Pricing weather risk at tender stage

The tender stage problem

Every construction tender includes a weather allowance. The problem is that most of them are wrong — not because the estimator is careless, but because the inputs used to build the allowance are not matched to the specific project.

The typical approach is to apply a rule of thumb: a percentage of programme duration, a number of days drawn from experience on similar projects, or a figure carried forward from a previous tender for a similar value of work. Sometimes it is a number agreed informally across the estimating team. Rarely is it calculated from measured weather data at the specific location where the project will be built.

That matters because weather varies significantly by location, by month, and by trade operation. A project in the north of England in January faces a different weather risk profile from the same project in the south-east in September. A roofing package running through winter has a different risk profile from an internal fit-out running through summer. Using the same rule-of-thumb allowance for both leads to systematic mispricing — sometimes too much, often too little.

The consequence of getting it wrong is absorbed somewhere. If the allowance is too generous, margin is surrendered at tender. If it is too low — the more common outcome under competitive pressure — the gap is absorbed on-site, becomes a dispute, or becomes an Extension of Time claim that has to be fought for after the event.

What weather risk actually means at tender

Before looking at how to quantify weather risk, it is worth being precise about what it is. Weather risk in a construction tender is not a single variable. It is the aggregate of:

• The number of days on which specific planned operations cannot proceed, or are materially constrained, because of weather conditions at the project location
• The cost of standing-down labour and plant on those days
• The cost of re-sequencing the programme around constrained days
• The programme risk — the extent to which weather-constrained days may affect the completion date, and the delay damages that follow

These are not the same thing. The financial cost of lost productivity on a weather day is quantifiable in isolation. The programme risk — whether lost days cause the project to overrun — depends on the programme structure, the amount of float on weather-sensitive activities, and the contract's provisions for extension of time.

At tender stage, both need to be understood. But the starting point for both is the same: how many unsuitable days should be expected, for which operations, in which months?

Why location and season matter more than most estimators allow

The single most common source of error in weather allowances is treating weather risk as a national average, or carrying forward an allowance from a previous project in a different location or a different season.

UK weather data illustrates this clearly. Average monthly rainfall varies by a factor of two to three across England, Scotland and Wales. Wind speeds at coastal and exposed upland locations are substantially higher than at sheltered inland sites even a few miles away. Frost days in January in the north of England are three to four times more frequent than in the south-east. These are not marginal differences. For operations sensitive to those conditions — concrete placement, roofing, cladding, crane operations — the difference between locations can represent several additional unsuitable days per month.

The month in which weather-sensitive operations are programmed is equally important. Rain in October falls more frequently and in greater intensity than rain in May at most UK locations. Wind speeds in winter months are materially higher than in summer. The programme that puts structural steel erection in November and internal fit-out in June is a different weather risk profile from the same programme reversed.

A weather allowance built from national averages and applied without regard to location or programme sequence is not a risk calculation. It is a guess — usually an optimistic one.

The role of historical data

The correct input for a tender-stage weather allowance is historical measured weather data at or near the specific project location — analysed by month, by weather parameter, and by trade operation.

Ten years of historical data provides a statistically meaningful sample for most weather parameters at most UK locations. It is also the benchmark used in many standard-form contracts. NEC4, for example, references weather conditions that occur less frequently than once in ten years as a compensation event trigger. A ten-year historical analysis simultaneously satisfies the requirements of the tender-stage risk assessment and establishes the baseline needed to support a future Extension of Time claim under the same contract.

The data required to build a useful tender-stage allowance includes:

• Daily rainfall totals — to identify days on which rainfall intensity exceeded the threshold for sensitive operations (concrete placement, coatings, applied finishes, earthworks)
• Daily and hourly wind speeds and gusts — to identify days on which wind conditions constrained or stopped lifting operations, scaffolding, cladding, roofing and work at height
• Daily minimum and maximum temperatures — to identify days on which temperature fell below the threshold for concrete placement, painting, and other temperature-sensitive operations, or exceeded the maximum for hot-weather concrete work
• Frost and ground-freeze days — for earthworks-heavy programmes running through winter
• Snowfall days — where relevant by location and season

For each parameter, the analysis produces a count of unsuitable days per month for each relevant trade operation — based on the specific threshold for that operation, not a generic "bad weather day" definition.

Unsuitable days by trade operation

This is the part of the analysis that most tender-stage weather assessments omit — and it is the most important part. The number of unsuitable days is not the same for every operation. Rain that stops earthworks may have no effect on internal fit-out happening on the same site on the same day. Wind that grounds the tower crane does not stop concreting. Cold temperatures that prevent external coatings do not prevent piling.

The practical implication is that a tender-stage weather allowance should be built operation by operation, month by month, matched to the programme. The total unsuitable days across the project is not a single number — it is a matrix: which trades are on site in which months, and how many unsuitable days does each trade face in those months at this location?

Building the allowance from the data

With historical unsuitable-day counts by trade and month, the tender-stage allowance can be constructed systematically.

The process starts with the programme. For each weather-sensitive trade package, identify:

1. The months in which the operation is planned to be on site
2. The number of working days planned in each month for that operation
3. The historical average unsuitable-day count for that operation in each of those months at the project location

The expected number of weather-affected days for that operation is the sum of the monthly unsuitable-day averages, pro-rated for the planned working days in each month. This gives a realistic expectation of weather disruption — not a national average, not a rule of thumb, but a figure derived from measured historical conditions at the specific location, for the specific operation, in the specific months planned.

