RTK Setting-Out Discrepancy on a Mandeville Pipeline
You are responsible for setting out a 2.4 km sewer rising main in Mandeville. The design engineer supplies coordinates in the Jamaica National Grid and a combined scale factor of 0.99987 for the pipeline corridor. Your RTK rover logs ITRF2014 geographic coordinates; the office transforms them to grid using a published seven-parameter transformation and a geoid model. On two published control stations, check shots differ from the design coordinates by +0.35 m north and -0.22 m east; ellipsoid heights differ by 0.80 m. The base station is set over a station whose grid coordinate you took from an old plan, and the manufacturer's default antenna height is 1.80 m while your pole is set to 2.00 m. The contractor wants to start trenching tomorrow. A total station is available, and the two control stations are 900 m apart with intervisible rooftops.
1. Diagnose the likely causes of the horizontal and vertical discrepancies. Which causes would you check first, and why?
Read the worked answer
The horizontal differences of +0.35 m north and -0.22 m east are far too large for normal RTK random error and suggest a systematic translation or transformation problem rather than noise. I would check first the base-station coordinate, because an error or a wrong station coordinate translates the whole RTK solution. Next I would check the ITRF2014-to-grid transformation: applying the seven parameters in the wrong direction, using the wrong epoch, omitting the transformation, or misapplying the combined scale factor can all produce metre-level or decimetre-level shifts. The antenna-height discrepancy of 1.80 m versus 2.00 m is mainly a vertical issue unless the pole is tilted or the antenna reference point is misunderstood, so it is a strong candidate for the 0.80 m height difference but not for the horizontal shift by itself. The vertical difference could also be geoid versus ellipsoid height confusion or a wrong geoid model. Poor RTK fix, weak satellite geometry, multipath, and unreliable old-plan control coordinates are secondary possibilities. I would not start trenching until the systematic cause is isolated.
2. Design a verification procedure using the available RTK and total station before any setting out is released.
Read the worked answer
Occupy each control station with the RTK and with the total station. Compare grid coordinates and heights, verify the antenna height and antenna type, and confirm that the transformation and geoid model are the ones intended for the project. Recompute the grid distance between the two control stations from their coordinates and compare it with the ground distance multiplied by the combined scale factor; this checks the projection and scale. If possible, run a static or rapid-static session on the base station and process it against independent control. Use the total station to establish two local control points near the start of works, tied to the verified stations, and set out a test point that is then checked by total station. Record every residual. If the RTK cannot be reconciled with the total station and the published control, use the total station for setting out and keep GNSS for reconnaissance only. The verification must be documented, not simply asserted.
3. What decision should you make about tomorrow's trenching, and how should you communicate it to the contractor and client?
Read the worked answer
I would not release trenching on the current RTK coordinates. The horizontal discrepancy is larger than normal construction tolerance for a rising main and could place the pipeline outside its easement, into other services, or at the wrong invert. I would tell the contractor to hold that activity and explain the checks already done and the release condition. If the total-station local control agrees with the design within tolerance, setting out can proceed from that control; otherwise the design coordinates or control must be referred back to the engineer. I would give the client a short technical note stating the coordinate system, transformation, scale factor, epoch, geoid model, check residuals, and the basis for the decision. This is more defensible than a vague assurance and protects both the client and the contractor from a costly error.