Predictive control of thermal equipment

Predictive control of thermal equipment, a key factor in reducing energy costs.


Used to regulate thermal equipment in industrial, commercial and service buildings, dynamic control is proving more effective than traditional control systems in reducing building energy costs.

This technology monitors the equipment’s behaviour using IoT sensors. It anticipates the weather and site activity to regulate and deliver the best building performance coefficient.


1. Impact of lack of dynamic control in an office tower

2. And in a cold store?

3. Reaction vs. Anticipation

4. Opting for dynamic control


1. Impact of lack of dynamic control in an office tower

Let’s take a first example in an office tower in Southeast Asia, where it’s often hot.

The air conditioning system is not dynamically controlled. This means that the building reacts to the elements.

The office rooms feel cold immediately when it rains. This is because the outside temperature drops a few degrees and the sun disappears behind the clouds, no longer heating the room through the windows. Of course, the fresh air supply is also colder, exacerbating the temperature drop in the building.

The climate control system will detect the temperature change and react accordingly.

However, this change does not happen instantly: several tens of minutes – usually 15 to 30 – are required. Prior to adaptation, the system would have continued to produce cold, reducing occupant comfort and increasing energy consumption.

This is a typical example of the kind of situation that our teams encounter in Asia.

The flaw in this process is that it relies on a “reactive” system, as opposed to a virtuous system that constantly anticipates weather changes.


2. And in a cold store?

In Europe, weather variations are observed between day and night and on a seasonal scale. Examples of the importance of dynamic control are different from Asia, but no less numerous.

Here, let’s look at the case of defrosting evaporators in storage. Classically, it is based on a schedule defined and parameterised in the BMS. For example: defrosting starts at 4 o’clock, then at 8 o’clock, and so on throughout the day.

However, more advanced control systems are available. For example, regularly opening a cell door increases the humidity level around the machines near the door, requiring more frequent defrosting. However, it may be optimal to limit defrosting at night when the door is not in use.

Nothing is done to adapt to the load level of the warehouse. However, a cell used at 30% of its capacity requires less cooling than a cell used at 100% capacity. Evaporator use is reduced ; theory suggests reducing defrost actions.

The same applies to the meteorological impact (temperature, humidity, wind, seasons, etc.). It is ignored by the systems because it is difficult to predict, but it is nevertheless the key to anticipating the need for evaporator defrosting.

In addition, defrosting, according to the technique, results in a variable energy consumption. While defrosting with ambient air is sufficient for products stored in a positive cold environment, this is not the case for products stored in a negative environment. There are therefore other forms of defrosting: hot water, gas, electric, etc.

Although necessary, these methods have a double disadvantage: energy intensive, they heat up the room, which has to be cooled afterwards.

So you better defrost when it’s really necessary!

Finally, defrost duration is often fixed. But isn’t it better to adapt to what’s really needed? Why should we systematically plan for 30 minutes, when sometimes 28 or 25 minutes would be enough?


3. Reaction vs. Anticipation

The interest in acting in advance is therefore both financial and to meet operational and comfort needs.

Instead of applying linear setpoints, it is important to understand the future thermal requirements and to adapt the heating and cooling systems accordingly to the weather variables and the building activity (e.g. at night, on public holidays, on Sundays…).

  • The difference between “reacting” and “anticipating” in terms of CO2 impact, consumption and cost is significant.

If the environment is dynamic, so must regulation.


The means to be used

Returning to our examples.

In our office tower in Asia, it is important to adjust the temperature settings at the right time to take full advantage of the natural cold and avoid discomfort in the building.

To do this, BeeBryte’s predictive optimisation system incorporates future weather data into calculating what the building needs to heat. By considering the thermal inertia of the building, we adjust the temperature setting when needed.

Comfort and reduced consumption: Dynamic control kills two birds with one stone.

If this first example illustrates the importance of considering the future weather, it is equally important to consider the level of activity: a building where half of the occupants telecommute will have a lower thermal demand than one that is fully occupied.

In warehouse optimisation, our solution monitors evaporators to understand their tendency to freeze depending on the environment. This allows us to detect the need and initiate defrost at the right time.

It is also adjusted in its duration. If 20 minutes is enough to defrost, let’s defrost for 20 minutes, no more!

This dynamic aspect allows both :

  • To avoid heating up a cell
  • To increase the availability of evaporators for cooling
  • To consume intelligently and to generate energy savings


If you would like to know more, please contact us.

"*" indicates required fields

This field is for validation purposes and should be left unchanged.


Share on :