Transformer Loss Evaluation: A and B Factors, DOE Standards, and Procurement Strategy

No-Load Losses vs Load Losses

Transformer losses fall into two distinct categories with fundamentally different characteristics. No-load losses (also called core losses or iron losses) occur whenever the transformer is energized, regardless of whether it is carrying any load. They are caused by the alternating magnetic flux in the core material, which induces hysteresis and eddy current losses. No-load losses are constant 24 hours a day, 365 days a year, for the entire life of the transformer.

Load losses (also called copper losses or winding losses) occur only when current flows through the windings and are proportional to the square of the load current. At 50 percent load, load losses are 25 percent of their full-load value. At 75 percent load, they are 56 percent. At full load, they reach 100 percent of the rated value. Most distribution transformers operate at average loading of 40 to 60 percent, meaning average load losses are typically 16 to 36 percent of rated full-load losses.

For a typical medium-voltage distribution transformer in the 500 to 2,500 kVA range, no-load losses might be 0.15 to 0.40 percent of rated capacity and load losses might be 1.0 to 2.0 percent at full load. Because no-load losses are constant and load losses are intermittent, the lifetime energy cost split between the two categories depends heavily on the transformer's loading profile. For lightly loaded units (below 30 percent average), no-load losses dominate total energy cost. For heavily loaded units (above 60 percent average), load losses dominate.

Understanding this split is essential for specification writing. A transformer with very low no-load losses but moderate load losses is ideal for a lightly loaded or variable-load application. A unit with moderate no-load losses but very low load losses is better for a consistently heavy-load application. The A/B factor method quantifies this tradeoff in dollar terms.

The A/B Factor Loss Evaluation Method

The A/B factor method converts future energy losses into present-value dollars so that transformers with different loss levels can be compared on a total owning cost (TOC) basis. The A factor represents the present value of one watt of no-load loss over the transformer's evaluated life. The B factor represents the present value of one watt of load loss at rated load. Total evaluated cost equals purchase price plus (A times no-load losses in watts) plus (B times load losses in watts).

The A factor is calculated from the cost of electricity, the hours per year the transformer is energized (typically 8,760 for continuous operation), the discount rate, and the evaluation period. For a transformer running 8,760 hours per year at an electricity cost of $0.08 per kWh, a discount rate of 8 percent, and a 20-year evaluation period, the A factor is approximately $6.87 per watt of no-load loss.

The B factor includes the same parameters as A but also accounts for the transformer's loading profile. Because load losses vary with the square of the load, the B factor uses the loss load factor (LLF), which is derived from the load factor (LF) of the application. A common approximation is LLF = 0.15 x LF + 0.85 x LF squared. For a transformer with 50 percent average loading, LLF is approximately 0.29, meaning average load losses are 29 percent of rated full-load losses. This significantly reduces the B factor relative to the A factor.

Typical A factor values range from $4 to $10 per watt. Typical B factor values range from $1 to $4 per watt, depending on the loading profile. Higher electricity costs, lower discount rates, and longer evaluation periods all increase both factors. Utilities and large industrial purchasers publish their A and B factors in transformer procurement specifications to ensure vendors optimize their designs for the buyer's specific economic conditions.

DOE Efficiency Standards for Distribution Transformers

The U.S. Department of Energy (DOE) sets minimum efficiency standards for distribution transformers under 10 CFR Part 431. The current standards, which took effect January 1, 2016, establish minimum efficiency levels based on transformer type (liquid-filled or dry-type), kVA rating, voltage class, and number of phases. These are minimum requirements, and loss-evaluated procurement typically drives specifications well beyond DOE minimums.

DOE efficiency is measured at 50 percent of rated load, which represents typical average loading for distribution transformers. For a 1,000 kVA liquid-filled transformer at 35 kV BIL, the DOE minimum efficiency at 50 percent load is approximately 99.1 percent. This means total losses (no-load plus load losses at 50 percent) cannot exceed 0.9 percent of rated output at half load. For dry-type transformers, the minimum efficiency is lower due to the inherently higher losses in dry-type designs.

The DOE has proposed updated standards (expected to take effect in 2027 or 2028) that would increase minimum efficiency by 2 to 10 percent depending on transformer type. Liquid-filled transformers would see the largest improvement requirements, with many sizes requiring amorphous metal core or equivalent low-loss designs to comply. Dry-type transformers would see more modest increases. Monitor the DOE rulemaking process if you are specifying transformers for projects with commissioning dates after 2027.

ENERGY STAR and other voluntary programs recognize transformers that exceed DOE minimums by a specified margin. Some utilities offer rebates for high-efficiency transformers that exceed DOE standards. These incentives can offset the premium for lower-loss units and should be included in the total owning cost calculation.

Finding the Peak Efficiency Loading Point

Every transformer has a specific loading percentage at which its efficiency is highest. This peak efficiency point occurs where no-load losses equal the load losses (adjusted for loading). Since no-load losses are fixed and load losses increase with the square of the load, there is always a crossover point. Below this loading percentage, no-load losses dominate and efficiency is lower. Above it, load losses dominate and efficiency again decreases.

For a conventional silicon steel core transformer, peak efficiency typically occurs at 40 to 60 percent of rated load. For an amorphous core transformer with very low no-load losses, peak efficiency shifts to 25 to 35 percent of rated load because the no-load losses are so small that they equal the load losses at a lower percentage. This means amorphous core transformers are most advantageous in lightly loaded applications where they spend most of their time near peak efficiency.

