Technical Overview of Broadband Powerline (BPL) Communication Systems

1
Technical Overview of Broadband Powerline (BPL)
Communication Systems

• Robert G. Olsen
• School of Electrical Engineering and Computer
  Science
• Washington State University
• Pullman, WA, USA
• bgolsen_at_wsu.edu

Presented at University of Idaho October 4, 2006
2
Injector (2 30 MHz)
Photo of a Typical BPL Installation (courtesy B.
Cramer EPRI)
3
The BIG Question
Do attenuation, background noise and restricted
input power due to emission limits result in the
need for financial investment (i.e., for
additional equipment, system conditioning, or
maintenance) that is incompatible with the
requirement that a BPL system be profitable
and/or useful to the electric utility? Here, we
will look at attenuation and emission issues.
4
Sources of Channel Attenuation

• Ohmic Absorption generally small
• 2. Reflection/Transmission loss
• (By far the largest component of loss)
• Reflection/Transmission loss will be examined

5
HF Transformer Equivalent Circuit
At BPL frequencies the transformer presents a
highly frequency dependent mismatch to the
transmission line
6
Reflections/Attenuation due to Isolated Junctions
Power Line Tap
T 2/3 3.5 dB loss
Overhead/Underground Transition
Typical Loss 15 dB
7
Because the reflections/transmissions from
each junction or connected element interact with
each other, it is necessary to evaluate the
entire system in order to evaluate the system
attenuation
Z02
Z01
Consider the simple BPL system shown next
8
Attenuation at higher frequencies is very
frequency dependent and may exceed 30 dB/km A
very messy channel
9
Broadband Systems for Complex Channels
Modulation schemes (such as Orthogonal
Frequency Division Multiplexing - OFDM) have been
developed for these very complex channels.
These schemes spread the signal among
numerous carriers over a wide bandwidth. They
are designed so that carriers can either be
turned off or not available due to attenuation
without losing the signal. The cost of shutting
off carriers to protect licensed operations is a
sacrifice in channel capacity.
10
Emission Issues

• Enough power must be used to overcome the
• attenuation so that the signal to noise ratio is
• sufficiently high for satisfactory communication.
  
• But, larger power causes larger emissions and
• these must be controlled so that
• FCC Numerical limits are satisfied
• 2. No harmful interference is created.
• Balancing attenuation, input power and emissions
• (assuming Part 15 as is) results in repeater
• spacing of less than 1 km.

11
Emissions from Balanced Systems
Balanced systems (i.e., those with equal
and opposite currents on two conductors such
as most telephone circuits) emit relatively
small fields. This is especially true if the two
conductors are closely spaced.
12
On Meeting Numerical Emission Limits
For a two wire power line with 1 meter spacing,
the sinusoidal current required to produce a 50
dB ?V/m equivalent electric field (i.e., the FCC
limit in a CISPR QP Rx) at x 3 meters is 47
?A! If Z0 550?, then the maximum power flowing
(without violating the FCC limit) on the
transmission line is 1.21 ?W or 29 dBm! This is
not much! It does not take much power to violate
the FCC limits. Any unbalance in the currents
will increase the fields. Note Spreading the
signal over a wide bandwidth will reduce
the measured emissions since CISPR RX bandwidth
is only 9 kHz.
13
Systems that are likely to be Unbalanced

• Systems with more than one return path
• Systems with significant displacement currents
• (i.e., stray capacitance)

