An experimental investigation into the electromagnetic compatibility aspects of high frequency power line communications
Power line communications technology, long established for low data rate
applications, is now charting new territory with respect to data rates and provided
services. This can only be achieved by increasing PLC operating frequencies from
the low frequency band (below 148.5 kHz) to the high frequency band (1 MHz and
upwards). There is now only one technical barrier to widespread deployment -
Existing low voltage power networks are optimised for the safe supply of electrical
energy. Low voltage cables are often pseudo co-axial in their cross section, but
when high frequency signals are coupled onto the network, part of the signal will be
radiated. There is therefore a potential for interference to be caused to legitimate
users of the radio spectrum.
This thesis, and the experimental program underlying it, seeks to quantify potential
problems and to propose mechanisms by which they could be mitigated to the extent
that wide scale deployment of PLC networks becomes possible.
The first part of the thesis offers a detailed introduction to the topics of electricity
supply networks, power line communications, modulation techniques and
electromagnetic compatibility. Existing EMC standards are examined and although
some do not directly cover power line communications networks, key principals are
drawn for later use in standards development.
The thesis then seeks to examine the mechanisms by which high frequency
interference might be caused. Radio propagation modes are discussed and a clear
technical distinction is drawn between localised interference from a single PLC
network to an individual radio user, and cumulative interference from wide spread
deployment of PLC systems. Both such scenarios are examined in detail.
The experimental program IS described quantifying radiated signal strength
regression from a number of power networks and at a number of operating
frequencies within the high frequency band. In this context, signal strength
regression is the rate at which electrical field strength reduces with increasing
The experimental setup uses a conventional signal generator to supply single test
frequencies of known power spectral density, which are coupled onto a power
network. The subsequent radiated signal is received via a conventional antenna and
radio receiver at a number of locations surrounding the power network at known
distances, and signal regression is derived. The experiment was repeated for a
number of different frequencies and at representative urban, suburban and rural
locations. Indeed, the experimental technique was evolved over a number of months
to allow increased portability of the signal receiving equipment, and hence the
number of measurements that could be taken.
From the experimental results, presented both In tabular and graphical format, a
number of conclusions can be drawn.
Firstly, based on these results, antenna factors in the order of 85 dB/m can be
expected of power line communication networks. It can be concluded that the field
strength regression to be anticipated from PLC networks is likely to be significantly
below the -20 dB per decade 'free space' regression figure that has often been used
in interference models. In fact a regression figure of -35 dB/decade IS more
representative of ground wave propagated interference from PLC networks.
It is also possible to conclude that the adoption of orthogonal frequency division
multiplexing as a multi-carrier spectral technique offers specific advantages in EMC
terms. Due to its nature, it is possible to apply a frequency 'mask' to an OFDM
based PLC system. Such a mask might be static, applied on a national or regional
basis in order to guarantee non-interference with specific frequencies, for example
those used for emergency radio channels. It would also be possible to add a
dynamic frequency mask, controllable on each PLC system, to mitigate interference
with radio services operating within the PLC operating band.