The problem of designing the estimation structure to perform a specific concentration estimation task in a binary distillation column with two temperature measurements is addressed within a geometric estimation (GE) framework. The structure design involves the choices of: (i) the measurement locations, (i) the complete or truncated estimation model, and (iii) the innovated-noninnovated state partition of the model. First, a structural analysis is performed on the basis of detectability measures in the light of the column characteristics, yielding candidate structures for a specific estimation task. Then, the behavior of the structures is assessed with estimator functioning and dimension considerations, yielding conclusive structural results. The proposed methodology is applied to a representative case example with experimental data.