Power Transformer Protection: Goals and Common Schemes

Why transformers need protection

A power transformer may appear to be a stationary device placed quietly on site, but once an inter-turn short circuit, inter-phase short circuit, or ground fault occurs inside it, the energy release is often very concentrated, leaving a very short time window for manual processing. Short-circuit current from the external power grid can also bring thermal stress and axial and radial electric power to the windings, which may cause winding displacement, deformation or support loosening. Therefore, the first goal of protection is never to simply trip the circuit breaker. It is to quickly remove internal faults, minimize equipment damage as much as possible, and at the same time provide alarms or backup actions for overload, overexcitation, external faults and cooling abnormalities. Whether a set of protections is qualified depends on whether it can cover the accident area before the fault expands.

There is no universal set of fixed values for protection schemes

Many people think that equipping a transformer with a fuse or an overcurrent relay is sufficient. In fact, power transformer protection schemes must be configured in combination according to equipment capacity and structure, system grounding method, available fault current, importance, and fault consequences. Details such as the number of windings, tap-in method, and available circuit breakers will affect which protections should be provided and how the fixed values should be given. There is no universal specification that can be determined solely by kVA, nor can the configuration of one project be copied to another as is.

How differential protection identifies internal faults

Longitudinal differential protection is one of the most common main protections. Its principle is to compare the current on each side of the transformer. When a phase or ground fault occurs within the protected area, the relationship between the currents on both sides will be broken, and the differential element will act accordingly. In actual rectification, a lot of compensation needs to be done, including ratio change, coupling group phase shift and current transformer error. Another point that is easily overlooked is that the excitation surge during normal closing may also manifest as differential flow, so differential protection usually uses harmonic or waveform criteria to avoid erroneous actions.

Overcurrent, heat and overexcitation are spread out one by one tube

Overcurrent protection is mainly used as a backup for external short circuits or internal faults. Thermal protection focuses on the risks of overload and winding hot spots, while V/Hz protection targets core overexcitation. Liquid-immersed products can also be configured with gas, burst pressure, pressure relief, and oil temperature protection, while dry-immersed products typically rely on winding temperature, fan status, overcurrent, and differential. The ground fault block, limiting ground fault protection, can improve the sensitivity of ground faults in specific windings, but its applicability depends on the configuration of the neutral point and current transformer.

We must think about protection, selection, and operation together

A complete power transformer protection should be implemented from the selection stage. The rated values and curves of fuses or overcurrent protection must simultaneously avoid permissible loads and excitation surges, and faults must be cleared within the equipment’s tolerance and the protection range of upper and lower stages. Before the first excitation, the impact of excitation surge on differential and overcurrent protection must be specifically evaluated to confirm that the circuit breaker, protection and system voltage conditions are all permissible. We must not arbitrarily expand the protection value just to close the gate without jumping, as that would be tantamount to negating the protection function.

What to do after tripping

Re-delivering power after a fault trip is not a matter of closing the gate. First, confirm the fault category, evidence of protection actions, equipment status, and system risks before deciding on the next step. For signals such as protection actions, abnormal gases, pressure release, and obvious deformation, the equipment should be isolated first, event records should be saved, and repeated trial delivery should not be performed.

A truly reliable design is to view power transformer protection schemes as a whole that is linked to transformer capacity, grounding system, and fault level. Differential, overcurrent, thermal protection and ground fault protection each perform their respective duties and are connected together through fixed-value coordination. Once you understand the protection objectives and then go back to configure each level of protection, the plan will naturally become clear.