Every year, the number of heat pumps installed in buildings increases, with most of them used for space heating and cooling. The use of heat pumps for domestic hot water (DHW) production is smaller in comparison but is increasing. Advanced control strategies enable effective management of energy flexibility in heat pumps, optimizing the use of variable electricity prices and enhancing PV self-consumption. This paper documents the implementation of a model predictive controller (MPC) in a centralized DHW production system with heat pumps. The use of centralized DHW system is a growing trend in Spain, replacing individual systems. The MPC requires real-time measurements of the system and data from electricity prices, external weather forecasts, and self-developed DHW, electricity consumption, and PV production forecasts. With this data, the MPC calculates the most cost-effective or emission-less operational path. The implementation of the MPC during one week led to a 6% reduction in thermal energy production and a 3.8% decrease in electricity consumption. However, it did not yield significant improvements in the COP or operating costs. Additionally, varying external conditions hinder a direct comparison.