In this paper, we examine the pore space geometry and topology of a North Sea sandstone reservoir rock based on multi-scale scanning electron microscopy. The reservoir was subjected to extensive diagenesis which has resulted in a complex pore space with a wide range in pore sizes. We quantify the pore size and pore coordination number distributions, and we find that the mean and standard deviation of the coordination number are power law functions of pore radius, where the scaling exponent varies from 0.3 to 0.5. We present a 2D stochastic algorithm to generate a pore network based on statistical information. The algorithm incorporates the concept of a weighted planar stochastic lattice which is a construction that naturally leads to scale-free character with power law behaviour. We validate the algorithm against SEM imaging by showing that it can reproduce the observed clustering and a realistic spatial distribution of pore space elements. We also try to explore the relationship between fluid flow properties of reservoir rock and 2D pore image features. Keywords Porous media • Pore network • SEM imaging • Multi-scale 856 G. Scott et al. r i Radius of element i r min Minimum element radius r max Maximum element radius r 0 Pore size distribution parameter, near the minimum radius r 1 Pore size distribution parameter, near the maximum radius r x Characteristic radius of largest pores which control permeability Z Coordination number Z i Coordination number of element i Z mean Mean coordination number Z stdev Standard deviation of the coordination number α Scaling index in the pore size probability distribution β Scaling index in the relationship between Z mean (r ) and r φ Porosity χ Connectivity function