The Carbonate Problem
Sandstone rules are so convenient that half the industry forgets they are local bylaws. Carbonates, holding half the world's oil, obey none of them: porosity made by biology and re-plumbed by dissolution, velocities ruled by pore shape rather than pore volume, Archie exponents that wander with the pore type, and permeability that laughs at the crossplot. Same logs, same physics, entirely different rulebook, and this path teaches the rulebook.
You can read a carbonate's depositional factory and its diagenetic overprint, model pores as inclusions and say why shape beats volume, classify the pore types and predict which hides from the sonic, run the M-N and matrix plots to unmix the mineralogy, choose saturation exponents with the pore system in mind, type the rock by flow rather than lithology, and treat the pre-salt microbialites as the exam the whole path prepared you for.
A different rock
Carbonate is grown in place by biology, reefs, platforms, microbial mats, so its architecture follows ecology and sea level, not river hydraulics; the depositional rules differ from the first grain.
Sandstone thinking treats porosity as an amount; inclusion models treat each pore as a shape in a solid, and shape is the variable that lets one porosity produce two velocities a kilometer apart.
Carbonates dissolve and reprecipitate on geologic whim: vugs and molds open after burial, invisible to the sonic that rides the matrix around them; secondary porosity is the diagenetic rewrite of the plumbing.
Interparticle, moldic, vuggy, micro: the pore taxonomy is the carbonate's real identity card, and placing a sample in it predicts stiffness, flow, and which logging tool will be fooled.
Logging the unruly
Calcite, dolomite, anhydrite, and quartz mix freely in carbonate sections, and the multimineral solver is the machinery that unmixes them; the tool that turns three logs into a mineralogy.
M and N collapse three porosity tools into two lithology-sensitive numbers: the classic crossplot that separates the carbonate minerals and flags secondary porosity while it is at it.
Density porosity is hostage to the matrix density you assumed, and carbonates change theirs mid-section; the matrix plots read the true grain density so the porosity stops inheriting a sandstone default.
Archie's m and n were calibrated on well-behaved sandstone; moldic pores push m toward three and fracture systems pull it under two, and the saturation error from a borrowed exponent can be the whole field.
The poro-perm crossplot that serves sandstone collapses in carbonates, where a vug network and a micropore mat share a porosity but differ in permeability by four orders; flow must be predicted from pore type, not pore count.
The carbonate mindset
The carbonate answer to heterogeneity: group the rock by how it flows, not what it is called, and the flow units carry the reservoir model where lithology labels cannot.
The case that assembles the mindset: equal porosities, wildly unequal velocities, resolved only by pore shape; the sandstone reflex fails and the inclusion picture succeeds, which is the whole lesson in one plot.
Lula's microbialite reservoirs are the final exam: a biological factory nobody had mapped, pore systems that defy the taxonomy, under two kilometers of salt; every carbonate skill in the path reports for duty at once.