Article from REHVA Journal 6, 2021 – Jan Kerdèl, senior consultant, Kerdèl Business Development; Pieter Pauwels, dr. Ir.-arch, TU Eindhoven
The installation sector is digitalising! As in the rest of the building industry, systems digitization and innovation have an increasing impact on the HVAC and installation sector. Far-reaching evolutions towards ‘digital twins’ and ‘smart buildings’, including predictive maintenance strategies and centrally linked systems, mean that the installation sector faces the next wave of digitization, or rather an AI (r)evolution.
Partly for this reason, ‘digitization’ is an important topic at the upcoming REHVA World Congress CLIMA 2022, the leading international scientific congress in the field of heating, ventilation and air conditioning (HVAC). The congress is organised every three years by one of the member associations of REHVA. The upcoming edition will be organised by TVVL in the Netherlands, in cooperation with Delft University of Technology and Eindhoven University of Technology. In addition to Digitization, CLIMA 2022 also covers the topics of Energy, Circularity, Health & Comfort, and Education.
How will current demands and influences evolve until 2030? What insights will we gain? How do the topics relate to each other? Will they reinforce each other or slow each other down? And how does the impact of climate systems relate to the building process and other influencing factors? What kind of teaching and learning is needed? With these questions in mind, CLIMA 2022 will have its EYE ON 2030.
Digitization is, therefore, an important topic at CLIMA 2022. However, typically digitization is not a stand-alone topic: it is often part of larger issues. No doubt other conference topics will use digitization to carry out complex research or solve certain issues. Digitization also plays an undeniable role in the design and operation of installations. And here, too, it typically serves specific purposes (e.g. system optimisation). In the following we provide a brief overview of the most important developments of this form of digitization for the HVAC sector.
Digital solutions that stimulate the energy transition in the built environment are a crucial topic. In many companies and organisations, solutions in the field of (predictive) digital twinning, data-driven smart buildings, data management, and continuous commissioning are high on the agenda.
Nowadays, digital solutions must be able to handle a wide range of HVAC systems and be self-learning in detecting trends and process deviations. While current systems often concentrate on monitoring, it is becoming clear that the future lies in predictive planning of interventions based on recommendations from an Al system.
It is expected that in an actual, physical building (physical twin in Fig. 1), several sensors and actuators will actively monitor data. In this environment, monitoring and surveillance are therefore carried out in a data-driven manner. On the other hand, the digital side means a virtual model (digital twin) that is usually strongly model-based. An information model and/or prediction model (e.g. neural network) is digitally available, including a number of simulated scenarios. These are compared to the input from the data-driven physical twin. Such comparisons make it possible to work on prediction, fault detection and self-learning (Fig. 1). In this way, model-based and data-driven approaches are combined, making for a powerful potential resource and important Al research topic, that can also be employed for HVAC systems.
The above-mentioned ambitions put the evolution towards dynamic HVAC systems under digitization pressure. Stand-alone or integrated solutions are possible, and a system and environment as shown in Fig. 1 is certainly achievable and already a reality in several places. However, there are some important preconditions or even obstacles for placing such systems on the market in a scalable way (performance, cost, speed):
- System architectures must be able to cope with large-scale implementation with various hardware (wired and wireless solutions, loT, cloud solutions, blockchain technologies), and at the same time, they must be flexible to continuously accommodate change (additional sensors, new users, change of provider, etc.).
- In addition to these large-scale and flexible infrastructures, the world of smart buildings also requires monitoring strategies that bridge the gap between building automation and control systems (BACS) and building information modelling (BIM) tools.
- Also, the recent COVID pandemic has led to research into digital-oriented design, monitoring and control of ventilation systems related to general comfort and health. This includes machine learning (ML) algorithms for fault detection and diagnosis, pattern recognition and anomaly detection: training a model based on sensor data; model-based prediction; intervention in the building system.
With the above evolutions in the HVAC field, the lifecycle costs of a building are expected to be easier to control, the comfort of the occupant or user will improve, and the system will be easier to monitor and maintain.
Building management systems for energy performance innovations are a prerequisite for better performing buildings. These are buildings that adapt to the changing climate, provide optimum comfort in an intelligent manner and, ideally, also produce energy (net positive buildings). The innovations are primarily expected from the building management system (BMS), which maintains a central reference point for a specific building (model-based and data-driven – Fig. 1). Based on this, a building management system can actively carry out interventions in the building.
