In this study we apply fracture characterization techniques, common to many other reservoir rocks, to coals. Additionally, the data gathered from this technique is used to determine fracture permeability and porosity based on a cubic relationship between permeability and porosity, as a function of aperture and spacing. However, these data are in an unconfined stress state. To relate this information to the subsurface, which is in a confined state, the results are then compared and calibrated to permeability data determined from laboratory triaxially confined samples, using the same sample material. A 1D scanline technique is used to gather data on cleat aperture, spacing, height, and frequency. Sixteen different coals from several European basins were examined, creating a large dataset (>8000 individual data measurements). This data was gathered from blocks of coals and from polished sections. The polished sections were digitally scanned under the microscope to create an image of the whole section. As a result, we examined a range of scales from micrometers to centimeters. Maximum, minimum and median values are determined for each attribute. To clearly define the limits of resolution of the data, a practical pore resolution is applied to a power law function of the data. Cleat attributes were further examined to determine a relationship with cleat type, e.g., face or butt, and lithotype. Power law distributions of butt cleat attributes are similar to the distributions found in lithotype. It was concluded that butt cleats are more strongly related to lithotype, and thus bed thickness and mechanical strength of the bed, than face cleats. Cleat aperture and spacing depend on cleat type. A face to butt cleat ratio of 5:1 given an aperture of about 12 lam was found. This is similar to a published anisotropy in permeability, where face cleats are five times more permeable than butt cleats. Cleat porosity from aperture and spacing is several times greater than cleat porosity calibrated to laboratory permeability (under stress). The calibrated porosity ranged from 0.01% to 1.8% and agrees with published values of in-situ cleat porosity. Thus aperture width due to an increase in stress is predicted to decrease four times. This data can be used to calibrate reservoir simulators for CBM production. (C) 2015 Published by Elsevier B.V.