5.3. Galactic Chemical Evolution of D
Numerical models are constructed to follow the evolution of the abundances of the elements in the ISM of our Galaxy.
The main parameters of the model include the yields of different stars, the distribution of stellar masses, the star formation rate, and the infall and outflow of gas. These parameters are adjusted to fit many different data. These Galactic chemical evolution models are especially useful to compare abundances at different epochs, for example, D/H today, in the ISM when the solar system formed, and primordially.
In an analysis of a variety of different models, Tosi et al.  concluded that the destruction of D in our Galaxy was at most a factor of a few, consistent with low but not high primordial D. They find that all models, which are consistent with all Galactic data, destroy D in the ISM today by less than a factor of three. Such chemical evolution will destroy an insignificant amount of D when metal abundances are as low as seen in the quasar absorbers.
Others have designed models which do destroy more D , , , , for example, by cycling most gas through low mass stars and removing the metals made by the accompanying high mass stars from the Galaxy. These models were designed to reduce high primordial D/H, expected from the low Yp values prevalent at that time, to the low ISM values. Tosi et al.  describe the generic difficulties with these models. To destroy 90% of the D, 90% of the gas must have been processed in and ejected from stars. These stars would then release more metals than are seen. If the gas is removed (e.g. expelled from the galaxy) to hide the metals, then the ratio of the mass in gas to that in remnants is would be lower than observed. Infall of primordial gas does not help, because this brings in excess D. These models also fail to deplete the D in quasar absorbers, because the stars which deplete the D, by ejecting gas without D, also eject carbon. The low abundance of carbon in the absorbers limits the destruction of D to < 1% .