Broadband over powerline (BPL) is always next year's technology; now it's never. Is never soon enough for you? For about the last 13 years, BPL was the going to be the third pipe into the home, supplementing the two incumbent wireline offerings of DSL and cable, which had developed into monopoly or duopoly controls most places in the world. Two years ago, with favorable FCC and upcoming EC decisions on BPL either released or about to happen, BPL seemed about to come into its own. I wrote a positive piece for The Economist based in large part on an enormous deployment that was contracted and underway in Texas, and a contract that had just been signed in France. These two events seemed like they would catalyze BPL.
About 18 months later, the Current Communicatins and TXU (now Oncor) Electric Delivery deal, which was expected to pass 2m homes by the end of 2008, is over, with Oncor purchasing the telecommunications network for $90m a few days ago. Oncor will use just the smart grid features that allow dramatically improved network monitoring--which is a well-understood aspect of data over powerlines, dating to much slower and primitive networks. The Dallas Morning News reports that just 64,000 homes were wired for BPL so far, and that Oncor will not offer Internet access. Oncor had agreed in 2006 to pay $150m for smart-grid features.
Google was a Current investor, which gave more credence to their plans in 2006. The company had already rolled out some smaller markets, overcome equipment problems, and had a positive relationship with the ARRL, the amateur radio society, in resolving interference issues. Hams have been the biggest complaintants with the FCC over BPL because hams are primary and secondary licensed users in the bands they use, while BPL is an unlicensed use.
The French deployment by SIPPEREC, a utility that manages power for the suburbs of Paris, stated that 1.5m homes would eventually be passed with BPL service, but no information has been released since Feb. 2007 about the project, which makes it likely that it simply didn't happen.
Even when I was researching the Economist piece, I was troubled by the many European deployments that were announced, went into trials, and then disappeared without a trace. Still, there were some active projects in Spain, Switzerland, and Ireland, and the rollouts in France and Texas seemed both committed (contracts were signed) and imminent. But the laws of physics always win, and I can only think that BPL equipment from whatever vendor simply cannot deliver results that work within budget and reliably enough to make network deployment for broadband make any sense.
The FCC's 2006 order that overruled a number of ARRL objectives stated, essentially, that interference was okay even with licensed purposes as long as it was within tightly controlled parameters. Part of the "BPL is dead" argument I make today stems from an appeals court decision in late April which affirms the FCC licensed/unlicensed approach, but which requires the agency to re-evaluate its information about interference. The FCC failed to disclose fully information from studies it relied on in setting rules, which violated public process. The ARRL wrote up the appeals decision on their site, and notes that a study in the UK that was fully released showed a much lower threshold would be needed.
The agency's need to redo some of its work, a potential shift of power to Democrats on the commission starting 20-Jan-2009, and the fact that other work shows the rules were established incorrectly could result in restrictions on BPL that make it even less likely to be rolled out. [Initial links via DSL Reports]