The strong international focus on reduction of CO2 emissions to the atmosphere involves many situations in which hydrate may form from water and CO2. Transport of CO2 in pipelines typically involves high pressures and low temperatures favourable for hydrate formation. Aquifer storage of CO2 also frequently involves regions of the storage sediments which are inside CO2 hydrate formation. CO2 hydrate can even be an active phase in CO2 separation technology. Hydrate formation, like any other phase transition, has two physically well-defined stages. The initial nucleation is an unstable phase in which the thermodynamic benefit of the phase transition competes with the thermodynamic penalty of pushing away the existing phases. The latter term depends on size as well as shape of the growing hydrate nuclei. The transition from nucleation over to stable growth is therefore generally dependent on the geometrical dimension of the crystal, which can range from one for a spherical core and upwards to complex crystal morphologies. A third stage which is not uniquely defined physically is the so-called induction time. This is the stage at which the hydrate growth is massive and visible by various detection methods like pressure change, laser detection of crystals or other imaging techniques. Frequently, the induction time is misinterpreted as nucleation time. Nucleation is a nano-scale process, but several situations of hydrate formation leads to changes in access to hydrate building blocks. Formation of a hydrate film on the interface between water and a hydrate former phase will substantially delay mass transport for further growth. The slow, mass transport limited, growth towards detectable hydrate is therefore frequently misinterpreted as absence of hydrate. Yet another set of misunderstandings arise from the fact that hydrates in industry and nature can never approach equilibrium. Some of these are well known for simple systems, but there are also some relevant phases that are rarely accounted for, for a deeper understanding of the hydrate. Finally, there are frequently misconceptions that only one type of hydrate is formed always. In this work we utilize classical thermodynamics with residual thermodynamics description for all phases, including hydrate, to analyse various routes to hydrate formation between CO2 and water. We utilize classical nucleation theory for simplified hydrate geometries in order to illustrate the range of likely hydrate nucleation times as well as ranges of different hydrate that can form. Hydrate forming homogeneously from dissolved CO2 in water can form a range of different hydrates since the composition will change depending on the concentration of CO2 in water. Hydrate nucleation times are very fast and in the nanoseconds range. Times for onset of massive growth in systems without stirring or other hydrodynamic forces that can break hydrate films can be very long due to slow transport through initial hydrate films.