In the September issue of ELECTRICAL CONTRACTOR, there was an Industry Watch story titled “Efficiency from the Wall to the PC,” about an organization that is campaigning for more efficient power “cords.” There was a reference to a quote indicating only 50 percent of the power that leaves the outlet reaches the PC because energy “leaks” out of inefficient power cords. However, a typical PC line cord has 0.3 ohms of resistance. If the computer consumes 720W, that is 6A on 120V circuit. The power loss in the cord is 6A^2*0.3 ohm = 7.8W. Relative to the computer itself, that is approximately a 1 percent loss of power delivered from the outlet, a 99 percent efficiency rating.
A few readers wrote to us about the inaccurate terminology. It should have read power “supplies,” and, of course, energy does not literally “leak” out of the power cord. It is lost within the inefficient components, such as the computer’s power supply.
As with most power quality issues, Ohm’s and Kirchoff’s Laws still apply here. These rules are how we calculate the voltage and current through an electrical distribution system or an electronic circuit. Next, we need another similar equation, where power = voltage × current (though not necessarily a simple math multiplication in most cases with today’s systems). Finally, the formula for efficiency is useful power output divided by the total electrical power consumed.
In electrical engineering terms used by the utility companies, this definition equates to the parameter called “power factor” (PF). Many people today use the term “true power factor,” which is the watts (W) divided by the volt-amperes (VA). This agrees with the definition above. The W is the power used by the load. It is divided by the VA, which is the power delivered by the utility company, to obtain the PF.
Of course, having a term called “true power factor” implies there is another term called “false power factor,” which isn’t correct; the opposite term is “displacement power factor” (DPF), which is the cosine of the angle between the voltage and current. In the old days prior to harmonic distorted and unbalanced systems, PF and DPF would be equal—but not anymore. However, that is a discussion for another day.
The Web site from which the misworded quotes originated explains the inefficiency of the individual voltage regulators on other boards within the computer, as well as the load of the heating, ventilating and air conditioning (HVAC) systems to remove the heat from the building, and so on, further reducing the efficiency. Since most HVACs today use adjustable speed drives to be more efficient, the efficiency of the heat removal depends on the load that is directly impacted by the ambient temperature and emitted temperature from the loads, among other factors.
The article on the Web site references an Electric Power Research Institute study on the efficiency of power supplies themselves. Efficiency of just a computer’s power supply depends on a number of factors, including the percentage of full-rated power that is being drawn from power supply, operating temperature, input AC voltage level and so on. What this graph doesn’t tell is, at lower loading, the current harmonic distortion goes up. So the real efficiency would have to account for the increased losses in the power transformers due to harmonic currents. Of course, the lower load means less power itself is being consumed, hence less carbon, energy and, therefore, money.
You can see that determining the true efficiency of a computer needs accounting for many variables that can make the answer vary quite significantly. So, I advise anyone to take any claims of money savings through efficiency with a grain of salt. What I really think is important is not the efficiency, but the effectiveness of the power supplied. Efficiency should not be confused with effectiveness: A system that wastes most of its input power but produces exactly what it is meant to is effective but not efficient. The term efficiency only makes sense in reference to the desired effect.
Examples include the incandescent light bulb—2 percent efficiency at emitting light; electronic amplifier—50 percent efficiency to speakers; and electric kettle—90 percent efficiency in boiling water. In a computer, the efficiency is significantly affected by the microprocessor. This has been brought to the forefront with AMD’s introduction of its newly minted quad-core processor, which is code-named Barcelona. According to InfoWorld, “the chip delivers more than twice the combined integer and floating-point performance of its two-core predecessor at the same thermal envelope … . That, to me, is a textbook example of a green technological advancement. It means AMD has minted a processor with significantly higher performance per watt than its predecessor, a metric that’s becoming increasingly important to datacenter operators.”
So to readers who spotted the miswording that, if true, would have resulted in changing the laws of physics with regards to power cords efficiency, thank you for helping to get this corrected. And to those concerned about efficiency, perhaps we should broaden the scope to encompass the complete picture of effectiveness—what power the electric utility must supply to perform the work the consumer requires, whether a computer or an entire facility. EC
BINGHAM, a contributing editor for power quality, can be reached at 732.287.3680.