In this article, we'll talk about how energy moves from a utility to a data center and the power distribution chain within the data center.
The best goal for a Class 4 data center would be to have two separate power lines or at least two main stations. This would provide two fully redundant power chains, including redundant access to the power grid. As this is very expensive and is usually not feasible. The most common option is to secure the power supply with diesel generators, a series of switches on both the power supply and the generator supply, to provide a level of redundancy for all equipment of the data center.
We start with the power supply, which comes from the power distribution and can be between 2 and 30 kilovolts. Since the power entering the building is usually 480 volts, there must be a transformer that can reduce the voltage to 480 volts. Power is then transferred from the transformer through one or more automatic transfer switches. The purpose of this switch is to be used when, due to power outages or blackouts, short term power fluctuations no longer provide enough power from the grid. When the power is off, the automatic transfer switch indicates that the standby generators that start the generators turn on to power data center equipment.
Many power outages can last less than 3 seconds, so the data center operator does not want generators to be turned on unnecessarily. It also takes a little time for generators to fully fire up and reach stable performance levels to bridge the gap between the end of the power supply and the power of the generators. We have a backup power source called uninterruptible power supply, called UPS. The reserve power supplied by the inverter is called transmit power and can be a battery or a flywheel system. Both systems store a certain amount of energy to save time between the power cut and the generator starting. Both types of systems have advantages and disadvantages. Once power is restored and stable. The automatic transfer switch returns to the utility grid and shuts down the fuse generators. The inverter then charges its energy storage capacities.
UPS power is typically connected to power distribution units, often referred to as PDUs. Note that these very large PDUs, and not the same as Rack PDUs, will be used for future reference. In these videos, I will be referring to these large PDUs as Floor PDUs to distinguish them from RAC PDUs.
Floor PDUs convert the output from 480 volts to 400 or 208 volts. Older data centers were built on 208 volts of electricity, while most new data centers use 400 volts of electricity. The reason for this development will be covered in another video later.
Ground PDUs are connected to multiple remote power panels called RPPs. These are larger versions of your home control panel.
If you provide three-phase power to the rack and connect a separate three-pin electrical wire, those three wires are on a line called a whip. These cables run under the raised floor, but at the far end of the whip data center is the power outlet.
Each rack should have two outlets from two different RPPs connected to two separate inverters as part of a redundant power chain.
Each rack's rack-mount PDU connects to outlets. Most servers now have two power supplies, and it is important to connect each power supply to a different rack PDU. The Rack PDU is then connected to these outlets on the racks.
That's all you need to know about the data center power supply chain. If you are looking for London Data Centre, click here.