a delay claim is rejected.

PROLONGATION is the cost of the site being open for longer. It is the time-related preliminaries — supervision, welfare, plant standing, scaffold hire, insurance — for the extended period, and it depends on establishing that the completion date moved and why. It is computed from the extension of time, which is a separate determination from the money.

DISRUPTION is the loss of productivity caused by working less efficiently, and it does not require the project to have finished late at all. A contractor can finish on time, having absorbed disruption by adding resource, and still have a legitimate claim for the extra resource. The two are independent, and a claim that asks for prolongation when the facts support disruption is asking for the wrong thing.

Disruption is also much harder to prove, because it is a comparison against what productivity WOULD have been. The credible methods compare a disrupted period against an undisrupted one on the same project, or against a measured baseline; a global claim that simply attributes all the overspend to the other party's actions is the form that fails most often.

Underneath both sits the distinction between an extension of time and money. An extension relieves the contractor of damages for late completion; whether it also carries cost depends on which event caused the delay. Some events give time but not money, which is a contractual allocation rather than an oversight.

Acceleration is bought at a poor exchange rate

Recovering time costs more than the time is worth on a straight-line basis, because every mechanism for going faster reduces productivity per unit of resource.

Sustained OVERTIME is the clearest case: output per hour falls as hours per week rise, through fatigue and absenteeism, and the studies that measure it find the effect compounding over consecutive weeks rather than stabilising. Several weeks of sustained overtime can approach the point where the extra hours deliver very little.

Adding RESOURCE to the same work face has its own ceiling. There is only so much room, so many access points and so much plant, and beyond a density the crews interfere with each other — so doubling the labour on a congested area does not halve its duration, and past a point makes it worse.

Extra SHIFTS avoid congestion but bring their own penalties: lower productivity on night work, supervision and lighting costs, and the handover time lost at each shift change.

The practical conclusion is that an acceleration decision should be taken against a measured comparison — the cost of the acceleration against the cost of the delay it avoids — and that the acceleration side of that comparison is always larger than the arithmetic of resource times rate suggests.

Lending against the asset: the bands are cliffs

Loan-to-value is the lender's risk measure, not the borrower's affordability measure, and it is priced in BANDS rather than continuously. Rates step at round thresholds, so a small change in deposit that crosses one changes the rate on the whole loan.

That produces the same shape the taxation paper describes in a different setting: a genuine cliff, where a marginal amount of extra deposit is worth far more than its face value because it re-prices everything behind it. Working out where the next band sits is usually a better use of a small sum than paying it towards the balance.

The other half is that LTV depends on the VALUATION rather than on the price. A valuation below the agreed price raises the loan-to-value on an unchanged deposit, which can push a case across a band the other way — and it is the point at which a transaction most often has to be renegotiated.

Borrowing against existing equity works from the same ratio applied to a current valuation, and it carries the consequence that dominates the decision: the debt is secured on the home. That is a different kind of borrowing from an unsecured loan at a similar rate, and the difference is not in the arithmetic.

Disposal: charged by weight, hired by volume

Waste costs are the one place on this page where the unit of charge and the unit of constraint are routinely different, and it produces a result people find counter-intuitive.

A skip or container is HIRED BY VOLUME, and it has a weight limit. Landfill and recovery gates CHARGE BY WEIGHT. So dense waste — soil, concrete, brick, plasterboard, tile — reaches the weight limit while the container still looks half empty, and light waste — insulation, packaging, timber offcuts — fills the container without approaching it.

Loading a skip past its weight limit is not a pricing question either: it cannot legally be lifted, so the practical outcome is a container that has to be partly emptied by hand before it can leave, which costs more than the second skip would have.

SEGREGATION is what reconciles the two, and it is also what makes recovery possible. Separated inert material, metals, timber and plasterboard each go to different destinations at different rates — and metals can be worth money rather than costing it, which is the one waste stream with a positive value and a market price that moves.

It is worth noting what the disposal calculations here do not include: transport, permits, duty where it applies, waste transfer documentation, and the hazardous classification that changes the destination and the price entirely. The pages return a rate applied to a quantity, and the classification is a regulatory determination rather than an arithmetic one.