A more efficient transformer can pay for itself many times over during its 25-year lifespan

The transformer plays a key role in an electrical system’s efficiency and power quality, yet 95% of buying decisions are based solely on first cost. Buying based on life cycle cost would save literally hundreds of thousands of dollars in operating losses over the installed life for transformers in a typical facility. In fact, the United States Environmental Protection Agency (EPA) estimates that 60 to 80 billion kWh annually can be attributed to transformer losses. These losses cost end-users $3 to $4 billion, and can tie up nine days of U.S. generating capacity annually.

When feeding the increasingly electronic nature of connected equipment, distortion of the voltage waveform can reduce the operating reliability of both the electrical system and the connected equipment. Users can lower energy rates and avoid high transformer losses by investing in energy-efficient transformers. This article compares the efficiency of different transformer types including low temperature rise, Energy Star, K-rated, as well as new transformers that are designed to minimize life cycle cost.

Underestimating loss data. Electronic equipment and other nonlinear loads now make up most of the load on transformers in many facilities. Even in the average office, many individuals plug in mostly computers, printers, scanners, and other electronics to 120V receptacles. The load profile of electronic equipment—from the computer in the office to the variable speed drive in the factory—drives both additional losses and unwanted distortion, according to IEEE Standard 519-1992. Since transformer manufacturers test only under ideal (linear) conditions, as called for in present construction standards, a substantial gap exists between published loss data and actual losses incurred after installation (Fig. 1). In fact, test results published in a 1996 IEEE Transaction paper documented an almost tripling of transformer losses when feeding 60kW of computer load rather than linear load.

Transformers have two major components that drive losses: the core and the coils. The typical core is an assembly of laminated steel. Core losses are mostly related to magnetizing or energizing the core. These losses, also known as no-load losses, are present the entire time the transformer is powered on, regardless of whether there’s any load or not.

Core losses are roughly constant from no-load to full-load when feeding linear loads. They represent a continuous cost, 24/7, for the 25- to 40-year life of the transformer. A common 75kVA commercial transformer has about 400W in no-load losses. At $.10/kWh, this represents a continuous cost of $350/yr or $14,000 over a 40-year life, eclipsing the purchase cost many times over. And remember, this is just the cost for powering the unit. The cost of powering the load itself far exceeds this cost.

The coil losses, commonly referred to as load losses, are associated with feeding power to the connected load. For linear loads, these losses are predominately I2R losses. In other words, load losses increase by the square of current from no-load to full-load, driven by the resistance of the coil. Fig. 2 shows a graphical representation of how transformer losses increase with loading.

Since a wide variety of transformers serve different purposes, actual losses incurred in the field will vary substantially from one installation to another. Load level varies widely, with some installations running very heavily loaded and others more lightly loaded. This difference substantially affects actual losses incurred.

To calculate the cost of these losses, one must refer to the billing structure of the electric utility. This varies across the country and may involve kWh, kW peak demand, and kVA charges. Taking an example from Fig. 2, at 60% load the transformer has about 1,500W of losses. If the user is being billed only on kilowatt-hour consumption at a rate of $.10/kWh, the operating cost would be 1.5kW x $.10/kWh, which equals $.15/hr or roughly $1,300/year—the same order of magnitude as the purchase price of the transformer. Although some utilities charge by kVA or kW, most charge a combination of a kWh rate and a peak demand charge. Additional distribution or environmental costs are also common surcharges included in electrical bills, so be sure to look beyond the cost per kWh.

Comparing transformer losses. Only a limited amount of field data is available on transformer losses due to the high cost of gathering detailed data from a reasonable number of individual transformers. Faced with this lack of comprehensive field data, the remaining graphs in this article represent our years of field experience with a combination of published efficiency data under linear and nonlinear load conditions and independent testing, as well as before/after field measurements to build a series of representative loss curves for different transformers as accurately as possible with the data available.

Standard transformer—The standard transformer is built to deliver its nameplate kVA rating under linear load only and is UL Listed on this basis. As it has the lowest purchase price on the market, it represents the majority of transformer purchases made across the country. When feeding electronic equipment, substantial derating is required—on the order of 50% or more—to prevent overheating and premature failure, according to IEEE Standard 1100-1992. Along with its high operating cost, other factors include a substantial loss in capacity and distortion of the voltage to connected equipment.

Low temperature rise transformer—Transformers with a low operating temperature rise have often been purchased with energy savings in mind, as published full load losses are substantially lower than those of many other transformers. These transformers are traditionally available in either 80°C or 115°C operating temperature rise, as opposed to the standard 150°C rise that represents the majority of low-voltage, 3-phase, dry-type transformer sales.

