Miniaturization of semiconductor devices is the driving force in the field of microelectronics. In the pursuit of this goal, various undesired and degrading effects are encountered that make a prediction of device behavior less reliable. One of the most important reliability problems is negative bias temperature instability (NBTI). This phenomenon manifests itself at high temperatures combined with negative voltages applied to the gate. During the last couple of years, various explanations of this phenomenon, based on different and contradictory atomic assumptions, have been put forward. It is well established that the degrading effect is caused by separating the hydrogen atom from the dangling bond. But it is still under debate whether the electrical field or the temperature initiate this reaction. In addition, the charge state of the hydrogen originating from the dangling bond has to be clarified. But the question arises, what happens with the hydrogen? First, two hydrogen atoms can bind together forming a molecule. Second, it may undergo reactions with defects located in the oxide. These defects occur mainly as oxygen vacancies but in various atomic formations as well as in different charge states. Last, it must be taken into account that the hydrogen in the oxide can change its charge state itself. Considering all the combinations of the different charge states and the different forms of the participants in these reactions results in a comprehensive model with a wide range of reactions. The necessary data, including reaction barriers and reaction possibilities, are not available through experiments alone. Therefore support of theoretical simulations that help, such as including only the relevant processes in the oxide, determining reaction probabilities, and ruling out impossible reactions, will be of considerable importance. This can be done within the framework of ab-initio simulations and will shed light on this matter.