Runway Grade Control: A Good Loop Closure and a Datum Mismatch
A 4.5-acre runway extension is to be graded to subgrade elevation 812.50 ft NAVD88. The survey crew runs a 12-turn differential level loop from BM A (NAVD88, 812.44 ft) back to BM A, a distance of 3,200 ft. The loop misclosure is 0.048 ft, which the crew accepts as 1:66,700. A side run from the loop to BM B (NGVD29, 815.09 ft) gives a difference that the contractor uses for grade control. At station 12+00 the contractor's grade check, using BM B, shows the existing ground at 812.61 ft, or 0.11 ft above the design subgrade. The contractor submits a change order for 800 CY of extra excavation. A check of the NGS datasheets shows BM A is a NAVD88 benchmark and BM B is a 1970s NGVD29 benchmark with no stated conversion. The crew's closure was computed only from BM A back to BM A.
1. Is the 0.048-ft loop closure acceptable, and does it validate the 0.11-ft elevation difference between the benchmarks?
Read the worked answer
The loop closure is good: 0.048 ft over 3,200 ft is about 1:66,700, which is well within typical differential-leveling standards for construction control. But a good closure only shows that the random errors in that loop are small and that no large blunder occurred within the loop. It does not validate a side run to a different benchmark, and it does not test the datum of BM B. The loop began and ended on BM A, so it never checked BM B's published elevation.
The 0.11-ft difference is therefore not a closure problem. BM A is NAVD88 and BM B is NGVD29. The two vertical datums are not the same, and differences of tenths of a foot are common. Comparing an NGVD29 elevation directly with a NAVD88 design elevation is a datum error. The 0.11 ft should be investigated as a datum or benchmark issue, not distributed as if it were random leveling error.
2. Quantify the effect of the 0.11-ft difference on the excavation quantity and evaluate the 800-CY change order.
Read the worked answer
One acre is 43,560 square feet, so 4.5 acres is 196,020 square feet. A uniform depth difference of 0.11 ft over that area is 196,020 x 0.11 = 21,562 cubic feet. Dividing by 27 cubic feet per cubic yard gives about 799 cubic yards.
That is essentially the 800 CY in the change order. The arithmetic strongly suggests the change order is driven by the datum mismatch rather than by a real 0.11-ft average overexcavation. However, the conclusion is not automatic: the actual ground surface is not perfectly uniform, and the true cut depends on the corrected elevations at each cross section. The point is that the quantity cannot be accepted until the elevations are all on NAVD88. If the contractor's grade check is 0.11 ft high because of the datum, the actual cut may be less than claimed, not more.
3. What steps would you take before certifying grades or quantities?
Read the worked answer
I would stop using BM B for NAVD88 grade control until its relationship to NAVD88 is established. I would obtain the full NGS datasheets for both benchmarks, check the date and adjustment of each elevation, and apply a published datum transformation such as VERTCON if appropriate. I would also run a new closed loop between at least two reliable NAVD88 benchmarks, with turning points chosen to avoid soft ground and with equal sight distances. If BM B must be used, I would establish its NAVD88 elevation by a direct differential run from BM A, not by assuming the NGVD29 value is equivalent.
After the datum is resolved, I would recompute the design grades and the earthwork quantities from the corrected field data. I would document the datum discrepancy, the source of each benchmark elevation, and the correction applied. I would not certify the 800-CY change order based on the existing comparison, because a good loop closure on the wrong datum can still produce a systematic elevation error.