Smoke Without Fire: How Long Can Thermal Cracking Sustain Hydrothermal Circulation in the Absence of Magmatic Heat?

Long‐lived hydrothermal circulation is now well documented along slow and ultraslow spreading mid‐ocean ridges, even though these settings only receive a moderate and intermittent supply of magma. This challenges the notion that hydrothermal convection must be sustained by a continuously replenished magma body. Here we investigate the possibility of sustaining hydrothermal circulation by infrequent magmatic intrusions separated by episodes of downward propagation of small cracks enabling fluids to tap heat from deep hot rocks. We focus on cracks nucleating from grain boundary flaws in response to the buildup of cooling stresses at the base of the convection system. We develop an analytical model describing the stable propagation of a percolation front and the associated heat transfer through hydrothermal circulation. Convection ceases when thermoelastic stresses can no longer overcome lithostatic pressure and hydrothermal circulation can no longer mine heat from underlying units. For lithospheric permeabilities greater than ∼10 −15  m 2 and over a wide range of grain and flaw sizes, this occurs within ∼100 kyr to ∼1 Myr from the onset of cracking, after the cracking front has moved by a few kilometers. We validate this analytical prediction by developing two‐dimensional numerical models of porous convection subjected to a lower boundary condition representing the dynamics of the cracking front. These models suggest that moderate‐ to low‐temperature hydrothermal venting in off‐axis, ultramafic‐hosted sites does not require an underlying magma sill at all times, but instead repeated sill intrusions every ∼10–100 kyr, which periodically reinvigorate hydrothermal convection without building a continuous crustal section.