Our current knowledge of production and destruction of light elements in astrophysical processes suggests that deuterium is produced during Big Bang nucleosynthesis and destroyed when cycled through stars. Primordial deuterium abundance can be determined by measuring the D/H ratio in a variety of astrophysical environments with different degrees of chemical evolution: the D/H ratio of unprocessed material directly gives the primordial value, while the ratio in processed material is expected to be lower and consistent with the predictions of galactic chemical evolution models. Here we focus our attention on deuterium abundance determinations of chemically processed material such as the interstellar gas in our Galaxy. Up to now, most of the determinations of deuterium abundance have been performed in the solar system or in local interstellar clouds. However, the overall accuracy of the measurements in local clouds is still insufficient to probe evolutionary trends. New D/H measurements in clouds at different locations in our Galaxy would be necessary to establish this issue, while interstellar measurements in nearby galaxies would give further constraints on the deuterium evolution in different galactic environments. With this goal in mind we have evaluated the capability of the Lyman channel of the SPECTRUM UV Rowland spectrography in determining deuterium column density in distant interstellar clouds. Three packages have been used to obtain realistic predicted spectra and to derive `observed' column densities: (1) the MIDAS package `CLOUD', to generate theoretical interstellar absorption profiles; (2) the `Synth' package developed in the IRAF environment by two of the authors to simulate spectroscopic observations of point sources obtainable with an astronomical spectrograph, (3) the FITLYMAN package inside the Lyman context of MIDAS to derive `observed' column densities from predicted spectra. The minimum exposure times, tmin, required to obtain a approximately 0.1 dex accuracy in the `observed' column densities, were derived by varying the input interstellar hydrogen column density. As a result, we show that the Lyman channel of the SPECTRUM UV Rowland spectrograph is up to the task of deriving accurate H and D column densities of low and medium column density interstellar clouds while it fails for N(HI) >= 1021 atoms cm-2.