Abstract
Model predictive control (MPC) approaches for HVAC systems in buildings have shown potential for significant reduction of energy use and for enabling flexible operation. However, a widespread adoption is long in coming, due to the high implementation costs and efforts, and data availability and quality. In this article, we demonstrate the implementation of a scalable, minimalistic and low-cost approach for replacing weather compensated control (WCC) with MPC in commercial buildings with legacy equipment. The concept utilizes the ventilation extract temperature as a proxy for indoor temperature to avoid excessive sensor installation and data treatment. The experiment demonstrates the applicability of the solution and indicates an energy cost saving of 33% during one month in early spring. While several points for improvement of the control algorithm are highlighted, the implementation shows robust behaviour even with up to eight hours gaps in the measurement updates.