Petrophysicist interview questions

I've been a petrophysicist for ten years, six in Oman. Started at an Indian operator on conventional sandstone reservoirs, then moved to PDO where I've worked on three major field types: mature waterflood sandstones, heavy oil thermal recovery, and most recently tight carbonate reservoirs. I cover the full workflow: log interpretation, core-log integration, rock-typing, saturation modelling, and increasingly the integration of machine learning for log prediction. I hold an MSc in petroleum geosciences and SPWLA membership.

Start with the core itself: depth-shift to logs, identify the rock types present (visual and CT scan), measure key petrophysical properties (porosity, permeability, capillary pressure, electrical properties). Calibrate the log responses against the measured core data; for example, deriving the right Archie parameters m and n from electrical-property measurements, not generic values. Build rock-typing schemes from core that you then apply to log data in uncored intervals. The output is a much better petrophysical model than logs alone could produce. Core is expensive; using it well is part of the job.

Standard Archie equation assumptions break down in complex reservoirs: shaly sands need shale-correction (Waxman-Smits, dual-water), carbonates often need rock-typing-specific approaches, low-resistivity pay zones need special treatment. My approach: identify which complications apply to the reservoir, choose the appropriate model, and validate against production data and capillary pressure curves. I quantify uncertainty: range of saturations consistent with the data, not a single value. For reserves calculations I use the range with a specific confidence interval (P50 typically, P10 and P90 for sensitivity).

ML is a tool for specific problems, not a replacement for understanding. I've used it for log prediction (predicting NMR responses from conventional logs when NMR isn't available), rock-typing classification (clustering log signatures to identify rock-type boundaries), and uncertainty quantification (ensemble methods showing the range of plausible interpretations). I'm careful about training data quality and over-fitting; ML models that look good on training data and fail on new wells are common. I also validate ML predictions against physics-based interpretation; if they disagree, I dig into why before trusting either.

Multi-disciplinary collaboration is the heart of subsurface work. I attend the geomodel review meetings even when not formally required; petrophysics needs to align with the geological interpretation. I share my interpretation results in workshops, not just deliverables; getting other disciplines into the room while my thinking is forming catches inconsistencies early. When a reservoir engineer's history match doesn't work, my saturation model is often part of the answer; I'm willing to update interpretation based on dynamic data. Subsurface studies fail when disciplines silo.

For a senior petrophysicist role with major operator experience in Oman I'd target OMR 2,300 to 2,800 total package depending on the bonus structure and asset criticality. Roles on flagship fields or with significant exploration scope command a premium. I'm on 90 days' notice. Beyond pay I'd value the technical scope; my career is built on the studies I've led, so a complex tight carbonate work programme is more valuable than routine sandstone analysis even at lower pay.