Issues to Consider When Substituting Large Power Transformers in Generating Stations
The power transformer is a reliable device, yet not failure-free. It has no rotating parts, consequently neither the typical faults of rotating machines. On the...
byRelu Ilie, Isidor Kerszenbaum
The power transformer is a reliable device, yet not failure-free. It has no rotating parts, consequently neither the typical faults of rotating machines. On the other hand, the large transformers are oil-immersed and suffer from other faults, mainly chemically or electrically related. A transformer internal fault may be very difficult to locate and to repair. In many cases, the owner may decide that it is faster and cheaper to buy a new transformer than to repair an old damaged one. Even when the repair is worthwhile, it may last for many months.
Almost all large transformers used in power plants are custom-designed. Keeping in storage suitable spare transformers is a common practice, for the purpose of avoiding long unplanned outages and high economic losses. In case an exact replacement is not available, the only option the generating utility has is to search for a substitute transformer that, as a minimum, is able to offer a temporary solution that may introduce some operational constraints. The goal of this paper is to mention the most important aspects to be considered when checking the interchangeability of such substitute transformer. The discussion is based on the various requirements included in relevant American and European standards.
The most common generating station arrangements are shown in Fig. 1: a unit generator-transformer block configuration, and a unit generator-transformer with generator breaker. The vital large transformers in a power plant are the unit step-up/main/generator transformer (UT), the auxiliary transformer (UAT) and the station service/reserve transformer (SST). A failure of any one of these transformers may lead to unit shutdown or start-up unavailability.
General layout
The UT may consist of a single three-phase unit, two half-size three-phase units, or three single-phase units. This is the most evident aspect to consider when a substitute transformer is needed.
When designing a new plant, the selection among these alternatives is generally based on consideration of some form of strategic reserve, as well as available space. Two half-size transformers may be selected in place of a single full-size transformer in order to reduce the cost of the spare. For similar reasons, a generator transformer may be made as a bank of single-phase units. Normally the cost, mass, and loss of such solutions are larger than for a single three-phase transformer; however, they may be preferable if transport size or weight limits apply. The three single-phase transformers provide independent magnetic circuits (see dedicated section below), representing high magnetizing impedance for zero-sequence voltage components. Therefore, a delta equalizer winding is normally provided, implemented by external connection between phase units.
Some layouts may further complicate the spare transformer availability, like UAT with three-winding design.
Dimensions and weight
Any physical size and weight limitations should be checked, for example for installation on an existing foundation. Special installation space restrictions may influence the insulation clearances and terminal locations on the transformer.
UT and UAT are connected to the generator through isolated phase bus ducts. The high/extra high voltage terminals of the UT may be connected to gas insulated switchgear. The medium voltage connections to plant auxiliaries are also normally done via rigid, non-segregated phase bus bars or cables. All these aspects must be checked and solved when looking for a transformer replacement.
Rated frequency
In the unlikely situation a transformer designed for a different frequency is considered, the following applies. The rated frequency radically affects the transformer design and operation. The general formula of voltage e induced by a variable flux φ in a coil with N turns is e = -N dφ/dt. Assuming a sinusoidal flux φ = Φm cos ωt, the induced voltage becomes e = ω N Φm sin ωt. Its rms value will be, in terms of core cross section A and flux density Bm [1]:
E = 2π/√2 f N Φm = 4.44 f N A Bm. (1)
According to (1), operating at 50 Hz a transformer designed for 60 Hz means that the same voltage can be achieved only by a substantial increase in the flux density. The core iron will become heavily saturated, the excitation current will rise and also the hysteresis losses (proportional to the area of the hysteresis loop), which could severely overheat and damage the laminations.
When operating at 60 Hz a transformer designed for 50 Hz, there is too much iron in the core. The hysteresis losses will be higher than the 50 Hz designed value (because iron volume and increased frequency), thus decreasing the efficiency. More importantly, the eddy currents losses will heat up the laminations, because they tend to increase as the square of the frequency. Operation of a 50 Hz rated transformer at 60 Hz may be sometime possible, but it may have to be derated from its nameplate MVA rating [2], depending on the new versus rated voltage.
To read the article, subscribe and choose the option which suits you best. We offer both free and paid options, and the registration takes only a minute.
Subscribe to Transformers Magazine
