• We study nucleate boiling on substrates of small thermal capacity using IR camera. • Microlayer does not fully evaporate due to limited thermal capacity of the heater. • Microlayer evaporation and rewetting are less important compared to thick heaters. • Most energy is removed by convective effects created by bubble growth and departure. • We re-derived heat flux partitioning model to quantify each heat transfer mechanism. In this work, we studied the wall heat flux partitioning during the pool boiling of water on thin metallic surfaces. We conducted boiling experiments on surfaces where we engineered nucleation sites by nanosecond-fiber-laser texturing. These nucleation sites form triangular lattice patterns with different pitches. We measured the time-dependent temperature and heat flux distributions on the boiling surface using an infrared camera. We developed post-processing algorithms to measure, based on these distributions, all the fundamental boiling parameters used in heat flux partitioning models (e.g., nucleation site density, bubble wait and growth time, and bubble footprint radius) and the actual partitioning of the heat flux, i.e., how much heat is transferred by evaporation of the microlayer, rewetting of the surface, and convective effects. This work reveals that the mechanisms of heat transfer on substrates of small thermal capacity are very different compared to substrates of large thermal capacity. With water, the bubble microlayer typically does not dry out and the surface temperature at rewetting is practically the same as the rewetting fluid temperature. These effects limit the efficiency of microlayer evaporation and rewetting heat transfer. Instead, convective effects generated by the bubble growth process remove most of the energy from the heated surface. This behavior is captured by a heat flux partitioning model that we re-derived from first principles to describe the heat transfer mechanisms on substrate of small thermal capacity.