Abstract. This work is concerned with conditional averaging methods which can be used for modeling of transport in porous media with volume reactions in the fluid phase and surface reactions at the fluid/solid interface. The model under consideration takes into account convection, diffusion within the pores and on larger scales, and homogeneous and heterogeneous reactions. Near the interface with fractal properties, the fluid flow is slow, and diffusion, as a transport mechanism, dominates over convection. Following the conditional moment closure paradigm, we employ a diffusion tracer as a reference scalar field that makes the conditional averaging sensitive to the proximity of a point to the interface. The resulting conditionally averaged reactive transport equations are governed by the probability density function (PDF) of the diffusion tracer, and this makes the study of its behavior an important problem. We consider a hitting time stochastic interpretation of the diffusion tracer, establish integral equations relating it to a subsidiary distance tracer, and obtain distance-diffusion inequalities. Assuming that the fluid/solid interface and pores themselves possess fractal properties which are quantified, in particular, by a variant of the Minkowski-Bouligand fractal dimension, we investigate the interplay between the interface and network scenarios of fractality in the scaling laws of the diffusion tracer PDF. We also discuss and employ several hypotheses, including a lognormal cascade hypothesis on the behavior of the diffusion tracer at different length scales.