That figure can then be converted into a cost allowance: the expected number of weather-affected days multiplied by the standing-down cost of the relevant trade (labour plus plant that cannot be redeployed). The cost of re-sequencing, and any residual programme risk, should be considered separately.

The contractual connection

The tender-stage weather allowance also has a direct contractual function that is often overlooked. Under many standard-form contracts, the contractor's weather risk is assessed by reference to what a competent contractor could reasonably have anticipated at the time of tender.

NEC4 defines a weather compensation event as conditions that are less frequent than once in ten years — that is, conditions worse than the historical 1-in-10 benchmark. JCT contracts refer to "exceptionally adverse weather conditions" — again, a threshold that implies a historical benchmark against which "exceptional" is measured. AS 4000 and NZS 3910 similarly reference conditions that could not have been reasonably anticipated.

In every case, the contractor's entitlement to relief depends on demonstrating that the actual conditions exceeded what was foreseeable at tender. The historical data that underpins the tender-stage allowance is the same data that establishes what was foreseeable. A contractor who has quantified the weather risk at tender — using the ten-year historical data at the specific location — is in a much stronger position to argue that conditions on a particular month exceeded the historical norm, because the analysis already exists.

A contractor who priced weather risk on a rule of thumb has no such baseline. They cannot demonstrate what was foreseeable, because they never assessed it.

Programme risk versus cost risk

Weather risk at tender has two dimensions that are related but not identical: cost risk and programme risk. Understanding the difference matters for how the allowance is structured and how the contract's risk allocation is negotiated.

Cost risk is the financial cost of weather-affected days: standing-down labour and plant, additional protection and reinstatement measures, the incremental cost of working in adverse conditions. This risk stays with the contractor unless the contract provides a mechanism for recovery — which most standard forms do not for weather conditions within the foreseeable range.

Programme risk is the risk that weather-affected days cause the project to overrun the contract completion date, triggering delay damages (liquidated damages or general damages depending on the contract). This risk is usually shared: the contractor bears the cost of acceleration and mitigation, but if weather conditions exceed the contractual threshold, an Extension of Time may be available — protecting the contractor from delay damages even if the cost of lost time stays with the contractor.

At tender stage, the distinction matters because: (a) the cost risk is almost entirely the contractor's, and should be priced accordingly; (b) the programme risk is partially transferable through the EOT mechanism, but only if the correct notice and evidence procedures are followed during the works. Pricing the cost risk without understanding the programme risk, or vice versa, leads to systematic errors in how the tender allowance is structured.

Using historical data to stress-test the programme

Beyond the average-case allowance, historical data also allows the tender programme to be stress-tested against realistic worst-case scenarios. The average unsuitable-day count for a given month represents the central expectation. But historical data also shows the distribution: in some years, the count in that month was well above average; in others, it was well below.

A programme with little float on weather-sensitive activities is exposed to the upper end of the historical range, not the average. If the average unsuitable-day count for roofing in January is six days but the worst recent January produced twelve, and the programme has only three days of float on the roofing package, the average-case allowance understates the programme risk by a material margin.

Understanding that distribution — and how much float the programme provides to absorb above-average weather — is part of a complete tender-stage weather risk assessment. It is also the basis for deciding whether to seek additional float in the programme, negotiate the completion date, or accept the risk and price it accordingly.

What a complete tender-stage weather risk assessment looks like

A properly constructed tender-stage weather risk assessment produces the following outputs:

• Expected unsuitable-day counts by trade operation and month, based on ten years of historical data at the project location
• A cost allowance for each weather-sensitive trade package, calculated from the expected unsuitable days and the standing-down cost of the relevant trade
• A programme sensitivity analysis showing which activities are exposed to above-average weather years and how much float exists to absorb that exposure
• A summary of the contractual weather risk allocation — what the contract defines as the threshold for relief, and how the historical data relates to that threshold
• A record of the analysis, which can be used to support any future Extension of Time claim by establishing what was foreseeable at tender

This is not an academic exercise. Each of those outputs has a direct commercial function — in the tender price, in the programme, and in the contract administration during the works.

How Construction Weather supports tender-stage planning

The Tender Planning tool in Construction Weather generates operation-specific unsuitable-day counts by month using ten years of historical weather data at the specific project location. Rather than working from national averages or regional rules of thumb, the analysis is location-specific: the data is drawn from the nearest available weather station to the project, and the unsuitable-day counts are calculated against the relevant thresholds for each trade operation.

For each trade selected, the tool shows the average number of unsuitable days per month — the figure the programme and allowance should be built from. The output is available in a format that can be incorporated directly into the tender build-up and the risk register.

The same ten-year historical dataset that underpins the tender-stage allowance is also used to benchmark actual conditions during the works — so the baseline for the tender assessment and the baseline for any future EOT evidence analysis are consistent. This matters if the project later moves to an Extension of Time claim: the contractor can demonstrate not just what conditions occurred, but how they compared to what was foreseeable at tender.

Weather risk is quantifiable. Price it properly.

Weather risk at tender is not unquantifiable. It is uncertain — but uncertainty can be modelled, measured and priced. The alternative to a rigorous assessment is a number drawn from habit and compressed under competitive pressure, which absorbs into margin or into programme risk that only becomes visible when it is too late to do anything about it.

The tools to do this properly — historical data at the specific project location, operation-specific unsuitable-day counts, a clear connection to the contractual risk allocation — are available before the tender is submitted. Using them produces a more defensible allowance, a more realistic programme and a stronger position if an Extension of Time claim is needed later.

Weather risk is one of the few construction risks that is genuinely measurable before the project starts. There is no good reason to price it without measuring it.