Assessment of commercial viability of a communication technology (or any technology for that matter) is only relevant in the context of its operating environment rather than the theory of communications in general. While the evolution of wireless communication for decades yielded significant characterization of the wireless medium resulting in huge advancements in wireless communications and communication technology in general, there was, in comparison, only small amount of characterization performed on the power line as a ubiquitous communication medium, and its specific challenges only now are becoming better understood.
The typical noise on the power line network is both time and frequency dependent. Some of the key characteristics of the power line environment, especially in the lower frequency region are:
• Impulse and tonal noise
• Significant and variable attenuation and propagation loss
• Often severe interference with time varying noise sources
• Dynamically changing channel due to load and noise variation
As one would expect, there are many sources of noise on the power line network. Some are due to the devices connecting to it and others due to the network itself, which in many cases is old and was never provisioned with communication in mind. Below are a few typical examples.
Activation of many kinds of devices can be a source of impulse noise. However, the most common impulse noise sources are light dimmers. These devices introduce high impulse noise, as they connect the lamp to the AC line part way through each half AC cycle. When the bulb is set to medium brightness impulses of several tens of volts are imposed onto the power lines at twice the AC line frequency.
Another form of noise is “tones”. The most common sources of tonal noise are switching power supplies, which are common in many electronic devices such as personal computers and electronic fluorescent ballasts. Many devices, such as televisions and computer monitors contain other high speed switching systems. The fundamental frequency of these systems is anywhere from 15 kHz to 1 MHz or more. The noise that these devices inject onto the power mains is typically rich in harmonics of the switching frequency. Figure 1 presents the spectrum of a typical DC charger injecting into the power main switching frequency harmonics at 70 kHz (main frequency), 140 kHz, 210 kHz, 280 kHz, etc.
Figure 1 – Typical DC charger spectrum
While typical channel simulations often rely on the principle of superposition, which inherently assumes linearity and time-invariance of the noise sources, neither of these assumptions holds for the power line environment, making theoretical analysis extremely difficult. As an example, the impedance at any point of the power line network varies with time as appliances are added, removed or change their power draw from the network. It is also not uncommon to observe different signal attenuation in different directions along the same path, i.e. signal attenuation from point A to B compared to the signal attenuation from point B to A.
Another challenge presented by the high variety of devices connected to the power line and the variation of load is the variance in line attenuation and its frequency dependence. Loads that present low network impedance at communication frequencies compared to the characteristic impedance of the wiring (e.g. heating elements), cause the wiring inductance, rather than its capacitance, to dominate the propagation effects of the communication signal.
Figure 2 presents an example of a real life power line channel that combines many of the noise sources discussed above.
Figure 2 - Unpredictable noise signature in narrowband channel (10-500KHz)
Click on the link below to check out the collection of Design Articles, Case Studies, Product How-To articles,... related to Power Line Communication that have been published on Smart Energy Designline.
Check back frequently. The list will be updated as new articles arrive.
It certainly is true that implementing a successful PLC system is a challenge, but the same can be said for wireless: meters are often installed in unfavourable locations like basements and in most of the world outside the US, houses are built using solid materials like stone or concrete which are far from transparant for radio waves. Using PLC is very attractive for grid operators as this is the only solution offering plug&play: you connect the meter to the LV-network and at the same time it can contact its concentrator. Noise and variable impedances at PLC-frequencies is a real and tough problem however and one can say that in the future one can only expect the number of noise sources in appliances to further increase (electronic ballasts for energy efficient lighting, Photo-voltaic convertors, electric cars, ...). We at Eandis have developed a solution for these problems: our patented technology uses filters to isolate the customers' LV-loads from the outside grid and we decided to use multiple gateways on each network segment to further increase reliability of the powerline communication. It's implemented in a test area with about 3000 customers and we are able to consistently read out over 95 % of the meters every hour and over 99% daily. In fact we had more issues with the firmware and software than with the PLC. By using multiple gateways we get a bandwidth of up to 9.6 kbps per 10 connected meters average, which we consider sufficient to be able to support IP over PLC in the future.
I am not ready to say narrowband PLC has out-lived its usefulness. The early implementations were really for fault detection in power transmission (though there were challenges to 'jump' across transformers) and the repair. From the utility provider's point of view, this was and still is a necessary piece of the communication needed in a system, even if it ends up being a backup/redundant.
Narrow band communication over power lines certainly would face a few challenges, but not the legal ones that broadband faced, caused by it's interference generation. The reason for seeking narrow band communication in some areas is because there is no internet there, so the assertion that internet is a better choice is sort of not the issue. It is possible that other methods might be better, but the questions about relative cost and reliability need to be answered in detail before that call can be made correctly. Glittering generalities have no place in the decision making process, only actual data.
PLC from a technology point of view isn't useless at all. It depends on the application but obviously there is not a single technology that fits all.
Even if get closer to standards on communications providing solutions for a wide range of scenarios, some special cases will always have their own rules.
Coming up with business models generating revenue with PLC isn't that easy but also a complete different story.
Research in PLC has brought up good results which diffused to other areas too. In some parts of the world PLC is used successfully. Beside that technology in general provides diversity in people's life and business. PLC is one of it.
The author sumarizes: "power line as communications medium presets unique challenges."
An understatement, to say the least.
The entire concept of PLC makes no sense, either from an engineering or business perspective. And I think you have to state your objective, in analyzing any proposal.
IF you want to enable remote or automatic meter reading, to save utilities labor expense-- forget about power comms--use available wideband--and either provide consumers a discount for automatic reading, or cut a deal with cable companies or cellphone carriers, for data backhaul from a smart meter.
IF you want to provide realtime command and control of the national grid, to provide smart power routing, and source switching, to match peak demand and alternative source variabiity, you need an agile wideband system which is separate from the network you want to control. And we've already got that over 95% of the US, with the internet.
From an investment standpoint, as well as a communications engineering perspective, PLC is a technology which offers nothing.
The only reason it's still around is the relative expense of wideband (way down from where it was 20 years ago, and falling) versus sunk cost in the existing utilities outside plant.
Overlay divested power generation and distribution companies on that backdrop, and it makes even less sense.
It's time for a national energy vision, and a policy for implementation. We're wasting time on PLC.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.