3 MINS READ
To address the growing concern of climate change, several countries have pledged to limit global warming and become carbon neutral by 2050. Carbon neutrality can be achieved through carbon offsetting, where the emissions made in one sector are offset by reducing in another. Improving the energy efficiency of systems that consume electricity is one form of carbon offsetting.
As much as 50% of a building or brick-and-mortar facility’s energy may be consumed by heating, ventilation, and air-conditioning (HVAC) equipment. Rising electricity costs have led to the deployment of various energy conservation methods in buildings to optimize HVAC energy, involving both passive (building construction) and active HVAC operations methods.
Fast-paced urbanization is adding on to building loads—factors that cause stress, deformations, or accelerations on the structure, and the concurrent need to ensure the biosafety of users, have given rise to both challenges and opportunities for facility operators.
A mix of control techniques can make next-generation buildings biosafe.
A building type, whether residential, commercial, or industrial, that determines who the user is and equipment schedules along with external environmental conditions are the key drivers for HVAC performance optimization. The end goal is to provide the building occupant a holistic convenience in terms of thermal, visual, acoustic, and biosafety.
To improve operational energy efficiency, building operators employ various energy conservation measures (ECMs) using control techniques of varying complexities.
Rule-based methods are by far the most common and simplest to implement. These involve exploiting durations of lean occupancy in buildings using higher set-back temperatures and adjusting ventilation, and scheduling equipment based on time-of-use electricity tariffs. However, these methods are mostly reactive, hence not the most viable or optimal.
Model-based predictive control (MPC) may offer a more optimal solution if there is an accurate thermal model of the workspace and performance models of the associated equipment. The downside of MPC is that model development can be cumbersome, particularly in case of large buildings and due to the lack of controller logic, which is usually proprietary.
The increasing popularity of IoT and AI-based technologies are helping reimagine legacy infrastructure. Deep reinforcement learning (DRL) is a class of machine learning under AI that is model-free and interacts directly with the environment. It is known to deliver near-optimal solutions. Classical DRL, however, holds the disadvantage as it calls for extensive online training and scalability. Recent advancements in transfer-learning and meta-learning techniques are helping resolve some of these issues.
Given that every control technique comes with its own challenges, a combination of these could count in the effort toward making next-generation buildings biosafe and carbon neutral.
Newer ways of work and social interaction have contributed in many ways, to how a facility consumes power. Building operators are aware of the need to transform traditional HVAC systems to fit these new requirements, ensure carbon neutrality, and maintain biosafety. All three are urgent and important sustainability goals.