Predictive Control of a Building Heating System

In the European Union (EU), it is estimated that the building sector consumes 40% of the total energy production. The EU has set the goal to reduce energy consumption by 20% by 2020. This ambitious goal requires to find ways to reduce energy consumption in buildings. When used in heating and cooling systems, an advanced process control methodology - model predictive control (MPC), can be beneficial compared to current control strategies. This control methodology allows to build a multivariate constraint model: key constraints such as the thermal capacity of the building and energy prices can be included while the predictive control uses weather forecasts to optimize resources and prepare for changes in outdoor temperature. Modeled in real time, MPC has reduced both energy consumption and costs.This paper provides an algorithm for such a model predictive control of multi-family buildings (MFB) based on a case study in Latvia. The building uses two heat pumps and electrical heaters as heat sources for space heating and two thermal accumulation tanks for balancing heat generation and the building's heat demand. Input values are weather forecasts and electricity tariff. MPC is used to improve the operation schedule of heat sources and to achieve financial savings.