Knowledge · Weather and climate
Wet weather delays and trade sequencing,
where a lost week becomes a lost quarter.
The rain day is not what does the damage. The damage is done at the trade boundaries behind it, where a lost window is replaced by the next one that trade has free. This reference works through the cascade stage by stage, explains why the dry-out period costs more than the rainfall record shows, and sets out the parts of the sequence a builder can actually control.
01 / Overview
A lost week does not cost a week
Ask a builder what a wet fortnight cost them and the honest answer is usually a number far larger than the fortnight. That is not exaggeration and it is not poor management. It is the arithmetic of a sequenced programme, where each stage is booked against a trade's calendar rather than against an idea of when the site will be ready. The weather takes the first bite. The sequence takes the rest.
The mechanism is simple enough to state in a sentence: a trade that loses its window does not wait, it books elsewhere. Everything else on this page follows from that. The builder who understands it manages weather delay as a queue problem with a short window to act in. The builder who does not manages it as a drying-out problem and is surprised in October.
Why it matters
Two reasons, one operational and one commercial. Operationally, the size of the delay is decided in the days immediately after the slip, when the next trade's slot can still be held or moved, and that window closes quietly. Commercially, the added weeks consume the same preliminaries as any other weeks; supervision, site facilities, insurances and finance keep running whether or not a trade turned up. Whether any of the time is claimable under the contract is a separate question, answered in inclement weather and extensions of time and extensions of time and delay, and ultimately by the specific contract. The sequencing problem exists either way.
Wider context for what eastern Australian builders have been building through since 2020 sits in weather and construction productivity. Build times over that period lengthened for multiple reasons and weather was one documented contributor among them, not the sole cause. This page stays on the mechanism a builder meets on their own job.
02 / The cascade
How three wet weeks in July become October
One worked sequence, stage by stage. Watch where the time is actually added; almost none of it is added by the rain.
- 01
Early July: the ground is saturated
The rain itself may only close the site for a few days. What follows is the ground, which holds water long after the sky clears. Excavation stops, the pad cannot be prepared, and the pour that was going in this week is not going in this week.
- 02
Late July: the slab finally goes down
The site dries, the pad is repaired and the slab is poured roughly three weeks later than planned. Measured on its own this looks like a three-week problem, and this is the point at which most builders think they know the size of the delay.
- 03
First week of August: the frame slot is gone
The frame trade was booked for the first week of August. They did not sit and wait for three weeks with nothing to do; they filled the slot on another job. The slab being ready has no bearing on whether they are available.
- 04
September: the frame crew returns
The frame goes up when the frame trade next has a window, which is September. The site was ready in late July. The extra five or six weeks are not weather, they are queue position, and no amount of drying out shortens them.
- 05
October: roofing instead of August
Roofing follows the frame, so it lands in October rather than August. The roofer’s own slot has moved for the same reason the framer’s did, and the lock-up date that everything internal depends on has now moved with it.
- 06
The fitout trades arrive into a different programme
By the time the fitout trades are needed, the programme has blown out by ten to twelve weeks. The rain accounted for a fraction of that. The rest was accumulated at each handover, where a lost slot was replaced by the next available one.
Read the cascade back and the shape is obvious. The weather cost roughly three weeks. The programme lost ten to twelve. The difference was created at two handovers, slab to frame and frame to roof, where a lost booking was replaced by the next available one. A third handover into fitout would have added more again if those trades had been tightly booked.
03 / The mechanism
The trade slot is the multiplier
This is the insight the page exists to teach, so it is worth stating plainly. A trade that loses its window does not wait. Trades run their own businesses with their own overheads, their own crews to keep employed and their own book of work. An idle week is a week they cannot afford, so when a builder's slab is not ready they fill the slot somewhere else, usually within a day or two of finding out.
The consequence is that re-sequencing depends on that trade's next availability, not on when the site dries out. Those two dates have nothing to do with each other. A site can be bone dry and rolling by the last week of July and still sit empty until September, because the framer's next free window is in September. The weather set the delay going; the trade's calendar decided how long it ran.
This is also why the effect compounds rather than adds. Each trade in the chain is booked against the completion of the one before it, so when the frame moves five weeks, the roofer is not being asked to move five weeks, they are being asked to move to a date that then has to be found in their calendar too. Every boundary in the sequence is another opportunity for the gap to widen, and the further down the chain the slip travels the more crews it has to be reconciled against. The mechanics of dependency behaviour are covered in critical path and dependencies, and the booking-lead-time side in lead times and sequencing.
