Raw surveys in, a definitive well position out
Every MWD survey leaves the rig carrying systematic errors: reference assumptions that don't match the local magnetic field, sensor biases and scale factors, and an assembly that bends the sensor away from the borehole axis. Survey Correction and QC is TALLENTECH's core engagement: we take the raw station data, apply the accuracy enhancements the well needs, quality control every station as it arrives, and return a definitive survey with quantified uncertainty.
The result on the published ISCWSA extended-reach test well, comparing an uncorrected survey against the referencing-plus-MSA chain:
| At TD · 2σ | Lateral | Vertical |
|---|---|---|
| Uncorrected, standard global modelMWD+SRGM | 190.5 m | 45.3 m |
| Referencing + multi-station analysis appliedMWD+HRGM+MS | 113.1 m | 45.3 m |
| Change | −41% | no change |
The correction chain
Each enhancement in the chain removes a specific error source, and they are applied in the order the physics demands:
- Reference first: Main Field Model referencing establishes declination, dip, and total field per wellsite; IFR1 adds the measured crustal correction where field data exists; and IFR2 corrects each survey at its timestamp with live observatory data.
- Sensors second: Multi-Station Analysis solves bias, scale-factor, and orthogonality corrections from the survey data itself.
- Geometry third: sag correction puts the inclination back on the borehole axis; high-frequency surveying resolves the trajectory between stations where the well demands it.
Not every well needs every step. Referencing is the foundation on all of them; the rest are added where latitude, geology, well geometry, or tolerance requires it. We recommend the stack per well plan and document the choice in the survey program.
QC on every station, not just at TD
Corrections are only half the engagement. As each survey arrives it is checked against the error models listed in the survey program: given the instrument performance model for that hole section, total field, gravity, and dip must each fall inside a predicted acceptance envelope. Meeting those limits is what keeps anti-collision and geological target analysis valid. A station that fails is flagged, with what failed and by how much.
Transparent corrections
Every correction is documented:
- which reference was used
- which model vintage
- which corrections were applied
- what changed