Setout of a low-gradient drainage channel with a suspect benchmark
A 1.6 km stormwater channel is being set out through a new subdivision. The design invert falls at 0.10% from chainage 0+000 to 1+600, with a minimum freeboard of 0.3 m. You have three benchmarks: BM1 at the outlet, BM2 mid-site, and BM3 at the inlet. A closed level run between BM1 and BM3 closes within 6 mm. The run to BM2 shows a 25 mm misclosure, and BM2 has been used by the contractor’s machine-control system. Recent setout stakes near chainage 0+900 are 30 to 40 mm above the design invert, and the excavator operator reports that the design surface looks wrong on the display. The contractor wants to proceed because the channel is behind program. The design engineer asks you to decide whether the stakes or the benchmark are wrong, and whether the works already placed near 0+850 should be reworked.
1. Diagnose the likely source of the discrepancy and describe the control checks you would perform before any further setout.
Read the worked answer
The closed run between BM1 and BM3 closing within 6 mm shows that the primary control at the ends is consistent. BM2 has a 25 mm misclosure, which is large for a local level run and makes it the prime suspect. The fact that the machine-control system has been using BM2 explains why the stakes and the display are both offset in the same direction. I would not adjust BM2 to force a closure. I would re-run levels from BM1 and BM3 to BM2 with a different level or a total station height traverse, check BM2 for disturbance, and compare its published level with the new observation. I would also check the machine-control file, coordinate system, design surface and total station setup, because a datum or file error can mimic a benchmark error. The stakes should be independently checked by total station from a known good benchmark. Only after the control is verified would I decide whether the stakes are wrong or the benchmark is wrong.
2. If the outlet invert at 0+000 is 12.450 m, calculate the design invert levels at 0+800 and 1+600 and explain how a 25 mm benchmark error affects the works.
Read the worked answer
The grade is 0.10%, which is 0.001 m per metre. Over 800 m the fall is 800 x 0.001 = 0.800 m, so the design invert at 0+800 is 12.450 - 0.800 = 11.650 m. Over 1,600 m the fall is 1.600 m, so the design invert at 1+600 is 12.450 - 1.600 = 10.850 m. A 25 mm benchmark error is a systematic vertical offset. If BM2 is 25 mm high, then levels set from BM2 will be 25 mm high everywhere, not just near BM2. That reduces freeboard and can reverse the intended fall if the error is combined with setout error. The 30 to 40 mm stake discrepancy is larger than the benchmark misclosure alone, so there may also be a setout or machine-control error. The design levels must be recomputed from verified control, and the as-built surface must be compared with the design before accepting any work.
3. Recommend whether to accept, rework or verify the 0+850 section, and what records you would provide.
Read the worked answer
I would not accept the 0+850 section on the current evidence. I would first verify it by an independent survey from a checked benchmark, measuring the invert level, channel grade and freeboard. If the as-built invert is within the specified tolerance and the freeboard is still at least 0.3 m, the section may be accepted with a non-conformance note and a record of the verified levels. If the invert is outside tolerance or the grade is affected, the section should be reworked before the next construction stage. I would provide the design levels, the control check, the benchmark misclosure, the independent as-built survey, the machine-control file check and a clear recommendation to the engineer. I would also require BM2 to be re-established or replaced before further setout. The record should show that the decision was based on verified survey evidence, not on the contractor’s program pressure.