Which trades multiply hardest is site-specific and regional. The general pattern is that trades with small crews, high demand or long travel are the ones whose next window is furthest away, and those are the boundaries where a builder's programme is most fragile. A builder who has run a few seasons in one area usually knows exactly who those trades are; that knowledge is worth writing into the programme rather than carrying in someone's head.
04 / The ground
Waterlogged ground outlasts the rainfall record
Rain stops earthworks directly, and that part is visible to everybody. What is less visible is what happens after it stops. Waterlogged ground stays unusable for days and sometimes weeks after the rain itself has cleared, because the site has to shed water, the pad has to dry to a state that will take plant and compaction, and any damage done by working it wet has to be made good first. The lost time therefore exceeds what the rainfall record shows, sometimes by a wide margin.
How long is site-specific. Soil type, site fall and drainage, the season and the amount of sun the site gets, whether the ground was already saturated from the previous event, and how much of the pad was disturbed before the weather arrived all move the number. The same 40mm falls on two sites and closes one for two days and the other for two weeks. That variability is exactly why a general rainfall figure cannot answer the question for a particular job.
The practical consequence is that a builder arguing from a weather bureau printout is arguing the wrong number. The printout describes the region, not the site. It does not record that the pad was under water on the Thursday after the last wet day, that the excavator was turned around at the gate, that the concreter was stood down or that the pump was cancelled. Those are entries in a site diary, written on the day, and they are the record that actually describes what the job lost.
The same point applies when the programme is being built rather than defended. A build period that carries an allowance for wet days but no allowance for the ground being unworkable afterwards is optimistic by construction. Whatever allowance sits inside your programme, it needs to cover the days the site cannot be worked, not the days it rained.
05 / The cost side
The cost base accumulates regardless
Running underneath the whole cascade is a set of costs that does not care whether anyone turned up. Supervision, site facilities, fencing and amenities, insurances and the finance carried against the job are all consumed by duration rather than by work done. A job open for ten extra weeks consumes ten more weeks of that base, and the weeks added by queue position cost exactly the same as the weeks added by the weather.
This is what makes the trade-slot mechanism a commercial problem and not just a logistical one. A builder who thinks of the delay as three weeks has priced the wrong exposure by a factor of three or four. It also changes where effort is best spent: an afternoon spent holding the framer's slot is worth more than a week spent trying to compress work later in the programme, because the cost being avoided is real weeks of overhead rather than notional time.
Whether any of that cost is recoverable is a contract question, and this page does not answer it. Contracts differ in what they treat as a qualifying cause, what allowance the build period already carries, and whether any cost recovery attaches to an extension of time at all. The machinery is described in inclement weather and extensions of time and extensions of time and delay, and the position for any particular job is set by the specific contract and is worth taking advice on. What is certain, and what belongs on this page, is that the overhead runs whichever way that question is answered.
06 / Regional capacity
When there is nowhere to re-sequence into
Everything above assumes the builder can find a next window, even if it is a poor one. After a major flood in the region, that assumption weakens. Tradespeople scheduled onto new residential construction move to repair and remediation, which is urgent, immediate and available in volume, and it absorbs regional capacity for months and often years. The pool a builder is re-booking into shrinks at precisely the moment they need it most.
The point that surprises people is that the builder's own site does not have to have flooded. A dry site in an affected region still cannot sequence trades that are committed to recovery work across that region, so the damage radius for construction capacity is wider than the flood radius. In practice this stretches every gap in the cascade above; the September window becomes November, and the ten to twelve weeks becomes something worse.
That dynamic, what it looked like in Lismore from 2022 and on the Mid North Coast from 2025, and how long the absorption runs after the water goes down, is covered properly in flood recovery and trade availability rather than repeated here. The relevance to sequencing is the single line above: when regional capacity is absorbed, re-sequencing gets harder, and float that looked generous in a normal year stops being enough.
07 / Controllables
What a builder can actually control
Not the weather, and not a trade’s calendar once they have committed elsewhere. These six are decided by the builder, and most of them are decided before the wet week arrives.
Which trades are booked with float
Float is not padding, it is the buffer that absorbs a slip before it becomes a re-booking. A programme where every trade is booked nose to tail converts every weather event into a queue problem. Deciding where the float sits, usually ahead of the trades hardest to re-book, is a decision made at programming time, not in the wet week.
