Anchor Rod Nonconformances Keep Repeating After a Reminder
Over six months an erector logged 41 nonconformances: 17 anchor rod misalignments greater than 1/4 in., 9 column plumbness readings outside tolerance, 6 bolt hole misalignments, 5 coating damage cases and 4 miscellaneous. The anchor rod issues cluster on three projects, all where the foundation contractor set rods without the erector's template and the erector's pre-erection survey was performed after the rods were grouted. Last quarter's corrective action was a toolbox talk reminding field crews to check anchor rods. Rework cost $86,000 and added 11 days across the projects. Management wants a plan before the next quarter.
1. Evaluate whether the prior corrective action was effective and explain what the data show about the real cause.
Read the worked answer
The prior action was not effective, and the data make that plain: 17 anchor rod NCRs occurred across three projects, and a toolbox talk does not change when the survey happens or who sets the rods.
The distribution matters. Anchor rods are 41 percent of all NCRs and cluster on projects with a shared condition: no erector template at setting, and survey after grouting. That is an interface and timing problem, not a motivation problem. A reminder asks workers to check a condition that has already become irreversible by the time they see it. Plumbness and bolt hole NCRs are smaller and may have different causes; treating all 41 as one behavior problem would obscure that.
A useful test of effectiveness is recurrence, not attendance. The reminder was delivered and the NCRs continued, so the action failed. The conclusion is that the corrective action addressed the wrong control point: it tried to improve detection after grouting instead of preventing misplacement before it.
2. Propose error-prevention changes upstream of the field crew and state who must own each change.
Read the worked answer
Shift the control to the point where the error is still cheap to fix:
- Template and setting requirement in the foundation contract. The erector supplies or approves the anchor rod template and the foundation contractor uses it. Owner: procurement and project manager, with engineering support.
- Hold point before grouting. No grouting until the erector or a surveyor has recorded rod position and elevation. Owner: superintendent, with the foundation contractor's agreement.
- Joint pre-erection survey with the foundation contractor and the erector present, using the same datum. Owner: project manager; the record goes to the QCM.
- Pre-job meeting agenda item on anchor rod responsibility and tolerance, not a toolbox talk after the fact. Owner: QCM.
- Feedback to design where rod patterns are congested; a small detail change can remove the interference. Owner: project engineer.
Each change is a documented interface, not a reminder. The most important is the hold point before grouting, because it converts an irreversible error into a correctable one.
3. Define leading indicators and a 90-day verification plan that would tell management whether the changes are working before the next quarterly NCR report.
Read the worked answer
Lagging NCR counts arrive too late. Leading indicators should measure whether the new controls are actually being used:
- Percentage of foundations where the template was used and recorded, before grouting.
- Percentage of hold points verified with a signed survey record.
- Number of joint pre-erection surveys completed, by project.
- Anchor rod NCRs per 100 rods set, tracked weekly rather than quarterly.
- Rework hours and first-time acceptance on the first 20 columns per project.
A 90-day plan:
- Days 1–15: baseline the indicators on active projects and issue the revised hold-point procedure.
- Days 16–45: weekly review of hold-point compliance; any missed hold point is treated as a stop-work event for grouting.
- Days 46–75: compare anchor rod NCR rate with baseline; interview crews to find where the procedure is being bypassed.
- Days 76–90: management review with three outcomes - continue, modify or replace each control.
The test is not whether NCRs fall in one quarter; it is whether the hold points are being used and whether the NCR rate moves in the expected direction. If compliance is high and NCRs do not fall, the root cause was misdiagnosed and the plan should change.