Shaly Sands Evaluation

SPWLA Region Conference March 2020, Bangkok

— DRAFT —

ABSTRACT

 

How to do Quick-Look Shaly Sand Log Analysis with Confidence

 

The operations petrophysicist faced with the daunting task of producing fit-for-purpose Quick-Look results in the freshwater, stacked shaly sands of S.E. Asia will understandably feel challenged.  The methods he/she adopts typically fail to provide the necessary information to distinguish hydrocarbon zones and permeability. Rather than using porosity/conductivity log analysis – a tool intrinsically unsuited to the task at hand – this paper sets out simple, cross-disciplinary techniques which cross-check and harden Quick-Look log analysis.  Commonly available special log, non-log and generic data are employed to this end.

 

First, a density-neutron total porosity, largely immune to common lithology and fluid effects is computed. If NMR, rotary sidewalls or offset core are available these data are used to calibrate and supplement the fluid density input. Alternatively PetroDB’s* default log driven fluid density equations are used.

 

Porosity is now used to compute an apparent salinity curve from standard Archie, and the maximum salinities used for Rw. This is analogous to a Pickett plot Rw but may be replaced by confident local knowledge of Rw.  A generic S.E. Asia shaly sand variable m formulation is now used for a temporary water saturated resistivity curve (Ro_temp).

 

Zones above FWLs, if any, are established by a single valid pressure; mudgas; shale corrected density-neutron separation or subtle departures of logged Rt over Ro_temp. If FWLs remain uncertain the most promising zone is assumed to be above a FWL.

 

With salinity and FWLs fixed the apparent standard Archie m curve is computed below FWLs. This m quantifies the presence of non-Archie (excess) conductivity. This m is now predicted (mv) from conventional logs and used in this paper’s Shaly Archie formulae.  mv is also used to compute the continuous Waxman Smits Qv and Dual Water Model shaly sand inputs.  If available, standard output NMR curves are employed to compute these inputs, providing certain criteria are met. We have now established water saturated resistivity (Ro), a critical but often neglected requirement.

 

If available, standard output NMR is now used for a resistivity independent Sw.  Alternatively, sidewall, offset or generic PetroDB core data and chart K-4 provide Swi.  If neither is available PetroDB’s generic vshale, porosity and tvdss provide permeability, Swi and full Saturation Height functions, where Sw>Swi. Under certain conditions dielectric curves may also provide Swi.

 

With a resistivity independent Sw now available the Quick-Look process calibrates the saturation exponents (n) for the three saturation equations.  If n values are non-feasible the assumption of FWL is revisited.

 

A final pass allows interactive adjustment of salinity, FWL, mv and n for Sw. Bulk volume water then calculates Timur Coates Permeability, which allows reconciliation of the entire process against any available core permeability: sidewall, offset or PetroDB.

 

This process is coded into a log analysis command file and completed within 3-4 hours. The diversity of inputs used affords a degree of confidence simply not possible with conventional shaly sand log analysis.  This method also checks and calibrates existing wells and geo-models.

 

The paper recommends the optimal data acquisition to enhance this Quick-Look technique.

 

____________

Dr Mark Deakin

PETROPHYSICS Pty Ltd