The low temperature rise transformer is designed to run cooler than a standard transformer when fully loaded. To meet this objective, manufacturers typically use a larger core and winding set, resulting in higher no-load losses (more core), but lower load losses (more coil). Since total losses are the sum of both core and coil losses, the low rise transformer will have higher losses than other transformers at low load levels where core losses predominate, but lower losses when heavily loaded, since coil losses predominate at high load levels.

From Fig. 3, it’s evident that at less than 60% load, it actually costs more to operate the 80°C rise transformer than the standard 150°C rise transformer. Depending on the size and manufacturer, the break-even point can be as high as 80%. Since many transformers are loaded to less than 50% capacity, use of an 80°C rise transformer is often a commitment to higher energy costs—the exact opposite of what was intended. Another limitation with the low temperature rise transformer is that it’s UL Listing applies when feeding linear loads only.

Energy Star transformer – In 1998, the EPA included a high-efficiency transformer program under the Energy Star banner. For a reference document, the EPA settled on NEMA TP-1 Guide for Determining Energy Efficiency for Distribution Transformers. The NEMA TP-1 standard establishes required efficiencies at 35% load for low-voltage, dry-type transformers, and at 50% load for liquid-filled and medium-voltage, dry-type transformers. In a bid to move the first-cost driven market to higher efficiency transformers, several states including New York, California, Minnesota, and Massachusetts, adopted NEMA TP-1 into law.

However, NEMA TP-1/ENERGY STAR transformer efficiencies reference test data under linear load conditions. This results in published efficiencies that are much higher than experienced in the real world due to the additional losses associated with the widespread use of electronic equipment. Ironically, transformers feeding harmonic-rich loads are exempt from meeting NEMA TP-1 benchmark efficiencies. Like the standard transformer, the Energy Star transformer is built to deliver its nameplate kVA rating under linear load and is UL Listed on this basis. And like standard transformers, Energy Star transformers exhibit increased losses, loss of capacity, and increased voltage distortion when feeding electronic equipment. The Energy Star-compliant transformer is more efficient than the standard transformer as shown in Fig. 4 (linear loading).

K-rated transformer—Unlike standard transformers, which are designed to feed linear loads only and lose capacity when feeding nonlinear loads, K-rated transformers are designed to feed nonlinear loads with harmonic content up to their nameplate rating. The UL Listing is maintained as long as the load profile has a K-factor lower than the K-rating of the transformer. Industry standard ratings include K4, K13, and K20, with K4 and K13 being the most frequently specified. A higher K-rating represents the capability to withstand higher harmonic content.

K-rating is a heat survival rating, not a treatment of associated power quality issues like voltage distortion, and efficiency isn’t typically discussed. Surviving the extra heat means using more core and coil material, and sometimes use of different construction techniques. Depending on the manufacturer’s design, harmonic losses may be reduced to varying degrees. Ironically, even though the designated use of the K-rated transformer is to feed nonlinear load, manufacturers publish their loss data under linear load conditions.

The need for commissioning. When energy savings are driving part or all of the justification for selecting a particular transformer, it’s important that these savings are indeed present once the transformer is installed. This means commissioning the transformer for energy performance after installation. In fact, some rebate programs and other life cycle-oriented programs like Leadership in Efficiency and Environmental Design call for ongoing product commissioning.

As electronic equipment has become more integrated into our daily lives, transformer losses have added a substantial hidden energy cost to the overall operating costs of many buildings. If properly applied, energy-efficient transformers can help deliver substantial energy savings and power quality improvements.

Ling is vice president – technology of Powersmiths Corp. in Irving, Texas.



Sidebar: Pay Me Now or Pay Me Later

Electric utilities have traditionally purchased their distribution transformers based on life cycle costing, or total cost of ownership (TCO), where the cost of losses is factored into the buying process—as they understand the cost of ongoing operating losses. As a result, the trend in this market segment is the use of higher efficiency transformers.

The opposite trend is in place in the commercial/industrial world, where the standard low voltage step-down transformer is widely considered a commodity. The only perceived differentiator is upfront cost, or purchase price, since the lowest first cost wins.

Commercial transformer specifications rarely set a minimum efficiency requirement. As you would expect, building a less efficient transformer is cheaper than building a more efficient one, so a typical low-first-cost transformer will have a low upfront cost but substantially higher operating cost. And the lifetime cost of the operating losses far exceeds the purchase cost.

The typical buying process makes the situation worse. Traditionally, the consulting engineer specifies a generic transformer, and a contractor purchases it from a wholesaler. The contractor typically focuses on first cost since the winning bidder is based on providing the lowest bid. The end user, who pays the electricity bill for the next 40 years, is neither involved in the selection process nor educated about the true operating cost of the unit or the potential savings from using a more efficient unit. As a result, the fight is over first cost, which for the transformer is on the order of 4% of the life cycle cost. In the end, the end user is stuck with high operating cost, the other 96% of the life cycle cost.