Energy Management Knowledge - Overview
Basics of Energy Management from a technical perspective
In reality, a system (such as a production facility, data center, or logistics center) is always a combination of different subsystems. Due to their interaction, it is difficult to determine the optimal combination settings of the subsystems. Each subsystem has an optimal working point for itself. In addition, there are many constraints to consider. For example, a compressor may only work with a minimum speed, otherwise the oil pressure is insufficient. It is an exiting task to optimize the different subsystems in an orchestrated way. So the total consumption is minimized, whilst the resilience might even improve, as resources are used more meaningful.
From our experience - in the optimization of data center facilities - we summarize a few examples here. They should be seen as suggestions. We are happy to support you in the planning and implementation of those projects.
Real saving rather than shifting
Don't squeeze the ballon!
Our goal is saving instead of redistributing. For example, we do not want to save gas and spend a lot more on electricity.
Or save on cooling, but use more in ventilation. The decisive factor is the acquisition and measurement of the essential consumers as a basis for effective cost reduction.
Often, the exact relationships between the parts of the plant are unknown. The overall complexity makes a simulation difficult. Therefore, it makes sense to determine the cause-effect relationship via suitable measuring methods and analytical instruments. Changes must be sufficiently documented. This allows to return to a baseline at any time.
An isolated optimization of subsystems can lead to rising total energy costs. The holistic view and measurement ensures that these negative effects are avoided.
An example: saving energy by raising the room air temperature in the data center
If the room air temperature in a data center is increased, the temperature of the cooling water can also be increased.
This reduces the operating hours of the refrigeration system and thus leads to significant energy savings.
However, modern room coolers, computers and routers have regulated fans for ventilation.
As the temperature rises, so the fans will absorb much more energy than was previously the case.
As the energy demand increases with the fan speed in the third power, their consumption increases significantly.
In addition, it can increasingly come to failures. And the lifetime of batteries may be significantly reduced.
It is therefore crucial to monitor the total energy consumption and include further quality characteristics.
Since the cooling energy mainly depends, on the outside temperature, adjustments should be made to draw the right conclusions.
Ideally, a measurement takes place at all levels. This makes it possible to clearly identify cause / effect relationships and develop an effective long-term, strategy.
fast streaming fluids (e.g. air) create under pressure at the surface. At a wing this causes the object to rise.
Prevention of air mixing through proper ventilation
In many data centers, cold air is blown into the cabinets to be cooled via a raised floor. If these cabinets are located very close to the cooling unit, it is likely that the air is not flowing into the cabinet. Instead it will suck warm air into the raised floor. This is a negative example of the venturi effect.
It is therefore recommended, to do air flow measurements in addition to temperature measurements. This can ensure that the cabinets are sufficiently supplied with cooling air. In the simplest case, a sheet of paper can indicate flow issues. To capture changes more accurately, an anemometer (airflow meter) is recommended.
To avoid the negative venturi effect, the dynamic pressure must be converted to static pressure. You can do this either:
- best would be to reduce the fan speed of the cooling units as this reduces energy consumption
- place the cabinets at least 1.8m away from the cooling units, or
- a guiding plate can be inserted into the stream.
Basically, the pressure drop should be low.
- it reduces the flow resistance and hence the required fan power.
- it reduces the Venturi effect and thus ensures direct action of the cooling medium
- it increase the volume of cooling medium and. This increases the efficiency of cooling.
example of the negative Venturi-Effect in improper data center design
reducing pumping power = increasing thermal transfer rate
less is more
less pumping capacity leads to lower delivery volumes and thus to increased efficiency of the heat transfer
Double energy savings through optimum delivery
In many cases, the more the more applies. However, this is not necessarily the case with heat exchangers. It can easily be seen that if a cooling medium has more time (δ t) for the heat transfer, it becomes more efficient. As a result, more heat is transferred and the temperature difference (δ T) between inflow and return increases.
Two effects reduce energy consumption:
- reducing the pumping power saves electrical energy for the pump drive.
- the higher return temperatures allow longer use of the "free air cooling" and thus reduce the operating hours of the more expensive "mechanical cooling." (See also: Extended use of free cooling by raising operating temperatures )
Of course, this procedure can only be used until at least flow rates are undercut or the temperatures in the rooms do not rise too far. So there must be suitable measuring and monitoring equipment available to control the system at any time.
Reference for Temperature and Humidity
The psychrometric diagram shows the work areas for temperature and humidity in computer rooms. By exploiting these areas, energy consumption can be massively reduced.
Data centers traditionally spend the most energy on IT and cooling. To avoid the risk of overheating the equipment, the setpoints for temperature and humidity are set very low. The tolerances are often too narrow.
At temperatures, this leads to unnecessarily high cooling use with corresponding costs.
Narrow air humidity tolerances provide high energy requirements when drying and during humidification.
Today's equipment can safely operate at temperatures above 40°ree;C. The humidity can vary between 10% and 70%.
The figure shows the permitted value ranges as recommended by the ASHREA and the ETSI.
It's time to check:
- which temperatures prevail?
- which equipment requires lower temperatures? and
- how is the temperature distribution in the room?
Then - ideally in several smaller steps - the temperature can be raised.
State-of-the-art data centers work with inlet temperatures above 27°ree;C. As a result, ambient air can be used for cooling most of the year. Expensive compression-based mechanical cooling systems are now a thing of the past.
extended use by raising temperatures
Increasing the cooling temperatures reduces energy costs for cooling.
Mechanical cooling can be largely replaced by cheaper free air cooling.
Cooling data centers often uses mechanical cooling . As with refrigerators, the refrigerant is compressed by a compressor and expanded through a throttle valve. The cooling water is cooled down to the desired temperature.
The energy input here is immense. It depends on the temperature difference between the input and output temperature. Pre-cooling via the outside air can therefore significantly reduce costs.
In the best case, the temperature of the environment is far below that of the cooling water. Then the mechanical cooling can be completely avoided. This is called Free Cooling.
If the ambient temperature is only a few degrees lower than that of the cooling water supply, the first step is to use free cooling. In the second step, the mechanical cooling has to bridge the temperature difference. This is then the Assisted Free Cooling .
As shown in the picture on the left, the free cooling percentage can be increased when the cooling water temperatures are raised.
Avoiding Humidification and Drying
effect of warm and wet air
The humidification and dehumidification of data center rooms can generally be completely eliminated in Europe. The reason is in the average humidity values. They are between about 40% and 100%.
Since the outside temperature is usually lower than the room, the humidity values automatically decrease due to the heating.
Added to this, there is condensation taking place in the cooling units.
Critical are only warm days with heavy rainfall. In these phases, the air exchange rate should be reduced anyway, in order not to have to cool warm fresh air. The further dehumidification is the provided by the cooling units.
The higher the humidity of the fresh air entering the cooling unit, the more moisture condenses during the cooling. So moisture is continuously removed from the room. The inflowing fresh air compensates for this loss of water. A psychrometric chart can be used to calculate these operations.