Why none of the 83 piles can be accepted yet
The screws holding up your house have been going into the ground since Monday. Nobody has proven they're strong enough, and nobody wrote down what "strong enough" means.
What's holding up the house
The soil under your lot is soft near the surface and wet below about 13 feet. It can't carry a house on its own, so the foundation sits on 150 steel screws called helical piles. Each one is a steel pipe with a round plate welded near the tip. A machine screws it into the ground until it reaches soil firm enough to hold its share of the house.
The concrete slab will sit on top of the piles. Once it's poured, you can't reach the piles again. So everything about them has to be proven before the pour.
How a screw proves it's strong: twisting force
You can't weigh-test every pile. Instead, the industry uses a rule measured in the lab: the harder a pile is to twist, the more weight it can carry. The machine's tablet shows the twisting force (torque) live. For Cantsink's pile, the official rule is:
weight it can carry = final torque × 9
To prove 60,000 lb, the pile must finish at 60,000 ÷ 9 = 6,667 ft-lb of torque or more. But there's also a ceiling: twist this pipe harder than 7,091 ft-lb and you risk damaging the steel or the bolted joints. Cantsink's official product rating (ICC-ES report ESR-1559) says installers "must not exceed" it.
So the acceptable finish is a narrow window, about 6% wide. Every pile has to land in it.
not proven 6,667–7,091
the window too hard
steel over its rating 10,000 ft-lb
Torque window for a 30,000-lb pile on Cantsink's 2⅞-inch shaft. Source: ESR-1559 (revised May 2026), Table 6: torque factor 9, maximum torsion 7,091 ft-lb.
The same number, two different forces
Torque can be measured in two units, the way distance can be in miles or kilometers. Foot-pounds (ft-lb) is the American unit. Newton-meters (N·m) is metric. One foot-pound is 1.36 newton-meters, so a reading of 7,000 means very different things depending on the unit:
7,000 ft-lb → carries 63,000 lb · 7,000 N·m = 5,163 ft-lb → carries 46,000 lb
Cantsink's log sheets write down the tablet's numbers and compare them to a target of 6,670 ft-lb. But Dmitry photographed the tablet on Sept 21, and filmed it again this morning, and both times the unit box reads "n-m". There's a second clue: the drive head Cantsink circled on the logs can only deliver 8,000 ft-lb, yet nine piles show readings above 8,000. That's impossible in foot-pounds and easy in newton-meters.
Switch between the two readings below. Each dot is one pile.
Either way, it fails
If the numbers are newton-meters, all 83 piles are too weak to prove their load, most by 10–25%. If they're foot-pounds, 59 of 83 were twisted past the steel's rating, which the product report forbids. Only Juan at Dunaway can decide what that means for the piles already in the ground.
How deep is deep enough
In April 2026, the soil engineer (Quartet Engineers) wrote an addendum for this exact lot, addressed to Igor. It says the helical piles should go "to a minimum depth of 20 ft below the existing grade." That document was in the project folder, but it never became a rule on the structural drawings or in the pile contract. Cantsink priced the job assuming 17 feet.
On site, Cantsink stops screwing as soon as the gauge hits the target. In soft soil with a firmer layer around 14–17 feet, that means many piles stop there, above the 20-foot line.
Installed depth of the 83 logged piles. Red bars are shallower than the 20-foot minimum: piles, ten of them at only 14 feet.
Dmitry also worries that the firm layer is thin, with softer soil beneath it. The deepest boring on your lot (Quartet B-3, 40 feet) shows fairly steady medium-dense sand from 14 to 40 feet, so that worry is only partly supported. It's one more question for the engineer, not a settled fact.
