Systematic patterns of detrital K-feldspar decline with increasing burial depth are revealed during deep burial (1.5–4.5 km) of sandstones in diverse sedimentary basins from three rift and two passive margin settings in the North Sea and USA Gulf Coast. K-feldspar destruction is predominantly by dissolution, though in-situ alteration to albite also occurs. Dissolution can remove up to 15% solid volume of the rock by 4.5 km, so that primary arkosic sandstones become diagenetic quartzarenites at depth, with increased secondary:primary porosity ratios. The quantity of K-feldspar destroyed during temperature increase is very variable: 0.05–1.0% solid volume/°C. Compaction continues during deep burial and petrographic textures diagnostic of secondary porosity are destroyed (so that many sandstones preserve <10% secondary porosity). Alternative explanations for K-feldspar reduction with depth, including changes in original sediment provenance controlling the observed variation of feldspar with depth, or that source areas changed during deposition, can be refuted. Deeply buried sandstones are not reliable records of depositional composition.
Exceptions to this scenario of detrital K-feldspar dissolution are noted. Proximity of K-rich evaporites, high SiO2 and K+ activity due to the breakdown of volcanic clasts or very low PCO2 can stabilize both K-feldspar and a suite of K-rich cements.