Setting Out a Transfer Plate: The 22 mm Column
You are the AP's surveyor for a 42-storey residential tower on a 3-level podium in Yau Tong. The tower columns are set out from a transfer plate at +18.5 m. The contractor's surveyor established a local control network on the podium roof using a three-point resection from two government control stations and one temporary nail in the slab. The resection geometry is poor: the two government stations subtend an angle of 178° at the instrument, and the temporary nail is 8 m away. The contractor used an unverified prism constant and did not apply atmospheric correction. During a routine as-built check at Level 6, you find column C12 is 22 mm east and 9 mm north of its design position. The contractor says the error is within 'normal tolerance' and wants to proceed with the next floor. The AP asks you to review and advise. The structural drawings specify a maximum positional tolerance of ±10 mm for columns at each level.
1. Identify the likely sources of the 22 mm error and explain how you would test each.
Read the worked answer
The error is large enough that several sources must be considered together.
- Resection geometry: two stations subtending 178° at the instrument give a very weak fix because the intersection is nearly a straight line. The temporary nail 8 m away adds little strength. I would test by re-observing from three well-distributed stations and comparing the computed position.
- Control coordinates: the government stations may be correct, but the local coordinates may have been transferred with a scale or orientation error. I would check the resection residuals and repeat the network with a closed traverse.
- Prism constant: an unverified prism constant can produce a systematic range error. I would check the prism against a known baseline or use a different prism and compare.
- Atmospheric correction: ignoring temperature and pressure affects distance, especially over longer sights. I would recompute the observations with measured atmospheric data.
- Instrument calibration: check the total station's collimation, trunnion axis, vertical index and EDM against a calibration certificate or a field check.
- Setting-out method: target miscentring, staff/target height, and transfer of control through the podium can all shift the column.
I would not accept a single explanation until the independent network closes within tolerance.
2. What verification survey would you carry out before advising on C12, and what would you report if the 22 mm is confirmed?
Read the worked answer
I would establish an independent control network from at least three well-distributed stations with good geometry, using a calibrated total station and a verified prism. I would observe C12 from two setups, measure the column faces and centre, check verticality, and compare with the design coordinates. I would also check the adjacent columns and grid lines to see whether the error is isolated or systematic.
If the 22 mm east and 9 mm north deviation is confirmed, it exceeds the ±10 mm positional tolerance on the drawings. I would report:
- the method, control, instrument and calibration details;
- the observed coordinates and the computed deviations;
- the likely cause, distinguishing systematic control error from local setting-out error;
- the effect on the structural grid, cladding, lifts and subsequent floors;
- a recommendation that the contractor submit a non-conformance report and that the RSE assess structural and dimensional implications before further floors proceed.
I would not certify the column as compliant. If the error originated in the control network, other columns set out from the same network may also be affected and should be checked.
3. The contractor argues that because the building is still stable, the AP should accept the as-built deviation. How do you respond, and what does professional liability require?
Read the worked answer
I would explain that stability is not the same as compliance. A 22 mm positional deviation can affect the structural grid, precast cladding, lift shafts, curtain wall brackets, drainage falls and future floors. The tolerance exists to control accumulation and interface risk, not only immediate stability. The fact that the building has not collapsed is not evidence that the deviation is acceptable.
Professionally, the AP has a duty to ensure that the works comply with the approved plans and that any non-conformance is properly assessed. The AP should not be pressured into accepting a deviation because the contractor wants to proceed. I would document the contractor's request, the measured deviation, the verification method and the advice given. The RSE should assess whether the deviation can be accepted, strengthened or remedied. If acceptance is proposed, it should be a recorded engineering decision, not an informal assurance.
I would also check whether the contractor's control network is still in use. If it is defective, the next floor will repeat and compound the error. My report would state the facts and the limitations, and I would avoid signing any document that implies the column is within tolerance when it is not.