POWER January 2015 - 46

EMISSIONS
2. Performance comparison. The VI curve for the single-phase base case and two lowripple
power supplies are compared. The two low-ripple options produce 20% higher voltage,
thus producing higher ESP particulate removal rates. Source: Babcock and Wilcox Power Generation
Group Inc.
1. Single phase 2. 3-phase 3. Switch mode
2,000
Sparking limit
by the spark comes from the energy stored in
the capacitance of the ESP field and not the
power supply.
Care should always be exercised with large
ESP fields (which increase capacitance)
and wide plate spacing (which increases
voltage), as spark energy is directly proportional
to the capacitance and the square of the
voltage.
Total Harmonic Distortion. ESP power
1,500
Peak-to-average-voltage ratio ≈ 1.0
1,000
Peak-to-average-voltage ratio ≈ 1.2
500
supplies connect to the power line and draw
power at the fundamental frequency and at
harmonic frequencies, which are whole number
multiples of the fundamental frequency.
This nonlinear load causes distortion of the
input waveform and can cause many problems
in the electrical distribution system,
including heating of conductors, nuisance
breaker trips, and interference with other
plant equipment. Therefore, it is important to
have a measurement of how much distortion
exists for each power supply type.
One widely accepted measurement is to0
10
20
30
40
50
Secondary
voltage (kV)
structed that considered the entire mechanical
configuration of the ESP. For example,
different discharge electrodes were configured
at various plate spacing, and ESP problems
such as close clearances and tracking
insulators were studied.
One example of the results of laboratory
testing compares the benefits of a low-ripple
power supply (Figure 2). In each test, the
power supply was operated from zero power
to the point where sparking occurred in the
ESP, and then a typical average voltagecurrent
(VI) curve was plotted. The selected
discharge electrode and the physical configuration
of the ESP determined the shape of the
VI curve and, as expected, the three power
supplies track along the same curve.
In all three cases, ESP sparking occurred
at 71 kV, which limited the power supply
from producing higher voltage. The high
ripple on the output voltage of the singlephase
power supply (shown as " 1 " in Figure
1) is clear. The peak voltage was 71 kV with
an average 57 kV, which produces a peak-toaverage-voltage
ratio of about 1.2. For the
low ripple three-phase and high-frequency
SMPS (points 2 and 3, respectively) also
operated with a peak 71 kV voltage and an
average 71 kV, the peak-to-average-voltage
ratio is ~1.0. A reduction in the peak-toaverage-voltage
ratio from 1.2 to 1.0 results
in a 20% increase in voltage. The net result
46
is up to ~35% more average current and
~50% more corona power available to the
ESP, which should result in additional ESP
particulate removal efficiency.
More Performance Issues
There are other power supply performance
issues that have significant impact on ESP
performance, such as spark and quench, harmonics,
site application, cost, and reliability.
A slate of four power supply sizes (24 kW,
32 kW, 72 kW, and 120 kW) were compared.
Commercially available, roof-mounted ESP
low-frequency single-phase and high-frequency
SMPS power supplies were compared
with the low-frequency, three-phase power
supply for each of the four design sizes.
Spark and Quench. When a spark occurs,
it dissipates all of the energy stored
in the ESP field and then the spark is extinguished.
In response to the spark, the power
supply quenches or turns off for a period of
time and then reapplies power to recharge the
ESP field.
However, the power supply does not turn
off the instant the spark occurs. There is a delay
based on the type of power supply, and
during this delay, energy continues to be delivered
to the spark from the power supply.
Each of the power supplies considered delivers
less than 0.1% of the total spark energy.
The majority of energy (>99.9%) dissipated
www.powermag.com
60
70
80
90
tal harmonic distortion (THD), which is a
summation of all of the harmonics present in
the system. The modeling results found that
the low-frequency design options exhibit the
lowest input THD and can therefore be expected
to provide significantly fewer installation
and maintenance harmonics problems.
ESP power supplies also produce harmonics
at the output. The DC waveform is made
up of many frequencies, including a fundamental
frequency and its harmonics. This
is particularly troubling in ESP power supplies
because its ground is a current-carrying
power lead and is energized with harmonic
frequencies. Since all of the plant equipment
and the neighboring facility plant equipment
are connected through ground, the potential
exists to cause interference with other plant
equipment, including other ESP power supplies.
This is particularly true as radiated radio
frequency (RF) emissions increase with
frequency. Manufacturers provide detailed
bonding and grounding specifications in
high-frequency designs, which must be meticulously
followed. Low-frequency designs
exhibit the lowest output THD and normally
experience fewer harmonics problems.
The physical internal electrical connections
inside the ESP are also important.
Historically, the ESP was constructed for
low-frequency operation with bolted or
friction fit connections. Both connection
types may be inadequate for high-frequency
operation, which may lead to voltage
drop at the connections, both in the highvoltage
distribution system and the ground
system. Also, voltage drop in the internal
ground connection causes crosstalk and
interference between ESP power supplies,
which is very difficult to detect, particularPOWER
| January 2015
Secondary current (μA)

http://www.powermag.com

POWER January 2015

Table of Contents for the Digital Edition of POWER January 2015