Add error propagation to coupling

[awegsche]

Error propagation is still missing but needed for e.g. corrections (at least the existence of an ERR-column).
The corresponding CERN note needs to be revised.

First step is adding a column of zeroes so that development of correction scripts can continue.