Which sequences are weather-exposed
Everything from site cut to lock-up is exposed; most of what follows is not. A builder who knows which stages sit in the open can see which weeks of the programme carry weather risk and which do not, and can be honest with the client about it before the season rather than after.
How early a slip is detected
A slip found on the day it happens is a phone call to the next trade. The same slip found three weeks later is a re-booking into whatever is left. Nothing else on this list has the same leverage, because early detection is what keeps the next trade’s slot in play at all.
Whether the dependency chain is visible
If the programme does not show what follows what, the downstream effect of a slip is discovered trade by trade as each one is called. A visible chain answers the only question that matters on the day: if this moves three weeks, what else moves, and who has to be told now.
What the next trade is told, and when
A trade told immediately that the slab has slipped three weeks can often hold or shuffle its window. A trade told the week before it was due to start cannot. The conversation costs five minutes and is the single cheapest intervention available in the whole cascade.
What is recorded while it is happening
Site conditions, the days lost, which trades stood down, when each was notified. The record has uses well beyond any claim; it is also how a builder learns, job over job, which trades in their area realistically hold a slot and which do not.
The pattern across all six is that the useful work happens either well before the event, when float and exposure are being decided, or within a day or two of it, when the next trade can still be reached. Very little useful work happens in between, which is where most of the effort usually goes.
08 / Common mistakes
Where wet weather sequencing goes wrong
Each of these is recognisable and mechanical, and most of them happen in the fortnight after the rain rather than during it.
Counting rain days instead of critical-path days
Eleven wet days on the bureau record does not mean eleven days lost, and it does not mean only eleven days lost either. Some of those days delayed nothing because the work in progress was internal. Others delayed a stage everything else depends on, and cost far more than a day each. The programme, not the rain gauge, is what converts weather into days.
Treating the dry-out period as zero
The rain stops on Tuesday and the programme assumes the excavator is back on Wednesday. Waterlogged ground stays unusable for days and sometimes weeks after the rain clears, depending on soil type, site fall and the season. A programme built on rainfall days alone will always be optimistic.
Not telling the next trade early enough
The single most expensive silence in residential building. The builder waits to see if the time can be made up, says nothing for two weeks, and by the time the call is made the trade has committed elsewhere. The slip was three weeks; the delay is now seven, and the extra four were created by the delay in picking up the phone.
Assuming the trade will wait
Trades carry their own overheads and cannot hold an idle week for someone else’s site. Re-sequencing depends on that trade’s next availability, not on the builder’s preference or on when the site dries out. Any recovery plan built on a trade waiting is not a plan.
No contemporaneous record
Nothing is written down while the site is closed, and the position is reconstructed months later from memory and a weather printout. The printout shows what fell in the region. It does not show that the site was under water, that the concreter was stood down, or that the framer was released to another job.
Rebuilding the programme only once, at the end
The dates keep being quoted from a programme that stopped being true in July. Every downstream conversation, with the client, with suppliers, with the trades still to come, is then measured against a baseline nobody believes. Re-sequencing at each slip is less work than one reconstruction in October.
09 / Practical example
Two builders, the same wet July
Illustrative only, not a benchmark. Two builders on comparable detached jobs are hit by the same July weather. Both lose the same three weeks at slab stage: roughly a week of actual rain, then a fortnight waiting on ground that will not take plant.
The first builder rings the frame trade on the Monday of the first wet week, before the slab has even been formally moved, and says the pour is at risk and the August start may need to shift by two or three weeks. The framer has one job either side and shuffles rather than releases the crew, taking a smaller job forward and holding the last week of August. The slab goes down in late July, the frame starts a fortnight late instead of six weeks late, and the roofer, told in the same week, moves by the same fortnight. The job finishes around three weeks behind, which is roughly what the weather actually cost. The site diary carries dated entries and photos for every closed day, so if the contract question is ever raised the evidence already exists.
The second builder waits, reasonably enough, to see whether the time can be made up. Two weeks pass with no call. By the time the framer is told, the crew has been committed to another site and their next clear window is September. Roofing follows into October, and the fitout trades, booked tightly behind lock-up, each land in their own next available slot. The programme is ten to twelve weeks behind and the site has carried ten to twelve weeks of preliminaries to get there. When the delay is finally reconstructed for the client, the supporting material is a rainfall printout showing eleven wet days, which is both the wrong number and a number that invites an argument.
