Weather days are the only delay claim on a construction project that arrives every single time, on every job, in a form nobody disputes in principle. They are also, on most contracts, governed by a sentence containing the word unusual, with no definition attached to it.
The standard clause and what is wrong with it
Most contracts say something close to this: the contractor is entitled to an extension of time for delays caused by unusually severe weather conditions.
Three words in that sentence are doing all the work and none of them is defined.
Unusually. Compared to what. Last year, the ten year average, the thirty year normal for this location. Each produces a different answer and the difference is measured in weeks.
Severe. How much rain stops work. On an excavation, a quarter inch may. On an interior fit out, four inches will not.
Caused delay. A day lost on an activity with three weeks of float caused nothing. The event has to hit the critical path.
The result is a clause that both parties agree with at signature and interpret differently in month eleven, at which point the contractor has daily records and the owner has a memory of it having been a wet spring.
Which is entirely avoidable, because weather is the one risk on a construction project with thirty years of published data behind it.
What a weather days baseline does
The fix is a table in the contract. Twelve rows, one per month, each carrying a number of anticipated lost days for that location.
The numbers come from published climate data for the nearest station, converted into working days lost using agreed thresholds. Those anticipated days are inside the contract price and the contract duration. The contractor carries them.
Days beyond the monthly baseline are extensions of time. Days below it are the contractor gain.
That single table removes the whole argument, because unusual now has a number attached to it for each month of the year.
It also makes the schedule honest. A baseline schedule that shows no weather contingency at all on a project pouring foundations through February is not a schedule, it is an aspiration, and the same reasoning applies as in whether the completion date is real.
Building the table takes about two hours using free public data, and it is the single highest return administrative task in the whole contract.
Defining the threshold, which is the other half
A baseline count is meaningless without agreement on what counts as a lost day.
Three definitions are common and the right choice depends on the work.
Measured conditions. Precipitation above a stated amount, temperature below a stated figure, or sustained wind above a stated speed, measured at a named station. Objective, verifiable, and the best option for site work.
Effect based. A day on which the contractor was prevented from performing critical path work for more than a stated number of hours. Closer to the commercial reality and harder to verify.
Hybrid. Measured conditions establish eligibility, effect establishes whether the day counts. Most defensible and slightly more administration.
Whichever is chosen, two further points need settling. Which weather station is authoritative, because site readings and airport readings differ. And whether a day counts when conditions occur outside working hours but make the site unworkable the following morning, which on earthworks is the majority of the argument.
These are small decisions that take twenty minutes at signature and produce months of correspondence when left open.
A worked example
Illustrative figures. Not taken from any client project and not a quotation.
A 26 million dollar warehouse on a site with significant earthworks, scheduled from October through the following September.
The contract carries the standard unusually severe weather clause with no baseline and no threshold.
In month seven the contractor claims 47 weather days across the winter and spring, supported by daily reports each recording the site as unworkable.
The owner disputes the claim on the basis that the winter was ordinary.
Reconstructed against thirty year climate data with a quarter inch precipitation threshold, the anticipated figure for those months is 29 days and the actual qualifying figure is 38. Of those, 31 fell on critical path activities.
The defensible entitlement is therefore 9 days rather than 47, and the argument about the other 38 took four months and two consultants.
A baseline table and a threshold definition, agreed in an hour before signature, would have produced the number 9 automatically and without a dispute.
Why owners resist the table, and why they should not
The commonest objection is that writing down anticipated weather days concedes delay before the project starts.
It does the opposite. Without a table, the contractor is entitled to time for anything unusual, and unusual is argued at the end from the contractor’s own records. With a table, the contractor has agreed to absorb a specific number of days per month inside the price and inside the duration.
The table is a transfer of risk to the contractor, made explicit. It is the owner instrument, not the contractor one.
The second objection is that the anticipated days lengthen the schedule. They do, by the amount that weather was always going to lengthen it. A schedule that ignored weather was going to be missed, and the miss would have arrived as a claim rather than as a plan.
The third is that it invites monthly administration. It replaces monthly administration with arithmetic, since the count is taken from a published record rather than argued from daily reports.
Money, as distinct from time
Weather delay is almost universally excusable and almost universally non compensable. The contractor gets time and carries its own extended cost.
That is the standard allocation and it is a sensible one, since neither party caused the weather.
Two consequences follow for an owner.
First, the extension is still expensive. The owner carries its own extended costs: financing, insurance, staff, and whatever the delayed opening is worth. Those are frequently larger than the contractor extended general conditions would have been.
Second, weather days interact with liquidated damages. A project that finishes 31 days late with 31 days of granted weather extension attracts no damages, and the owner absorbs the whole consequence. That is the correct legal outcome and it is worth understanding before relying on a damages clause as protection, which is the argument in what liquidated damages are worth.
Which makes the baseline table valuable in a second way. Days inside the baseline produce no extension at all, so every day the table absorbs is a day of owner exposure removed.
There is one exception worth drafting for. Where weather damages work already in place, the cost of repair is not a weather delay question at all. It is an insurance question, and it belongs under the builders risk policy rather than under the extension of time clause.
Contracts that fold storm damage into the weather clause tend to produce an argument about whether the repair time is an excusable delay, when the more useful question is whether the policy responds and what the deductible is.
Keeping the two separate at signature takes one sentence and removes an argument that arrives at the worst possible moment.
What we do
We build the baseline table from published climate data for the site, agree the threshold definitions and the authoritative station, and check that the baseline schedule actually contains the anticipated days rather than assuming clear weather for twenty two months. That work is part of the schedule and procurement risk review. Where a claim is already on the table, dispute and claims support works from the same records under the direction of your counsel.
Questions people ask
What period should the baseline use?
A thirty year normal is the usual reference and is the most defensible, because it smooths individual wet and dry years. A ten year period is sometimes argued as more representative of current conditions. Either works provided both parties agree it in the contract rather than after the winter in question.
Does a weather day always mean a lost day?
No, and that is why the critical path test matters. Work can stop on one activity while continuing on others, and a day recorded as unworkable on excavation costs nothing if the structure above is proceeding. The daily report shows what was worked, which is the record that answers it.
Should extreme events be treated separately?
Yes. Hurricanes, floods and named storms belong under force majeure rather than under ordinary weather, because the consequences include damage and site restoration rather than lost days alone. Keeping the two clauses separate prevents an extreme event from being absorbed into a routine weather count.
This is general information about construction contracts and is not legal advice.