Positive energy should also become possible in the field of energy management for buildings. Solutions in this direction are expected in the areas of (predictive) digital twinning and data-driven smart buildings, in which building performance is monitored and displayed in real time through dashboards that relate building data and measured values (time series). Recent developments link this to information models and metadata standards for data management such as the Industry Foundation Classes (IFC), Linked Building Data (LBD), Brick, and Haystack.
The latest developments are:
- Energy transition measures for existing buildings
- Net positive building developments
- Building performance monitoring with digital twins
- Data-driven smart buildings: monitoring based on time series data
- Linked building data for digital twinning (LBD, Brick, Haystack, etc.)
Design for automation: from SIM models to BACS
Typically, there is a Building Information Model or BIM model available for modern buildings. Such a semantic 3D model allows architects, engineers and building professionals to plan, design, build and manage a building better and more efficiently. However, the BIM model is usually limited to the design and construction phase. This means that valuable information in the operational phase is either collected by another dedicated management system or (not infrequently) lost.
Research and development are trying to close the gap that still exists between BIM and Building Automation and Control Systems (BACS). The information available in the BIM design model could form an excellent basis for the start-up and design of the control technology in a BACS. Since the orientation of the building, the use of space, the intended use and the desired comfort classes are known, SIM routines could be developed to suggest BACS solutions and control strategies. Monitoring strategies should follow to maintain quality and control costs throughout the lifecycle. Hereby, information from SIM models can become a basic reference for cost control over the lifecycle of a building.
The latest developments:
- BIM for indoor climate control design
- Building automation design from BIM environments (BIM & BACS)
- Automation of maintenance and monitoring: self-sufficient buildings
- From data to decision-making: standards and best practices
- Facility management design
- Digitization of design and engineering of HVAC installations
HVAC control and health monitoring
Particularly after the outbreak of the Covid pandemic, much research has focused on rethinking ventilation strategies and their design, monitoring and control. This leads to an intensive research project on the role of ventilation in comfort and health monitoring in a building. Within Clima2022, several contributions are expected around the digitization-focused research on the design, monitoring and control of ventilation in a building in case of a pandemic.
Furthermore, the use of Al techniques for this purpose is encouraged, in particular the use of Al algorithms for fault detection and diagnosis, but also pattern recognition and anomaly detection in building use and HVAC systems. It is important to use these digital technologies for critical control and risk mitigation strategies, rather than signaling every minimal error or deviation within tolerances.
Rather than displaying all available information, the key question is how these techniques can be used to proactively predict where and how systems will fail and how this risk can be mitigated.
The latest developments are:
- Ventilation strategies in pandemics (design, monitoring, control)
- Health and comfort monitoring
- Pattern recognition and anomaly detection in building use
- Algorithms for fault detection and diagnosis
- AI for critical control and risk mitigation strategies (proactive vs. reactive)
Integration in existing buildings: Upgrade of buildings
Buildings and the energy infrastructure are undergoing a transition to a carbon-free society by the year 2050 (Paris Agreement 2015). The approach differs from country to country: in countries with a temperate climate, the building envelope is often of insufficient quality and requires renovation or upgrading. In addition, low temperature (LT) heating and cooling solutions are often implemented, even though they are more sensitive to failures. In order to deploy these systems appropriately and make renovation possible, large-scale monitoring at an affordable cost must be developed to continuously monitor commercial buildings.
Research on this topic typically focuses on the process of building improvement and poses new challenges for building management; this topic is thus also expected to be discussed at Clima 2022. Solutions must be able to handle a wide range of HVAC systems and be self-learning in detecting trends and process deviations. In renovating and implementing systems, system architectures must include large-scale implementation. Both wired and wireless solutions are possible, and in the case of implementing loT in a renovated building, a professional approach to loT (lIoT as a quality standard) is recommended (lowest risk level).
The latest developments are:
- loT and industrial loT
- Cloud solutions for building management (e.g. Kubernetes, MS Azure)
- Security, control and authorisation in cloud-based BMS systems
- Edge computing: on-device computing for building automation
- LoRa and LoRaWAN: training connected and loT devices
- Local area networks, WiFi networks and 5G networks in embedded infrastructure