A two stage procedure for control structure analysis and design is proposed and investigated on a heatpump control structure synthesis problem. In the first stage qualitative process knowledge is used to suggest possible control structures, ie. measurements, actuators and feasible loop pairings for multiloop SISO (scalar) control. In the second stage frequency dependent SVD (Singular Value Decomposition) is employed on a quantitative model to indicate, whether a multivariable control strategy is called for. if scalar control is inferred to be adequate, the RGA (Relative Gain Array) method is used to synthesize loop pairings for multiloop SISO control. The proposed procedure is demonstrated using a heatpump control structure analysis and design problem. The heatpump is used on an industrial scale energy integrated distillation column. Based on qualitative knowledge the process interactions are illustrated using a DYnamic QUantity Interaction Diagram (DYQUID). From qualitative considerations a multiloop SISO control strategy. which concurs with intuitive engineering knowledge, is suggested. A linear state space model obtained from a non-linear simulation model is used to quantitatively analyse the control structuring problem. SVD analysis shows, that a multiloop control strategy is justified, and RGA analysis suggests the same loop structure as inferred qualitatively. Finally the overall control structure for the energy integrated distillation column is discussed in relation to experimental results.