Grid, Ground, and 0.16 Feet of Sewer
Your firm is providing layout control for 2,400 ft of 8-inch sanitary sewer in a state plane zone. The design engineer's plan coordinates are state plane grid values, but pipe lengths, slopes, and manhole spacing were computed on the ground. At the site, the zone's mapping angle is negligible, the published zone scale factor is 0.99993, and the site is 1,250 ft above the ellipsoid (use mean earth radius 20,906,000 ft). The contractor's tolerance for structure location is ±0.10 ft. The plans show a 1,200.00 ft reach between two manholes. A record control monument nearby has a published state plane coordinate with stated accuracy 0.05 m (0.16 ft), but your GNSS-derived value for it differs by 0.09 m north and 0.04 m east. The project schedule allows one mobilization for layout.
1. Compute the elevation factor, the combined factor, and the grid-versus-ground difference over one 1,200 ft reach and over the full 2,400 ft. Does it matter against the ±0.10 ft tolerance?
Read the worked answer
Elevation factor: R/(R+h) = 20,906,000 / (20,906,000 + 1,250) = 20,906,000 / 20,907,250 = 0.99994.
Combined factor: 0.99993 × 0.99994 = 0.99987.
Ground distance corresponding to a grid distance: ground = grid / CF = grid × 1.00013 → about 130 ppm.
- Over 1,200.00 ft of grid: 1,200.00 × 130 × 10⁻⁶ = 0.156 ft.
- Over 2,400 ft: 0.31 ft, cumulative.
Yes, it matters. The single reach already exceeds the ±0.10 ft structure tolerance by more than 50%, and the error accumulates along the line. Setting stakes at grid distances would push each successive manhole progressively further off, and the pipe slope computed from a 1,200.00 ft plan length would be installed over a 1,200.16 ft ground length. The two effects are separate: coordinates are consistently shifted (a systematic scale), but individual structure spacing is what the contractor checks with a tape.
2. What is your technical fix, and how do you document the basis of coordinates and bearings so the contractor and engineer are not misled?
Read the worked answer
Preferred fix - publish a project ground coordinate system: choose a project origin near the site, scale all state plane coordinates about that origin by 1/CF (1.00013), and issue a control table in both systems with the transformation stated. Advantages: distances computed between coordinates equal ground distances, so the contractor's tape and the total station agree everywhere on the job, and the tie to the state plane is preserved for GIS and record.
Acceptable alternative: keep state plane coordinates and require every layout distance to be scaled by 1/CF at the time of computation, with the scale factor printed on each stakeout sheet. This works but is a standing invitation for someone to use a raw inverse.
Documentation that must accompany either:
- The datum/realization and epoch (e.g., NAD83(2011)), units (U.S. survey feet vs. international feet - a 2 ppm trap that also exceeds this tolerance on long lines), and the zone.
- The basis of bearings, with meridian convergence and arc-to-chord (t−T) applied or explicitly noted as negligible at this site.
- The combined factor, the project origin, and a worked example of the conversion.
- A statement that plan distances/slopes are ground values and coordinate inverses will not match unless the factor is applied.
Do not leave the engineer's grid coordinates and the contractor's ground tape as two unlabeled systems on the same job.
3. The record monument disagrees with your GNSS-derived position by 0.09 m N and 0.04 m E. How do you handle it before holding control for layout?
Read the worked answer
0.09 m ≈ 0.30 ft, roughly twice the stated 0.05 m (0.16 ft) accuracy - this is a flag, not a rounding issue, and layout should not proceed on a point whose status is unresolved.
Diagnostic order:
- Datum/epoch mismatch. A NAD83(1986) or NAD83(CORS96) value compared with NAD83(2011) can differ at the decimeter-to-meter level in some regions. Check the published metadata before assuming a blunder.
- Published accuracy and source. A 0.05 m value from a third-order traverse is not equivalent to a modern GNSS value; the discrepancy may be entirely within the monument's own error budget plus yours.
- Your solution quality. Fixed vs. float, PDOP, observation duration, number of redundant vectors, antenna height, and any nearby obstruction/multipath. A 0.09 m horizontal outlier with a short single-vector observation is more likely yours than theirs.
- Monument condition. Disturbed, leaning, or reset monument; check ties, witness posts, and any remonumentation record.
Action: re-observe the monument with a redundant session (or a static/OPUS-style occupation) and use at least two other control points to check the network internally. Do not force the network to a single point by translating everything to match it - that hides the problem and corrupts the rest of the control. If the discrepancy persists, report it in writing and on the deliverable, note which control you held and why, and let the engineer decide with the facts in hand.