The Two Measurements That Decide Repair or Replace on Steel
By Ed Brannigan, Marine contractor. Reviewed by Dana Whitcombe, technical reviewer.
Seawall Repair | Boat Lift Installation Late October, 2017. Panama City. A cold front had the channel rolling, whitecaps all the way across Saint Andrews Bay. I was kneeling on a finger pier, water slapping the stringers, trying to keep the ultrasonic probe level on a steel H-pile that hadn't been painted since Bill Clinton was in office. The marina owner stood behind me in a safety vest, asking the same question he'd asked three contractors already: 'Patch it or jackhammer it out?' I didn't answer right away. I was busy measuring. Because for thirty years, on the Gulf and Atlantic coasts, I've learned that this question—repair or replace—comes down to two numbers. Not the color of the rust, not how it sounds when you tap it with a hammer. Two measurements. If you get them right, you spend a quarter of what a replacement costs. If you guess, you'll be back there in three years, cutting out a bigger section. I made the mistake of guessing once. In 2001, on a dock in Mobile Bay, I told a client we could weld a doubler plate over a bad spot. I didn't measure the run of corrosion, just the worst point. The plate did fine for a year, then the rust just kept walking underneath it. I lost the client and a season of work. After that, I made a drill: two measurements, no exceptions.
Measurement One: Remaining Wall Thickness
The first number I write down is the actual wall thickness of the steel, measured dead straight with an ultrasonic thickness gauge. I grind a clean spot on the pile, wipe it flat, smear coupling gel, and press the probe steady. A decent gauge is under a thousand bucks—mine reads in thousandths of an inch. You need the original specification to know what you're starting from. If it's an HP 14x73, the flange and web are rolled to standard. But drawing or not, that number tells you how much metal is actually left between the sea and the structural core. After three decades, I've seen a 3/8-inch flange worn to 0.220—a 40% loss—and the pile still standing. But I've also seen a sharp pit burn through and a finger push right through it. The first measurement is about whether the steel is still there to do its job.
Measurement Two: The Continuous Run of Deterioration
The second number is the length of the corroded zone. I run a tape from the highest obvious pitting or rust scale down to where clean, uncorroded steel begins. That dimension goes along the flange or the pile face, not around the circumference—you're looking for the vertical run. A pile might lose half its thickness at one specific spot at the mud line, but if the run is only eight inches long, you can box it out and weld a collar. If the rust runs twelve feet down the same pile, even with decent thickness remaining, no patch will stay bonded. There's no way to stop corrosion at a seam. That measurement tells you the scale of the repair and the real risk. In Pensacola in 2018, two piles on the same seawall had the same 60% remaining thickness. The first had an 18-inch run. We welded a doubler, coated it, and it's still there. The second had a 12-foot run. We drove a new pile. Same number one, different number two, opposite outcomes.
Reading the Two Together
If the remaining thickness is below what the engineer says is the minimum section modulus for the load, you replace. Period. That's non-negotiable. But if you've got a number that's okay, the run length decides whether you can trust a repair. A short, shallow section can be ground out and patched, because the steel around it is still sound. A long, continuous section means corrosion is active all over the surface. You can't weld a new boundary on ragged steel. I've seen contractors patch a 15-foot section on a sheet pile and pour concrete behind it. Two years later the crew is back, cutting it out again. The cost of repair with the wrong measurements is worse than replace. The two measurements give you a defensible reason to do either one.
Why I Always Make These Measurements on the First Visit
When I'm on a dock and the owner wants an answer right there, these two numbers are the fastest way to a recommendation. They also protect me. If I say 'replace' and the measurements show only a short run and decent thickness, I'd better have a good reason. If I say 'repair' and the thickness is gone, I'm setting them up for failure. So I take the time, write the numbers, and share them. It kills the argument. And the owner sees it's not a guess. The measurements are the difference between a piece of business you can stand behind and a lawsuit waiting on the tide.
Frequently Asked Questions
Can't you just weld a patch over a thin spot? You can, but only if the remaining thickness is enough for a solid weld. If the steel is too thin, the weld burns through or creates a hard spot that cracks. The patch holds if the run is short and the metal around it is sound. If the measured thickness is below spec for the pile, no patch will make it structurally whole. That's why we use the second measurement before ordering steel. How do you know the original thickness if you don't have the drawings? Most steel piles are standard rolled shapes—H-piles, sheet piles, pipe—and the dimensions are published. You can also find a protected point near the top, under the cap, where the original thickness survives. For custom steel, we sometimes have to cut a coupon. But you need that baseline to interpret the thickness number. Do I need an ultrasonic gauge, or can I just tap with a hammer? A hammer tells you if it's sound to the ear, but not by how much. A decent UT gauge costs a few hundred dollars and has paid for itself on the first job. I've tapped steel that sounded solid and found it was 80% consumed, and I've tapped bad-looking scale that cleaned up to 90% remaining. The numbers don't lie. If you're maintaining a dock or a seawall, I'd rather see you buy a gauge and measure than trust a hammer.