Error (Measurement and Survey)

Key Takeaways

• Wellbore position error accumulates along the wellbore and creates an expanding uncertainty ellipse with depth — every survey measurement (inclination and azimuth at each survey station) has an associated measurement error, and as these errors are integrated along the wellbore trajectory calculation, the positional uncertainty grows with depth; near surface, the wellbore position may be known to within centimeters of the GPS-surveyed wellhead; at 10,000 feet of depth in a deviated well, the positional uncertainty ellipse may be 30-100 feet in the lateral direction and 5-15 feet in the vertical direction, depending on the survey tool quality and the wellbore geometry; the ISCWSA (Industry Steering Committee for Wellbore Survey Accuracy) error model provides a standardized framework for computing wellbore position uncertainty from the individual instrument error sources, enabling operators to design multi-well pad spacing and anti-collision rules that maintain adequate separation with known probability even when the exact wellbore position is uncertain.
• Systematic calibration errors in logging tools create depth-coherent biases that are more dangerous than random errors — a random measurement error tends to average out over many measurements and produces scatter in the log reading; a systematic error (such as an incorrectly calibrated density source count rate, a neutron tool with an incorrect matrix response, or a depth measurement system with incorrect cable calibration) produces a consistent offset that biases every reading in the well by the same direction and approximate magnitude; systematic errors are particularly insidious because they are not obvious from the data alone (the log looks consistent and clean) and may not be detected until the log data is compared against core measurements or offset well data from a different service company; pre-job tool calibration against API standard formations, comparison of log readings against core at the same depth interval, and cross-validation of porosity logs against each other (neutron-density cross-plot consistency checks) are the primary methods for detecting and correcting systematic calibration errors before they propagate into reservoir characterization and reserve calculations.
• Depth error between logging runs and perforating runs can cause completions to target the wrong intervals — wireline depth measurement relies on the length of cable paid out from the surface drum, corrected for cable stretch (which depends on tool weight and cable tension, varying with depth) and temperature effects on cable length (steel cables change length with temperature); if the depth measurement system is not properly calibrated for cable weight, cable tension, and temperature, the recorded depth may differ from the true measured depth by several feet in a shallow well or tens of feet in a deep well; in a completion where the target perforations are in a specific 10-foot interval identified on the gamma ray log, a 5-foot depth error in the perforating gun run (due to different cable stretch assumptions than the logging run) will perforate the wrong interval — a very common and very expensive completion error that has been documented many times in the industry; correlating a GR collar locator run with the perforating gun against the original log collar positions is the standard quality control method for depth matching between logging and completion operations.
• Reserve calculation errors from measurement uncertainties compound through the reserve volumetric calculation — the probabilistic reserve calculation V_HCPV = A × h × ø × (1-Sw) × recovery factor involves several terms each with their own measurement uncertainty; porosity (ø) from wireline logs has typical uncertainties of ±1-2 porosity units; water saturation (Sw) from resistivity has uncertainties of ±5-10% saturation units; net pay thickness (h) depends on permeability cutoff choices that have substantial uncertainty; area (A) from seismic and well control has uncertainty that dominates in undrilled areas; and recovery factor is often the most uncertain parameter; Monte Carlo uncertainty analysis that propagates individual measurement errors through the reserve calculation produces a probabilistic reserve distribution (P10, P50, P90) that is more honest about uncertainty than the single-point deterministic estimate and is required for SEC reporting and portfolio risk management in major oil companies.
• Measurement while drilling (MWD) survey errors are particularly consequential in geosteering operations — in horizontal wells drilled to stay within a specific reservoir interval (geosteering), the driller uses real-time formation evaluation data (gamma ray, resistivity) to maintain the wellbore within a target interval that may be only a few feet thick; if the MWD survey system has inclination errors of ±0.1° (typical for magnetic MWD tools near magnetic interference from steel casing), these errors translate to uncertainties in the calculated depth below the formation top that may be comparable to the target interval thickness; in such cases, the formation evaluation log (gamma ray trending down or up) becomes the primary geosteering indicator, with the survey being used for directional control and the petrophysical response being used for stratigraphic depth control; advanced gyroscopic MWD systems with uncertainties of ±0.025° provide more accurate position relative to the formation top and are preferred in precision geosteering operations in thin reservoirs.