Hurricanes don't ruin schedules. The decisions you made in June do. The storm is the day everyone watches; the work is in the eleven weeks before, and the three weeks after.
Francine made landfall on September 11, 2024 at 5 p.m. Central, near the mouth of the Atchafalaya River on the St. Mary – Terrebonne Parish line, roughly thirty miles south-southwest of Morgan City. The National Hurricane Center fixed it as a Category 2 with one-hundred-mph sustained winds and a minimum central pressure of 972 millibars. By the time the eye crossed I-10, we were ninety miles inland in northeast Rapides, watching a 3-to-6-inch rain band climb the parish line.
We were mid-cut on a fourteen-acre industrial pad east of the Red River. [VERIFY-WITH-JOE] Subgrade was running two days behind plan because August had already been wet. The question wasn't whether we'd lose the day of landfall; that was free. The question was whether we'd eat ten days on rebound, or two.
Ninety miles inland, the hurricane is a rain event.
The headline is always the coast. The structural damage, the surge, the eyewall — that's for the parishes south of I-10. Up here, Francine was wind that gusted 20 to 30 and rain that fell long enough to put four inches into a terrace-clay subgrade that hadn't fully bled off the August tropics. The NWS Lake Charles MRMS estimate puts the heaviest inland band, three to six inches, right across northeast Rapides and northern Avoyelles. That is the band we were standing in.
The relevant number on a Central Louisiana site during a landfalling Gulf hurricane is almost never wind. Industry practice on most mobile cranes is to follow the manufacturer's chart, which usually puts the operational cutoff between 20 and 30 mph; OSHA's codified numeric trigger is on personnel platforms at 20 mph sustained or gusting. None of that decided our day. We had stopped lifting forty-eight hours earlier on the same chart that decides every summer thunderstorm.
The wind doesn't kill the schedule. The four inches of rain on a profile that was already two points wet of optimum kills the schedule.
There is “pause work,” and there is “secure site.” They are not the same call.
Pause work is the call you make every spring afternoon. The operator stows the bucket, the foreman clears non-essential personnel, and we wait the cell out. Secure site is a separate playbook, and on the September 9 forecast cycle, two full days out, we ran the long version.
What gets done in the 72 hours before landfall: temporary erosion protection laid down on every open slope and stockpile — we run a heavy double-net coir blanket on cuts steeper than 3:1 and a silt fence and wattle line on the rest. Plate steel goes over every open trench wider than four feet and every bell hole that can't be backfilled. Crew trucks and light plant move to the yard. Tracked equipment moves to the high quadrant of the site and gets oriented into the wind, not broadside to it. Fuel gets topped because we'd rather burn the gallon than wait three days behind the FEMA convoy.
And the paper. Parish-level burn bans and storm-stop notifications get filed before the office closes Tuesday afternoon, not Wednesday morning when half the parish staff is already evacuating family. The owner and the GC get a two-line e-mail: site secured, photos attached, reopen decision pending storm verification. Nobody on the GC side wants a phone call from the contractor on Wednesday; they want a paper trail that says we already moved.
Three dates, one tactic, one made-up week.
We secured the site Tuesday, September 10. [VERIFY-WITH-JOE] Full crew on site at 06:00, plate steel and erosion blanket laid by 14:00, equipment staged on the north pad by 17:30, office out by 19:00. No work Wednesday, September 11 — the day of landfall. We were back on site at 07:00 Thursday, September 12, walking the pad, photographing every joint where the silt fence had taken weight, and running a moisture density check on the top six inches of subgrade before any machine moved.
The moisture readings on that Thursday morning told us what we expected: the pad was three to four points wet of optimum on the south half and closer to two on the north half, where the cross-slope had drained better than the design suggested it would. [VERIFY-WITH-JOE] We did not start compaction Thursday. We did not start compaction Friday. We pulled the dozer hands into a haul-road rehab on the access spur and let the pad bleed off through the weekend.
The make-up tactic was not heroics. It was a second compaction crew, pre-arranged with our standing labor partner on the June seasonal call, on site Monday September 16 to run day-shift and an extension shift to 20:00 with light plant. We also re-sequenced the building pad to run north-south instead of east-west, because the north half had drained first. By the following Friday, September 20, we were back on the pre-storm critical path. [VERIFY-WITH-JOE] Nine working days lost to landfall and rebound, six made up by the end of the month.
Catahoula clay does not read the rebound chart.
We wrote in the first field note that a Proctor curve is a starting point on this ground, not an answer. The corollary on a hurricane-rebound jobsite is that the moisture-content rebound curve is also a starting point. A four-to-six-inch rainfall event on a Pleistocene terrace clay or a Catahoula uplands fill returns to optimum on a timeline somewhere between ten and twenty-one days, depending on temperature, wind, slope, and how deep the cut was before the storm. [VERIFY-WITH-JOE]
The two points that matter for sequencing: it is not symmetric — the top six inches bleeds off faster than the next eighteen, which means proof-rolling the surface tells you almost nothing about whether the lift below it will pump. And it is not patient with the calendar — if the post-storm week happens to be overcast and humid, you can lose the next week, too. The dry-down clock starts when the sun comes back, not when the rain stops.
A subgrade that proof-rolls clean on Friday can pump on Monday. We test, we wait, we test again. The penalty for being two points wet of optimum on a hurricane-rebound pad is the entire pad.
The hurricane year is decided in June.
The same three patterns show up after every named storm we've worked through. The teams that lose the schedule lose it for the same three reasons. The teams that hold the schedule hold it for the same three reasons.
- 01
The pre-season setup pays for itself in a hurricane year.
Erosion blanket, plate steel, and the storm box of fuel, tarps, and signage live on the yard from May 15 to December 1. We pay for that real estate twelve months a year because we will not negotiate it from a phone in week-of-landfall. The crews who scramble for plate when the cone tightens already lost a day; the crews who scramble for blanket lost three.
- 02
Moisture content is patience, not a hammer.
A four-to-six-inch rainfall event on a Pleistocene terrace clay or a Catahoula uplands fill does not bleed off on the schedule the project manager wants it to. We test, we wait, we test again. Bringing the roller back two days early is the single most expensive mistake on a hurricane-rebound jobsite, and we have watched out-of-region crews make it on every storm year we can name.
- 03
Procurement decides whether you make up time or eat it.
If your select-fill and your fuel and your erosion materials are all sole-sourced to one yard in Lafayette, you are sharing a Gulf-coast logistics window with every other contractor on the storm side of the cone. We pre-stage at two yards north of I-49 and one east of the Atchafalaya before the season opens. The premium is two percent on the materials line and it has saved a week of float on every named storm since 2020.
Atlantic hurricane season is June 1 to November 30. The peak is the second week of September. We do not write this down because we forget — we write it down because the contractor on the other side of the bid table doesn't live here, and the owner needs to know whether the schedule respects what that calendar actually does.
— Joe Burns, Founder