Predictive models at the service of HVAC systems

The HVAC systems (Heating, Ventilation and Air Conditioning) include plants for the production of thermal energy, transportation of carrier fluids, as well as treatment and distribution terminals for air flows. A study conducted a few years ago by Frost & Sullivan entitled “Analysis of the Global HVAC Equipment Market” highlighted this market. A market that in 2014, was equal to 80.67 billion dollars, estimating growth by 2020 , which is expected to reach $ 108.93 billion. The research focused on unitary equipment, applied systems and refrigeration systems, heat pumps and non-ducted systems. Even then there was a strong demand for energy efficiency and therefore for low-consumption models. A question that would almost certainly be confirmed today, given that it is estimated that around 50% of the energy consumption in buildings is attributable to HVAC systems. Today there are predictive models at the service of HVAC systems.

The refrigeration units inside the HVAC systems

A calculation that, probably, should be revised upwards in industry and large-scale distribution. Just think, for example, of the impact that refrigeration units, used both for air conditioning and for the production of cooling fluid, have on the energy costs of many business sectors, starting with agri-food. In these types of systems, the main energy absorber is represented by the compressor, with a very wide range of thermal powers, from 0.1 kWf to some MWf. This type of plant is subject to frequent need for adjustment to vary the thermal load to be removed. They are determined by the external temperature or by the phases of the production cycle for industrial processes.

This means that, for a rational partialisation of the load, the habit of building water chiller systems with multiple compressors in parallel, which stop and start automatically in sequence as the cold demand changes, is invalidated. In addition, each compressor is equipped with a regulation system, with the possible availability of inverters, in order to modulate the performance with the same efficiency.

Because even the best technology is not enough

The manufacturers of refrigeration systems in particular, and of HVAC systems in general, are increasingly oriented towards finding solutions that can affect the costs of the energy carrier through improvements in machinery efficiency (and the use of inverters is one of these solutions) . However, even the most advanced technology cannot escape dispersion problems or failures that could arise after installation. Without forgetting that the company is not always in a position to replace obsolete systems with new machinery and must therefore try to optimize the existing one.

Either way, what you need is an accurate analysis of consumption data. This takes into account several factors including: vector, ignition sequence, external temperature, water inlet temperature, cooling energy etc. All these data, entered within predictive models, allow to identify measures to reduce consumption often at very low costs, helping to recover energy efficiency and to confer a competitive advantage that derives from the enhancement of strategic assets.

Greater efficiency and savings with predictive models

The predictive models for HVAC systems combine consumption data, energy drivers and efficiency objectives. That is, they allow:

  • improve energy efficiency up to 25% already at first use, and in any case according to the quantity and quality of the data available;
  • ensure the optimal functioning of the systems through the prior identification of deviations of energy behavior;
  • improve productivity by providing a diagnostic point of view and concrete support in solving anomalies.

In essence, predictive models are used to characterize the system’s operating modes in such a way as to predict its ideal performance in real time, highlighting potential waste, inefficiencies, anomalies and malfunctions. Through the use of the CuSum cards, they allow to read together the trend over time of the individual energy drivers and to check the variations of the slope in conjunction with the changes in the values of the energy drivers. The models do not identify inefficiencies due to the adoption of unsuitable or obsolete technologies. In the audit phase, they are decisive for assessing the profitability of any investments aimed at purchasing more efficient technologies.

If you want to learn more about predictive models and their advantages, contact us.

Scroll to Top