Energy quality is defined as any alteration in the waveform of the voltage or current in the power supply. Any distortion in one of them will, in turn, generate a distortion in the other.
In the data center environment, this is often a topic addressed after various incidents or infrastructure failures have occurred. This has a more significant impact beyond equipment failures and the cost of repairs.
One of the most important impacts is on the business level, such as:
• Service interruptions.
• AIncreased energy costs due to losses.
• Loss of information and erroneous data in IT equipment.
• Deterioration of reliability.
• Fines for poor service.
This publication will review the types of power quality disturbances, their causes, consequences, and solutions for each one, concluding with the recommended approach to proactively address the issue, thus avoiding events that could lead to downtime or service problems.
A Bit of Theory
The relationship between current and voltage is explained by Ohm's law for AC, which in effective values is: V = I x Z.
As seen in the equation, impedances Z (loads of electrical circuits) will be an important part of the causes of distortions in the waveforms of both variables.
Current/Voltage Waveform Disturbances
Types of Disturbances
The main disturbances are mentioned below:
·
Overvoltages and undervoltages: According to duration. Sags
Sags
Surges
Transients
Spikes s
· Interrupciones.
· Ruido: Señales externas de alta frecuencia.
· Armónicos.
· Desbalance de fases.
· Variación de frecuencia.
Causes, Consequences, and Solutions s
Analyzing the most common disturbances, the following causes, consequences, and solutions can be found within the data center's infrastructure:
Harmonics
Sinusoidal voltages or currents whose frequencies are integer multiples of the fundamental frequency (50 Hz)
Consecuencias: Pérdidas por efecto Joule en conductores y transformadores.
Solution: Harmonic filters
Transients
Increase (greater than 110% of Vrms) in voltage for a period less than half a cycle.
Consequences: Damage to electronic equipment, data loss, and incorrect information.
Solution: Protections, grounding systems.
Undervoltages
Decrease (less than 90% of Vrms) in voltage for a period longer than half a cycle.
Consequences: Damage to electronic equipment, data loss, and incorrect information.
Solution: Undervoltage corrector.
Phase Imbalance
Causes: Phase load imbalance.
Consequences: Higher energy consumption due to losses and current through the neutral wire.
Solution: Proper phase load balancing.
Conclusion: How to Address the Issue?
To proactively address power quality issues in data centers and minimize their impact on business, the following steps are proposed:
· Infrastructure monitoring.
· Identify symptoms as described in the previous section based on monitoring.
· Determine the cause to identify the type of disturbance.
· Design the appropriate solution for each type of disturbance.
For this process to be truly efficient, it is essential that monitoring is done through a centralized platform where instant and historical values can be viewed, and alarm thresholds can be configured for values outside the standard operating range. To achieve this, it is necessary to incorporate sensing equipment such as amperometric clamps and power analyzers that can provide monitoring with a resolution of less than 1 second.
Finally, it is worth mentioning that safety is another important aspect when managing loads in a data center installation. Like the impact on infrastructure and business, addressing this issue is based on following this process and the recommendations mentioned.
