I recently did an undergraduate thesis which included some projections of what would happen were we to fill the gap in oil production with synfuels from coal, analysed in a similar way to the spreadsheet on this page:
EVWorld Blogs
but I took into account many other factors, such as:
the projected demand for oil,
the projected proportion of electricity generation from coal (currently 40%),
projected demand for electricity,
projected energy required for transport,
projected proportion of transport directly fuelled from oil,
projected proportion of electrically powered transport (more efficient engines),
average transmission losses in power networks,
the effficiency of coal fired power stations (may rise with new technologies)
estimated total recoverable coal remaining.
The main one being an assumption that a certain level of electrically powered transprot will replace current technologies. I used a polynomial fitted to a Hubbert curve derived from production figures in the BP statistical review with a peak around 2005-2006, as my production figures. I didn't have the data to make a more sophisticated prediction.
The upshot of this is coal is exausted by 2110 in my optomistic scenario, and this also assumes no peak coal, i.e. production increases until the moment we run out. My pessimistic scenario has us run out by 2047