formance at the rated load
on the nameplate. The
input power applied to the
machine is set to the nameplate values for frequency
and volts.
The dynamometer
then loads the machine to a
specific torque value and the
test operator measures and
records the volts, amperes
and watts, and resulting
speed of the machine. This
step is repeated for various torque values, giving
the tester the total losses
under a loaded condition.
Accuracy of the measured
motor losses is enhanced by
allowing the machine to stabilize at the operating temperature and performing the
test at the rated load. Testing the efficiency of an 800-hp, 5012 Frame, 2-pole, 5300-lb vertical solid shaft motor.
A no-load test is then Photo credit: Jim Stroope and Tom White
performed at different input
voltages to measure the resulting amps and watts. This helps
determine the core, friction and windage losses of the machine
under the assumption that they are independent of the load.
With this data, the stray loss can be calculated by subtracting
the sum of the measurable losses from the total losses.
The concept of “smoothing” is a statistical method that
correlates the resulting stray loss, minimizing error and optimizing the reliability of the data (refer to IEEE 112 – 6. 4. 2. 8
for details). A clear indication that the test results are valid is
when the correlation factor is 0.9 or larger.
Method F: By Equivalent Circuit
Methodology
IEEE 112 Method F relies on an equivalent circuit calculation to determine efficiency. Data on items such as amperes,
watts and winding temperature are determined through various tests, including a Direct Current test, a locked-rotor test
and a no-load test.
But because a dynamometer is not involved in the process, characteristics, such as stray loss and rotor resistance,
need to be obtained by completing a set of calculations. These
may or may not represent the motor under actual running
conditions.
Under Method F, the stray losses are determined either by
taking direct measurement or by using an assumed percentage
of rated output that is determined in Table 2 of IEEE 112. By
using the predetermined percentage, stray losses may not have
a direct relationship to the design, nor do they take into consideration the machine’s manufacturing variation. Similarly,
the direct measurement option obtains results through a complicated process of calculations open to the chance of human
error.
Take, for example, the stray losses of a standard efficient
motor versus a premium efficient motor. While their efficien-cies are different, IEEE 112 Table 2 treats their stray loss values
the same. Stray losses are typically 8 percent to 15 percent of
the total losses of the machine. If the table is calculating the
stray losses as 8 percent of the total losses, but the stray losses
are actually 15 percent, the overall efficiency rating of the
motor may be operating 0.3 percent lower in efficiency.
The only way to uncover the stray-loss differences among
the motors under Method F is to measure the fundamental
frequency and high-frequency components of the stray-load
loss and then calculate the sum of the results. In addition to
the complications associated with this process, the calculations
have the capability to produce erroneous results. Although
Method F offers a way to measure several variables when
efficiency cannot be measured at the motor’s rated operating
point, the opportunities for results that differ from Method B
extend well beyond stray losses.
Another problematic issue related to Method F is the calculation for rotor resistance, which determines stator losses,
and the assumption used. Stator and rotor leakage reactance
are portions of magnetizing reactance and the coefficients of
these variables are typically assumed.
Method F calls for measuring these resistance and reactance parameters with a locked-rotor test that produces a significant disconnect from the actual operating point. Under
the locked-rotor test, alternating current frequency for rotor
