Low-Loss Oil-Filled Power Transformer Core: Grade Selection and Stacking Practice

What Sets No-Load Loss in an Oil-Filled Transformer Core

The core provides the closed magnetic path that links the primary and secondary windings, and unlike the windings it is magnetised for as long as the transformer is energised. No-load loss is therefore a core property rather than a winding property. Hysteresis loss depends on the alloy and on the flux density the designer chooses; eddy-current loss depends on lamination thickness, surface insulation and the quality of the stacking. For 50 Hz networks, cold-rolled grain-oriented (CRGO) silicon steel with roughly 3.0-3.3% silicon and a thickness of 0.23-0.30 mm remains the reference material because it combines high permeability along the rolling direction with low loss at the high flux densities used in power transformer design.

In an oil-filled design the core and windings are immersed in insulating oil, which removes heat from the active part, provides dielectric strength between conducting parts and excludes moisture and oxygen from the insulation system. The core is therefore designed as a compact, well-clamped assembly, because oil circulation must reach the limb and yoke surfaces that carry the loss.

Grade Selection and Guaranteed Loss Values

Grain-oriented grades are self-describing. In the GB/T 2521.1 designation a name such as B27R090 states 0.27 mm thickness and a maximum specific total loss of 0.90 W/kg measured at 1.7 T and 50 Hz; IEC 60404-8-7 specifies the same class of material through thickness, magnetisation at 800 A/m and loss at 1.7 T, 50 Hz. When a buyer moves one step down the loss table the gain is not cosmetic: on a core of several tonnes, 0.1 W/kg of specific loss is a permanent, measurable load on the asset for 25-30 years.

Thin gauges reduce eddy-current loss but raise the number of laminations, the amount of shearing and the risk of burr and short circuits at the cut edges, so 0.23 mm material is normally reserved for units where loss is the binding constraint.

Stacking, Joints and Coating Practice

Loss is not only a material property. Mitered step-lap joints distribute the reluctance of the corner region across several laminations, which limits local flux distortion and the extra loss it causes. Uniform clamping pressure, controlled burr height and clean shearing are equally important: a short circuit between laminations creates a loop for circulating current and can destroy the gain obtained from a lower-loss grade. Insulating coatings specified to ASTM A976 must survive shearing and any stress-relief anneal while keeping interlaminar resistance high; surface insulation resistance is measured to IEC 60404-11. A stacking factor of 0.96-0.97 is normal for coated 0.23-0.30 mm material, and a low stacking factor forces a larger window, more copper and more oil for the same flux path.

Verification Before Dispatch

Core-plate properties are verified on the incoming strip by the Epstein frame method of IEC 60404-2 or by the single-sheet test of IEC 60404-3, and the finished transformer is then proven by the no-load and load-loss tests defined in IEC 60076-1, with temperature rise limited according to IEC 60076-2. Insulating oil is a separate acceptance gate: unused mineral insulating oil is specified to IEC 60296, breakdown voltage is tested to IEC 60156 and water and acidity are held within the limits of that specification, because moisture in oil degrades both dielectric strength and the cooling of the active part.

Frequently Asked Questions

Q: Which grade should be used for a low-loss distribution transformer core?
A: A 0.23-0.27 mm high-permeability grain-oriented grade such as B27R090 at 0.90 W/kg or B23R080 at 0.80 W/kg at 1.7 T, 50 Hz is the usual starting point, with the final choice driven by the guaranteed no-load loss in the purchase specification.

Q: Does a thinner lamination always lower total loss?
A: It lowers eddy-current loss, but it also increases lamination count, handling and burr risk. The net benefit must be evaluated against stacking factor and manufacturing cost, not on gauge alone.

Q: Why is stacking factor important in an oil-filled design?
A: Magnetic material occupies only 96-97% of the built core volume. A poor stacking factor means a bigger core window for the same flux, which raises copper, oil and tank size.

Q: How is core loss proven on the finished transformer?
A: By the no-load test of IEC 60076-1 on the assembled unit, supported by strip tests to IEC 60404-2 or IEC 60404-3 and by coating tests to IEC 60404-11.

Q: What oil properties matter most for the core and winding assembly?
A: Dielectric strength, water content and oxidation stability, specified to IEC 60296 and tested with IEC 60156 methods, because they govern both insulation margin and long-term cooling.

Q: Can a wound or cut core replace a stacked core?
A: For distribution sizes a wound core can reduce joint loss and assembly time, but it constrains the coil winding sequence; the loss guarantee, not the construction, should drive the decision.