Fundamental Problem Since currents are
continuous, every current must return to its
source. The further this return current is from
its source current, the greater the emitted
fields.
14
1. Power lines are multiconductor transmission
lines as shown below. Current on one phase
conductor can return either on the other or on
the neutral wire. Unbalanced or common mode
currents that flow on the neutral conductor cause
greater emissions
15
Consider a Typical Excitation System
By expanding the voltages and currents into
common and differential modes, it can be shown
that both common and differential modes are
excited and that their amplitudes are roughly
equal. Tentative conclusion More complex
excitation schemes that excite only differential
modes appear to be warranted.
16
The reality is not so easy. Consider what
happens when currents are incident on an
unbalanced connected element.
A differential mode current will create a common
mode current and vice versa
17
2. Systems with significant displacement currents
Source Current
Open circuit current
18
At low frequency, the open circuit current is
nearly zero as expected. But not at higher
frequencies. Here, displacement currents become
important
19
Decay rate of BPL fields away from a power line
Why is this controversial?
FCC Section 15.31(f) says an attenuation factor
40log10(d2/d1) should be used to extrapolate BPL
field measurements made at d1 meters to a
location d2 meters from the power line. If the
field actually decays more slowly than this,
measurements made close to the line and
extrapolated to 30 meters (the standard
distance) will indicate smaller fields than the
actual fields measured. Measurements have been
reported that support a slower decay rate than
suggested by the FCC Do these measurements make
sense?
20
Decay Rate Calculations
simplified geometry, assume dpn gtgt dpp
General expression for the magnetic field along x
axis
I1 I2 I3 0
Messy, but simplified soon! Most important x,
ds compared to ? Note ? 150 meters at 2 MHz
and ? 10 meters at 30 MHz
21
Suppose x, (x-dpp), (xdpn) ltlt ? This is a
low frequency or near field
approximation Then the Hankel functions can be
approximated using small argument
approximations. Then
A lot simpler!
At large distances (i.e., x gtgt dpp , dpn) the
decay is as 1/x2
Id and Ic are the differential and common mode
amplitudes
22
If the common mode current is zero (i.e., Ic 0)
The field decays as 1/x2 as long as x gtgt dpp
(reasonable)
This is the decay rate assumed by the FCC in its
Part 15 Rules (i.e., 40 log10(d2/d1) or 40
dB/decade). But, the common mode is not always
zero, the frequency is not always low and near
field approximations may not always be allowable.
23
If the return currents are very far away (i.e.,
dnp gtgt x)
If, further, the differential mode amplitude is
0, then the fields decay as 1/x (i.e.,
20log10(d2/d1) or 20 dB/ decade). Since the
reality is probably in between these extreme
cases, it is no surprise that measurements
indicate decay rates between 20 and 40 dB per
decade.
24
At higher frequencies different approximations
can be made and
The decay rate in this case might be even smaller
than 20 dB/decade
The bottom line is that 40 log10(d2/d1) generally
does not properly describe the decay rate of BPL
fields.
25
Harmful Interference Clause
(b) Operation of an intentional, unintentional,
or incidental radiator is subject to the
conditions that no harmful interference is
caused. Harmful interference is defined as
any emission radiation or induction which
endangers the functioning of a radio navigation
service or other safety services or seriously
degrades, obstructs or repeatedly interrupts a
radio communication service operating in
accordance with this chapter.
Harmful interference is still not completely
defined!!!
26
Different World Views on Defining Harmful
Interference
1. Consider Broadcast TV for example For Grade
B, UHF television service (Channel 2 55 MHz) a
signal strength contour (i.e., 47 dB µV/m) is
defined that results in satisfactory reception
in 50 of the locations, 50 of the time.
Interference is deemed negligible if certain D/U
(desired to undesired signal) ratios defined by
the FCC are exceeded. Note There is no
mention here of the noise floor. Therefore, it
may be permissible for an interference signal to
exceed the noise floor. CISPR QP Receiver
120 kHz Bandwidth
27
2. Consider Amateur Radio This service is
frequency agile and operators often look
for parts of the frequency spectrum with low
noise (includes man-made interference).
Measurements at 28 MHz in a quiet area indicate
that a noise floor of -10 dBuV/m is not
uncommon. Note The noise floor defines an
acceptable interference level Therefore, it is
not acceptable for an interference signal to
exceed the noise floor. Position of the ARRL
not endorsed by the FCC CISPR QP Receiver
9 kHz Bandwidth
28
World View Comparison
Commercial TV (Ch 2 55 MHz) Amateur
Radio - 28 MHz
47 dBµV/m
Grade B signal level
BPL _at_ 30 m
D/U guard
29.5 dBµV/m
BPL with 20 dB filter
9.5 dBµV/m
Noise Floor
-10 dBµV/m
Acceptable Interference Ranges (Red)
BPL rejected by the ARRL since it exceeds the
noise floor
29
The Harmful Interference Wild Card
Is harmful interference any signal that
contributes to a measurable increase in the
noise floor or should it be defined as signals
that exceed a defined amount below a
specified protected signal level? At how far
from the source must the noise be deemed
acceptable (i.e. should points further than 10
meters be protected? 30 meters?) These are
questions that only the FCC can answer. And,
they will be answered only as the FCC responds to
complaints
30
Where are we now?

• BPL has not grown as quickly as its proponents
  thought. Some utilities have dropped BPL
  programs.
• I know of no one who claims to be making a
  profit on BPL.
• In some cases, there have been no interference
  complaints, but the business case still has not
  been satisfied.
• Independent economic analysis suggests a
  difficult business case.
• Utilities have many applications that are
  narrowband. It is possible that narrowband power
  line communication (i.e., the old PLC) could be
  used on medium voltage and low voltage systems