Why it took an outsider to notice
The root cause isn't one mistake. It's that no one owned the rule. Dunaway's drawings say the helical piles are "engineered by others." Quartet's addendum says the pile company should do the design. As far as the email shows, the pile company never sent a signed design. So there was no written depth, no written torque, and no written unit for anyone to check against.
| Who | Their role | What happened |
|---|---|---|
| Cantsink pile installer | Install to the required torque; supply the pile design | Graded its own work Compared N·m readings to a ft-lb target. No signed design found. Logs edited after sending. |
| Keystone foundation contractor | Hires Cantsink; passes records along | Passed through Forwarded logs and said "OK to proceed" without questioning them. |
| Dunaway structural engineer | Sets the loads; the final sign-off on the foundation | Rule never set Drawings say "by others". Asked for a pile submittal Sept 8; no reply yet to Sept 23 email. |
| Quartet soil engineer | Says how deep is safe | Said it 20-foot minimum, April 16, 2026. It didn't reach the drawings. |
| RockSolid your inspector | Watch installation and report to you | Report missing On site Sept 21 and 23; paid for the first visit; no field report or photos received yet. |
| Dmitry Ultimate Homes' advisor | General contractor's technical check | Caught it Photographed the tablet, did the math, reviewed every log. |
What I got wrong
On Sept 23 I told you Dmitry had probably read the gauge in the wrong unit. I was the one who was wrong: his photos and video settle it. I also missed the 20-foot minimum, even though Quartet's addendum was sitting in your own Geotech Reports folder when I reviewed the pile costs on Sept 21.
The records changed after we asked questions
Cantsink emailed log sheets on Wednesday morning, and we flagged pile 25 and the 14-foot piles to Juan and Dmitry. Wednesday evening, Cantsink re-sent the logs, and some entries had changed. Edits like these are why the engineer should get the tablet's own data files and the original handwritten sheets, not just typed copies.
| Item | Wednesday morning | Wednesday evening | Why it matters |
|---|---|---|---|
| Page 1 (Sept 21) | 2 piles | 16 piles | Dmitry saw 3 piles stop at 17 ft that day; neither version shows that |
| Pile 2 depth | 27 ft | 22 ft | The old segments only added up to 22 |
| Pile 25 final reading | 6,190 | 6,812 | Changed after we flagged it as under target; depth still says 14 ft |
| Pile 4 | not shown | 29 ft, final 3,803 | Far below target in either unit |
| Piles 105–111 | not shown | depths, no final torque | Can't be accepted without a final reading |
How we got here
- Two soil studies drill borings on your lot (Geoscience, then Quartet with borings to 25 and 40 ft). Soft, wet soil near the top.
- Quartet addendum to Igor: helical piles, minimum 20 feet deep; also remove and replace the top 4 feet of soil.
- Cantsink proposal: $86,700 for 150 piles, priced at 17 feet each, $28 per extra foot.
- Dunaway's structural drawings: piles "engineered by others", loads of 30,000 and 50,000 lb.
- Juan asks for an updated helical pile submittal for review. No sign it was sent.
- Installation starts. Cantsink warns piles will run deeper than bid. Dmitry photographs the tablet showing "n-m".
- Dmitry's first report: piles likely too weak. Cantsink projects 3,100–3,350 total feet; Keystone says OK to proceed.
- You email Juan six questions (no reply yet). I send Igor the theory that Dmitry misread the units. Cantsink re-sends edited logs.
- Dmitry's addendum and video confirm N·m. My check of the product rating and soil reports backs him on torque and depth.
What happens now
This is serious, but it was caught at the right moment. Piles that haven't been capped can usually be screwed deeper with extensions. The engineer will pick one or more of these fixes:
Extend the piles
Add extensions and drive the existing piles to 20 feet or more. Most likely fix. Costs $28 per extra foot.
Re-test after a rest
Re-twist a sample of piles a day or more later to see whether they hold their torque.
Load-test one or two
Physically push on a pile to prove its capacity before the pour.
Add piles
Where a pile can't be fixed, put another one beside it.
Three emails are drafted and waiting in your Gmail drafts. None has been sent:
To Keystone: ask Cantsink to hold installation, and send the engineer the tablet's data files, the original field sheets and a pile design.
To Juan: set up a call today or tomorrow with Cantsink, Dmitry and RockSolid; rule on the unit, the 20-foot minimum and the piles already in.
To RockSolid: send the missing field reports and say what the tablet showed.
The decisions that are yours: whether to pause the crew for a day or two (standby costs are small next to redoing piles), and whether to keep Dmitry involved through the pour. I'd do both.