A two-dimensional, one by two-kilometer section through the seafloor was simulated with a numerical model to investigate coupled fluid and heat flow resulting from basalt intrusions in a buried spreading center. Boundary and initial conditions and physical properties of both sediments and basalt were constrained by field surveys and drilling in the Guaymas Basin, central Gulf of California. Parametric variations in these studies included sediment and basalt permeability, anisotropy in sediment permeability, and the size of heat sources. Faults were introduced through new intrusions both before and after cooling.

Background heat input caused fluid convection at velocities ≤ 3 cm a⁻¹ through shallow sediments. Eighty to ninety percent of the heat introduced at the base of the simulations exited through the upper, horizontal surface, even when the vertical boundaries were made permeable to fluid flow. The simulated injection of a 25–50 m thick basalt intrusion at a depth of 250 m resulted in about 10 yr of pore-fluid expulsion through the sea-floor in all cases, leaving the sediments above the intrusions strongly underpressured. A longer period of fluid recharge followed, sometimes accompanied by reductions in total seafloor heat output of 10% in comparison to pre-intrusion values. Additional discharge-recharge events were dispersed chaotically through the duration of the cooling period. These cycles in heat and fluid flow resulted from the response of the simulated system to a thermodynamic shock, the sudden emplacement of a large heat source, and not from mechanical displacement of sediments and pore fluids, which was not simulated.

Water/rock mass ratios calculated from numerical simulations are in good agreement with geochemical estimates from materials recovered from the Guaymas Basin, assuming a bulk basalt permeability value of at least 10⁻¹⁷ m²/(10⁻² mD). The addition of faults through intrusions and sediments in these simulations did not facilitate continuous, rapid venting. Increased heat input at the base of the faults resulted in temporarily greater fluid discharge, but the flow could not be sustained because the modeled system could not recharge cold fluid quickly enough to remove sufficient heat through the vents.