Transformer losses are broadly classified as no-load and load losses. No load losses occur when the transformer is energized with its rated voltage at one set of terminals but the other sets of terminals are open circuited so that no through or load current flows.
In this case, full flux is present in the core and only the necessary exciting current flows in the windings. The losses are predominately core losses due to hysteresis and eddy currents produced by the time varying flux in the core steel.
Load losses occur when the output is connected to a load so that current flows through the transformer from input to output terminals. Although core losses also occur in this case, they are not considered part of the load losses.
When measuring load losses, the output terminals are shorted to ground and only a small impedance related voltage is necessary to produce the desired full load current. In this case, the core losses are small because of the small core flux and do not significantly add to the measured losses.
Load losses are in turn broadly classified as I2R losses due to Joule heating produced by current flow in the coils and as stray losses due to the stray flux as it encounters metal objects such as tank walls, clamps or bracing structures, and the coils themselves. Because the coil conductors are often stranded and transposed, the I2R losses are usually determined by the d.c. resistance of the windings.
The stray losses depend on the conductivity, permeability, and shape of the metal object encountered. These losses are primarily due to induced eddy currents in these objects.
Even though the object may be made of ferromagnetic material, such as the tank walls and clamps, their dimensions are such that hysteresis losses tend to be small relative to eddy current losses.
Although losses are usually a small fraction of the transformed power (<0.5% in large power transformers), they can produce localized heating which can compromise the operation of the transformer. Thus it is important to understand how these losses arise and to calculate them as accurately as possible so that, if necessary steps can be taken at the design stage to reduce them to a level which can be managed by the cooling system.
Other incentives, such as the cost which the customer attaches to the losses, can make it worthwhile to find ways of lowering the losses. Modern methods of analysis, such as finite element or boundary element methods, have facilitated the calculation of stray flux losses in complex geometries.
These methods are not yet routine in design because they require a fair amount of geometric input for each new geometry. They can, however, provide useful insights in cases where analytic methods are not available or are very crude. Occasionally a parametric study using such methods can extend their usefulness beyond a specialized geometry.
No comments:
Post a Comment