Flux ropes are twisted magnetic structures, which can be detected by in-situ measurements in the solar wind. However, different properties of detected flux ropes suggest different types of flux-rope population. As such, are there different populations of flux ropes? The answer is positive, and is the result of the analysis of four lists of flux ropes, including magnetic clouds (MCs), observed at 1 AU. The in-situ data for the four lists have been fitted with the same cylindrical force-free model, which provides an estimation of the local flux-rope parameters such as its radius and orientation. Since the flux-rope distributions have a large dynamic range, we go beyond a simple histogram analysis by developing a partition technique that uniformly distributes the statistical fluctuations over the radius range. By doing so, we find that small flux ropes with radius R < 0.1 AU have a steep power-law distribution in contrast to the larger flux ropes (identified as MCs), which have a Gaussian-like distribution. Next, from four CME catalogs, we estimate the expected flux-rope frequency per year at 1 AU. We find that the predicted numbers are similar to the frequencies of MCs observed in-situ. However, we also find that small flux ropes are at least ten times too abundant to correspond to CMEs, even to narrow ones. Investigating the different possible scenarios for the origin of those small flux ropes, we conclude that these twisted structures can be formed by blowout jets in the low corona or in coronal streamers.