Measurement After Drilling

Key Takeaways

• The data rate limitation of mud pulse telemetry is the fundamental reason that MAD data provides dramatically better resolution than real-time LWD data: mud pulse telemetry systems transmit data by creating pressure pulses in the circulating drilling mud column using a valve or siren in the MWD tool, and the physics of pressure pulse propagation through drilling mud at typical drill string lengths (3,000-10,000 meters) limits the practical data rate to approximately 1-12 bits per second for standard systems, improving to 30-60 bits per second for advanced electromagnetic telemetry systems; at these data rates, deep-reading resistivity, density, neutron porosity, and gamma ray measurements can only be transmitted at a rate of approximately one sample per 0.5-2.0 meters of depth; the same sensor operating in memory mode can record data at 5-30 times higher frequency (effectively one sample every 10-20 centimeters of depth), capturing bed boundaries and formation heterogeneity that the real-time telemetry resolution misses entirely.
• The risk of MAD data loss due to tool failure or memory data corruption is the most significant downside of the MAD approach compared to real-time telemetry: if a logging-while-drilling tool fails mechanically during the bit run (from shock, vibration, mud invasion, or electronic component failure), the memory data stored in the tool is typically lost along with the tool, and the formation evaluation information for the affected interval cannot be recovered; real-time telemetry provides a continuous stream of data that is recorded at surface as it is received, so that partial data is available even if the tool fails later in the run; for this reason, MAD data is typically used as a supplement to real-time data rather than as a replacement, with real-time data providing the operational decision support (formation tops, pore pressure indicators, drilling hazard detection) and MAD data providing the high-resolution petrophysical analysis after the bit run is complete; in wells where real-time data quality was poor but memory data was successfully recovered, MAD data sometimes provides the only adequate formation evaluation for the well.
• Depth correlation between MAD data and real-time data is often imperfect because the depth at which measurements are recorded in the tool's memory is derived from the drilling depth (the depth of the drill floor kelly bushing minus the length of the drill string), which includes errors from pipe stretch, temperature-induced length changes, and the discrete nature of stand-length depth measurements; when the MAD data is downloaded and depth-matched to the well's master depth record, small depth shifts (typically 0.1-1.0 meter) may be required to align features in the memory data with their correct geological depth; this depth reconciliation process uses sharp formation boundaries visible in both the real-time data and the high-resolution memory data as control points, and the depth-corrected MAD data is then merged with the real-time dataset to produce the best-available composite log for the well; the composite log represents the combination of the real-time data's operational reliability with the memory data's petrophysical resolution.
• MAD data from rotary steerable system (RSS) runs often provides the highest-quality formation evaluation data of any LWD approach, because RSS drilling produces a smooth, low-tortuosity wellbore that allows LWD tools to maintain stable positioning relative to the formation during measurement, unlike motor-slide drilling where tool vibration during the slide phase can degrade measurement quality; in long horizontal wells drilled with RSS, MAD data has proven particularly valuable for geosteering quality assurance — reviewing the high-resolution memory resistivity and gamma ray data after the run to confirm that the well path followed the intended reservoir target, identify intervals where the well exited the pay zone (which the real-time data may have detected but at insufficient resolution to precisely locate the exit point), and plan corrective steering for subsequent bit runs.
• Memory data from formation tester tools (MDT-type or RCI-type formation testing while drilling tools) captures the complete pressure transient response during each station test at full time resolution, which is not transmitted in real time due to data rate constraints; the full pressure buildup curve recorded in memory allows the reservoir engineer to perform pressure transient analysis (calculating the formation's permeability-thickness product, the skin factor, and the pore pressure) using the same techniques applied to conventional wireline formation tester data; real-time telemetry from the same tool can only transmit a few pressure sample points during the test, giving the driller confidence that a test was conducted but not providing the data quality needed for rigorous petrophysical interpretation; downloading the memory data after tripping the BHA thus converts a real-time qualitative pressure measurement into a quantitative reservoir characterization tool equivalent in data quality to a wireline formation pressure test.