Participating companies | GELSENWASSER AG and DABBEL – Automation Intelligence GmbH |
Project location | Germany |
Industry | IT |
Technologies used | Utilisation of new control technology. Utilisation of AI for optimised HVAC systems control |
Energy efficiency | Reduction of energy consumption: 201,660 kWh/yr (30% HVAC systems) |
CO2 savings | 48 t CO2e/yr |
Economic efficiency | Monthly usage fee (approximately 30% of savings) |
Payback period | 0–3 years |
Funding programmes | No funding programmes |
GELSENWASSER AG is a large, municipal and supra-regional company that supplies energy and water and treats waste water. The goal of the submitted project is to save on energy costs and reduce CO2 emissions in its so-called ‘White House’, the headquarters of Gelsenwasser AG, which has a total space of 6,000 square metres. Prior to project implementation, the existing building management system controlled the heating, ventilation and air conditioning system (HVAC system) conventionally, based on reactive control regulations, and could not be adapted to meet the actual demand.
However, the AI software from DABBEL – Automation Intelligence GmbH can make this possible, which led to it being installed in the ‘White House’ in 2019. The AI system allows the HVAC system to respond dynamically to changes in outdoor and indoor conditions. The AI is driven by model predictive control (MPC). This enables the software to predict future conditions. The current situation as well as a wide range of variables such as insulation coefficients, solar radiation, humidity, CO content, room temperature, external temperature, number of people, and so forth are taken into account. Since variables are constantly changing, the software must constantly react to a new building environment. The installation was performed remotely without any additional hardware within just one week.
The software entered into an almost two-week-long learning phase in the ‘White House’, during which it adjusted to the building’s thermodynamics. The heating and cooling systems can now be controlled much more efficiently and operated at close to the optimum efficiency. Energy consumption was reduced by 30 per cent by using this intelligent, predictive control.
Transferability
The approach of using AI-supported HVAC systems control to increase energy efficiency is transferable to many industrial buildings. This approach can also be transferred to commercial buildings, such as administrative buildings, schools, clinics, hotels and so on. The prerequisite for this is that a building management system that uses BACnet, IP or Modbus TCP as the main communication protocol is in place in the respective building.
