Data center isolation transformers and normal transformers have the same function to isolate data centers but different ways to achieve them. A data center isolation transformer is equipment that can protect the main power supply from the side of faults in the secondary network to ensure the data center’s regular operation.
A normal transformer can isolate the power supply by itself. Therefore, data center isolation transformers are usually used together with normal transformers in one, two, or three-phase transformers.
Data Center Isolation Transformers
A normal transformer transfers electrical power from one circuit to another circuit with the same voltage but a different current. However, a data center isolation transformer transfers electrical energy from one isolated circuit to another isolated circuit with the same voltage but a different current. The difference between the two is that an isolated circuit has no other circuits connected and thus would not exchange power with other circuits.
To avoid potential damage to electronic components, each piece of IT equipment must be connected to an individual transformer that steps down the voltage while isolating it from problems in the grid. There may be one transformer per rack or cabinet or groupings of racks. The amount depends on how much power is being used by the IT equipment and how much excess capacity there is in the transformer.
This arrangement can be complicated because of a wide range of factors that affect utility companies’ abilities to provide reliable power sources, including:
- The availability of sufficient transmission lines and substations.
- The capacity of power plants.
- The ability to implement new technology quickly enough for it to be helpful.
- The health and efficiency of transmission lines and substations.
When there’s an issue with any of these factors, it can significantly affect the reliability of the electrical service provided by utilities.
Data Center Isolation Transformer Applications
The primary function of a data center isolation transformer is to provide electrical power safely by isolating circuits. It can prevent short circuits and overloading, which can help maintain the structural integrity and safety of the system. Datacenter isolation transformers are required at each point where distribution lines deliver power to equipment racks in a data center.
This equipment might include the following:
- Uninterruptible Power Supply (UPS) units.
- Generators with automatic transfer switches.
- Main power distribution units provide power to the data center racks from incoming utility feeders (AC mains).
- Automatic Transfer Switches (ATS) or Manual Transfer Switches (MTS).
Because the purpose of these devices is to isolate the secondary side of the transformer from the primary side, they are also called isolation transformers.
A facility engineer will install these devices to prevent damage to the equipment listed above. The transformer may only succeed if it is appropriately rated in a data center environment. It will not provide electrical isolation between the incoming utility power supply and circuit feeding a critical load such as server or storage enclosures.
Worth to read: Reducing Emissions in Data Centers as a Today Top Priority
Data Center Transformer Protection
There are several types of protection based on their characteristics:
- ground fault,
- phase-to-phase fault,
- and phase-to-ground fault protection.
Different types of protection are used for various applications.
Overcurrent protection is the most common type of fault protection used in data centers. It prevents equipment damage, such as cables and circuit breakers, from current overloads caused by short circuits or individual equipment failures. The most common type of overcurrent protection used in data centers is the circuit breaker; however, some applications in telecom facilities use fuses instead of circuit breakers.
Voltage protection can be applied at either the source or the power distribution equipment and is used to prevent damage to equipment due to voltage overloads caused by miswiring or cable faults. This type of protection has two basic configurations: line-to-ground voltage protectors and line-to-neutral voltage protectors. The line-to-neutral voltage protectors are typically preferred in telecommunications facilities because they provide a better indication of a fault condition than line-to-ground protectors.
Ground Fault Protection
Ground fault protection is primarily a safety issue that prevents electric shock when an unintentional current flow occurs between an energized conductor and one or more grounded conductors. There are two basic types of ground fault protection: differential protection and ground fault trip circuits.
Ground fault trip circuit devices (GFCIs) are considered the most reliable type of ground fault protection because they are designed to detect even tiny ground faults. Differential protection devices are less commonly used, but they work by detecting voltage differences between the hot and neutral wires.
Phase-to-Phase Fault Protection
A phase-to-phase fault is a short circuit from one phase to a different phase. The fault current travels from the first phase to the second, bypassing the transformer. The transformer is designed to withstand single-phase faults, but a phase-to-phase fault can cause significant damage to the transformer.
Installing a differential relay is the best protection against this type of event. The differential relay consists of two relays mounted next to each other. The first relay detects phase-to-neutral faults, and the second relay detects phase-to-phase faults.
If a fault occurs, then the differential relay actuates to trip all three sets of breakers at once. This allows for much quicker clearing time than if only one set of breakers is tripped. The differential relay also aids in fault location because it will automatically clear once the reset button is pressed.
Phase-to-Ground Fault Protection
A phase-to-ground fault is a short circuit in the power distribution system. It is one of the most hazardous conditions in a distribution system and requires immediate action to prevent equipment damage, personal injury, or death. Under normal operating conditions, the voltage between any two phases will be nearly identical concerning the ground.
However, if a fault exists between phases, the internal impedance of the transformer will cause current to flow from one phase to another. This causes an instantaneous change in voltage magnitude and polarity between these two phases relative to the ground, which can result in severe damage to equipment and personal injury or death.
- The first line of defense against phase-to-ground faults is the transformer’s insulation system which must be rated for operation under such circumstances. If a fault should occur, it will be contained within the transformer until corrective action is taken by either automatic reclosing or manual switching.
- The second line of defense is provided by overcurrent protection devices, which typically consist of fuses or circuit breakers.
Transformers are also equipped with differential relays to detect phase-to-ground faults. Other differential protection schemes, such as current transformers, are used to protect specific equipment downstream of the transformer, such as motors, circuit breakers, and switchgear.
The data center isolation transformer is designed to increase the voltage to levels higher than the primary supply of electricity that enters a building. The higher voltage is used to power servers within the data center, which requires more energy than typical household appliances.
The average home in the United States uses about 11 kilowatts/hour per day and has a peak usage of about five kW/h. For comparison, an individual server at a large data center may use 2-3 MW/h and peak at over 10 MW/h.
When used in applications such as the ones described here, data center isolation transformers are designed to provide a relatively low voltage output that can be safely connected to a home or office where people may be present.
The lower voltage output also reduces the need for expensive fiber optic systems by enabling longer cable runs in which signal loss is less of a concern than at higher voltages.