Same rain, same ground, same three weeks lost on site. The difference was made in one phone call in the first week, and in whether anyone wrote down what was happening while it happened.
10 / FAQ
Common questions.
Because the site drying out and the next trade being available are two different events, and the programme runs on the later of them. When a trade loses its booked window it does not hold that window open; it takes work elsewhere, and the builder then re-books into whatever slot that trade has next. Each handover in the sequence can add its own gap for the same reason, so the total is the original weather delay plus the accumulated queue time at every trade boundary behind it. That is why the compounding effect, rather than the rain day itself, is what does the damage.
No, and it usually understates it. Rain stops earthworks directly, but waterlogged ground then stays unusable for days or weeks after the rain itself has cleared, so the site is out of action well past the last wet day on the record. A builder arguing from a weather bureau printout is arguing the wrong number in both directions: the printout does not capture the dry-out period, and it does not distinguish a wet day that delayed nothing from a wet day that moved the whole job. What the site diary records, conditions on the ground and what actually stopped, is the number that describes the job.
Sometimes, and the answer is a contract question rather than a sequencing question, so it is not answered here. What qualifies, what allowance the build period already carries, what notice is required and within what window are set by the specific contract, and the general machinery is covered in inclement weather and extensions of time and in extensions of time and delay. The point this page makes is separate and applies either way: whatever the entitlement position turns out to be, the programme still has to be re-sequenced around trade availability, and the cost base keeps running while that happens.
Because a large share of a residential job’s running cost is priced by duration rather than by work done. Supervision, site facilities, fencing and amenities, insurances and finance keep being consumed for every week the job is open, whether or not a trade turned up. A programme that runs ten weeks longer consumes ten more weeks of that base. This is why the compounding effect matters commercially and not just logistically; the weeks added by queue position cost the same as the weeks added by the weather.
It makes re-sequencing harder, because the pool of trades to re-book into is smaller. After a major flood, tradespeople scheduled onto new residential work move to repair and remediation, which is urgent and available immediately, and that absorbs capacity for months and often years. A builder whose own site was never flooded still cannot sequence trades if those trades are committed to recovery work across the same region. The mechanism, and what the Lismore and Manning River events showed about how long it runs, is covered in flood recovery and trade availability.
How early a slip is detected and how quickly the next trade is told. Weather is not controllable, and neither is a trade’s calendar once they have committed elsewhere. What is controllable is the window between the slip happening and the next trade hearing about it, because that window is when their slot can still be held or shuffled. Everything else on the controllables list, float placement, knowing which sequences are exposed, a visible dependency chain, exists to make that window shorter.
11 / Terms
Glossary for this topic
Trade slot (the window a trade has reserved for a job, which is released to other work when the job is not ready), compounding delay (the effect where a delay grows at each trade boundary rather than staying the size of the original event), dry-out period (the time after rain stops during which the ground is still unworkable), float (the spare time between dependent tasks that absorbs a slip before it forces a re-booking), dependency chain (the ordered set of stages where each one cannot start until the one before it is complete), re-sequencing (rebuilding the programme around trade availability after a slip), critical path (the sequence of tasks that sets the finish date), contemporaneous record (site conditions and events written down on the day rather than reconstructed later). Definitions for the wider vocabulary live in the construction glossary.
The dates, notices and diary entries that hold a re-sequenced programme together are only useful if they exist while the job is running; the applied version of that discipline is in the residential construction scheduling guide.
12 / Keep reading
Related knowledge
13 / Further reading
Primary sources
- Australian Bureau of Statistics, Home Building Through the Pandemic, for the published average time from commencement to completion for new private houses, quarterly by state.
- Reserve Bank of Australia, Recent Trends in Australian Productivity, Bulletin, September 2023, which identifies wet weather among the factors that constrained construction output.
- Bureau of Meteorology, rainfall and temperature long-range forecasts, and the ENSO history record, for the seasonal outlook a programme is built against.
- Brad Caldon, The weather nobody in housing is talking about, Sourceable, July 2026, the operator account this cluster draws its worked cascade from.
- Your own job records. The most useful source for which trades in your area hold a slot and which do not is the last three years of your own programmes and site diaries.
See what a slipped stage does to every trade behind it, on the day it slips.
VIABUILD models the dependency chain, so when one stage moves the downstream effect on every following trade and on the claim schedule is visible immediately rather than discovered a month later.