The peak efficiency loading point can be calculated from the nameplate data. The percentage load at peak efficiency equals the square root of (no-load losses divided by rated load losses), expressed as a fraction. For a transformer with 1,000 watts no-load loss and 5,000 watts rated load loss, peak efficiency occurs at the square root of (1,000 / 5,000) = 0.447, or approximately 45 percent load.

When specifying transformers for a known load profile, matching the peak efficiency point to the expected average loading maximizes energy savings. If your facility typically loads the transformer at 30 percent, specifying a unit with peak efficiency near 30 percent (achievable with amorphous core) saves more energy over the transformer's life than a conventional unit with peak efficiency at 50 percent.

Amorphous Metal Core Transformers

Amorphous metal (also called metallic glass) core material reduces no-load losses by 60 to 80 percent compared to conventional grain-oriented silicon steel. The material is produced by rapidly cooling molten metal alloy onto a spinning wheel, creating thin ribbons (about 25 micrometers thick, compared to 230 to 270 micrometers for silicon steel) with a non-crystalline atomic structure. This structure has very low hysteresis and eddy current losses.

A 1,000 kVA liquid-filled transformer with an amorphous core might have no-load losses of 300 to 500 watts, compared to 1,000 to 1,500 watts for a conventional silicon steel core of the same rating. At $0.08 per kWh and 8,760 hours per year, the energy savings from reduced no-load losses alone is $370 to $700 per year. Over a 25-year life at an 8 percent discount rate, the present value of this savings is $3,900 to $7,400, which often exceeds the price premium for the amorphous core design.

Amorphous core transformers have higher purchase prices, typically 15 to 30 percent more than conventional designs of the same rating. They are also larger and heavier because amorphous metal has lower magnetic saturation than silicon steel, requiring a larger core cross-section. For indoor substations or pad-mounted applications with tight space constraints, the size increase may be a limiting factor.

The economic case for amorphous core is strongest in lightly loaded, continuously energized applications where no-load losses dominate total energy cost. Residential distribution transformers, which average 15 to 25 percent loading, are the ideal application. Commercial and industrial transformers with higher average loading still benefit, but the relative advantage narrows as load losses become a larger share of total losses. The A/B factor calculation will identify which applications justify the amorphous core premium.

Harmonic Effects on Transformer Losses

Non-linear loads such as variable frequency drives, LED lighting, switch-mode power supplies, and electric vehicle chargers draw current in pulses rather than smooth sine waves. This distorted current contains harmonic frequencies (odd multiples of 60 Hz: 180, 300, 420 Hz, etc.) that increase transformer losses beyond what the rated load losses would predict.

Harmonic currents increase load losses through two mechanisms. First, winding I-squared-R losses increase because the total RMS current including harmonics is higher than the fundamental-frequency current alone. A load with 30 percent total harmonic distortion (THD) draws about 4.5 percent more RMS current than a linear load of the same fundamental power, increasing I-squared-R losses by roughly 9 percent.

Second, and more significantly, stray losses in the windings, core clamps, and tank walls increase approximately with the square of the harmonic order. The fifth harmonic (300 Hz) produces stray losses 25 times greater per amp than the fundamental. The seventh harmonic produces 49 times greater. For a transformer supplying a load with significant 5th and 7th harmonic content, total stray losses can double or triple compared to a linear load of the same kVA.

The K-factor rating system quantifies the additional heating from harmonics. K-1 is a standard transformer suitable for linear loads. K-4 handles moderate harmonic loads (office buildings with computers). K-13 handles severe harmonic loads (VFD-heavy industrial plants). K-20 handles the most severe harmonic environments. A K-rated transformer has heavier windings, lower flux density, and better cooling to handle the additional losses without exceeding temperature limits. Alternatively, a standard transformer can be derated (loaded to a lower percentage of nameplate rating) to accommodate harmonic heating.

Writing Loss-Evaluated Transformer Specifications

A well-written transformer specification includes loss evaluation factors that allow vendors to optimize their designs for minimum total owning cost rather than minimum purchase price. The specification should include the A factor (dollars per watt of no-load loss), B factor (dollars per watt of load loss), maximum guaranteed no-load losses, maximum guaranteed load losses, and the penalty formula for units that exceed guaranteed losses.

State both the evaluation factors and the hard limits. Evaluation factors drive the vendor to minimize losses, but hard limits prevent the vendor from offering a design with extremely low losses in one category at the expense of the other. For example, a vendor might offer very low no-load losses with a very heavy (and expensive) amorphous core, while the load losses are only marginally better than standard. Hard limits on both loss categories prevent this gaming.

Require the vendor to guarantee losses with a tolerance of plus or minus zero percent. Any unit that tests higher than the guaranteed no-load or load losses should be subject to a penalty equal to the applicable A or B factor times the excess watts. Any unit that tests lower than guaranteed should receive a credit. This creates a direct financial incentive for the manufacturer to meet or beat guaranteed values.

Include a requirement for factory testing at the transformer's rated conditions with witnessed tests for units above a specified kVA threshold (commonly 2,500 kVA). The test report should include open-circuit test results (no-load losses and exciting current) and short-circuit test results (load losses and impedance voltage). Compare factory test results against guaranteed values to determine if penalties or credits apply.

For large purchases (multiple units or large kVA), consider requesting loss evaluation bids where each vendor submits a base design and an alternative low-loss design. Evaluate both on TOC. The low-loss alternative often has a higher purchase price but lower TOC, and providing both options lets the procurement team make an informed decision rather than defaulting to lowest bid price.