BHA Misalignment

BHA alignment and survey error

Directional sensors measure the orientation of the sensor package. Gravity, borehole curvature, and bent assembly components can make its axis differ from the local borehole axis. The amount depends on the assembly geometry, materials, support locations, and hole conditions.

Misalignment in the highside plane produces an inclination error. A lateral component can also produce an azimuth error. The signed inclination correction is therefore different from the total angle between the sensor and borehole axes.

For example, a constant 0.1° inclination bias over a 3,000 m horizontal interval produces approximately 5.2 m of TVD error. This comparison assumes the same entry position and otherwise identical measurements. It illustrates accumulated bias, rather than the placement error of a particular well.

Sag errors can remain correlated across stations and accumulate along the well. Their magnitude and sign can change as assembly and hole conditions change. Repeated surveys alone do not establish that the error has cancelled.

Sensor calibration does not establish alignment with the borehole axis. Normal total-field, gravity, and dip checks can pass despite BHA misalignment. Mechanical modeling estimates a correction; independent borehole measurements can help assess that estimate. See the ISCWSA field-acceptance guidance.

Estimating the correction

TALLENTECH uses a finite-element model of the recorded assembly. Component dimensions, material properties, sensor position, stabilizers, and motor bends describe the BHA. Supplied inclination, toolface, fluid density, and estimated hole size define the survey conditions. The model estimates deflection and wall contacts, then evaluates the signed inclination correction at the sensor position.

The 3-D solver accepts separate build and turn inputs and can report predicted lateral misalignment and azimuth error. Current automatic survey processing supplies local build curvature with turn set to zero and applies the accepted sag correction to the inclination solution. Turning-hole analysis therefore requires the corresponding 3-D inputs and review. The predicted azimuth effect is a diagnostic output, rather than an automatically applied azimuth correction.

Inclination errors can affect northing, easting, and TVD. Their position effects depend on the trajectory. Highside and vertical directions coincide at a horizontal station, but differ elsewhere along the well.

An illustrative uncertainty comparison

The figure uses TALLENTECH's EOU engine on the published ISCWSA #1 extended-reach test profile, with 8,000 m measured depth. It compares the generic MWD+HRGM and MWD+HRGM+SAG error models.

At horizontal TD, the 2σ highside and vertical uncertainty values decrease from 45.3 m to 26.3 m, approximately 42%. The lateral value remains 182.0 m in this example. These values are semi-axis measures, rather than full ellipse widths.

This is an error-model comparison. The figure calculates uncertainty under two sets of assumptions; it does not measure the BHA solver's correction accuracy for a particular well. Applying a station correction can change the calculated trajectory, while changing the error model changes the estimated uncertainty that remains.

Inputs and model limits

Several inputs interact to determine the correction:

• Stabilizer placement and gauge are important controls on support and contact locations.
• Stabilizer wear and hole size change clearances and contacts. They can change the correction's magnitude or sign. Expected wear is a supplied estimate, and an assumed hole size may differ from the measured borehole.
• Fluid density affects buoyancy and effective assembly weight. Stiffness and contact conditions then control deflection.
• Inclination changes the gravity component across the assembly.
• Toolface changes the orientation of motor bends and other assembly features relative to gravity and curvature.
• Build and turn rates describe local curvature over the modeled assembly span. The model does not reconstruct unmeasured curvature changes from those rates alone.
• Component dimensions and material properties determine stiffness and weight. Steel, monel, and titanium components can respond differently.
• Sensor position, operating loads where modeled, and bit constraints affect the predicted alignment. The selected conditions must represent the survey measurement.

A mechanical model and a correction table both need an applicable operating range. A model can also give a poor estimate when its inputs or assumptions do not represent the BHA and borehole. Review input records, contact behavior, solver status, and sensitivity to uncertain conditions before accepting a correction.

Residual uncertainty and records

A calculated correction does not remove all sag uncertainty. Uncertain hole size, wear, curvature, loads, and model assumptions can leave a residual error. Sensitivity analysis and independent validation help assess that error. The uncertainty model must represent the remaining error for the accepted correction method and survey conditions.

The example above uses generic error-model assumptions. It does not establish a 42% reduction for another well or justify assigning a smaller sag term solely because a solver returned a correction. See the ISCWSA error-model definition and Studer and Macresy's residual-error methodology.

For review and definitive-survey use, record the model version, assembly and sensor geometry, supplied conditions, estimated inputs, correction sign convention, solver status, and application status. Retain the basis for the residual uncertainty assessment with those records.

References

• ISCWSA: Error-model definition, Rev5.13, including borehole misalignment, sag, and the diagnostic test profiles.
• ISCWSA: Field acceptance criteria, including the limits of normal QC checks for tool misalignment.
• Studer and Macresy (2006): SPE 102088 abstract, describing 3-D BHA modeling and residual sag-error evaluation.