For most of a decade, the FCC tried to turn the electric grid into a broadband network — and in doing so, turned utility poles across America into HF antennas. Amateur radio fought it, documented it, took it to federal court, and won. Then the balance sheet finished the job the lawsuit started. BPL didn't lose on interference. It lost on arithmetic.
In the early 2000s, American broadband was a duopoly. You got cable, or you got DSL, and in a lot of places you got neither. The Federal Communications Commission wanted a third pipe — and someone pointed at the one utility that already ran a wire to every building in the country.
The electric company.
Broadband over Power Lines — BPL — proposed to inject data onto the medium-voltage distribution lines that run down every street on the poles overhead, and pull it back off at the transformer serving your house. No trenching. No new poles. No spectrum auction. The grid was already there, and it went everywhere. On paper it was the most elegant broadband idea anyone had in a decade.
It did not work. And the reason it did not work is a lesson in RF engineering that the amateur radio service spent eight years explaining to a federal agency that did not want to hear it.
Here is the whole problem in one sentence: an overhead medium-voltage distribution line is an unshielded, unbalanced conductor, strung dozens of feet in the air, above a ground plane, with no termination.
At 60 Hz, that description is irrelevant. The wavelength is thousands of miles. Nothing about the geometry matters.
At 10 MHz, the wavelength is about 30 meters. Now that same conductor is many wavelengths long, it sits a fraction of a wavelength above ground, and it has discontinuities — transformers, taps, splices, dead ends — every few spans. It is not a transmission line at that frequency. It is a long-wire antenna with a random impedance profile, and every discontinuity is a place where energy that was supposed to keep traveling down the wire radiates into the sky instead.
BPL took a wideband digital signal — typically OFDM — and put it on that wire. Not at one frequency, but across a swath of the high-frequency spectrum. And HF is not just any spectrum. It is the band that propagates by ionospheric refraction, which means an HF receiver in a quiet location is listening to the entire planet, and its usable sensitivity is set not by its own noise figure but by the noise floor around it.
Raise the noise floor in a neighborhood and you have not degraded HF reception there. You have deleted it.
The interference ran both ways, too. A power line radiating into HF is also a power line receiving HF. A nearby amateur station running legal limit into a decent antenna could desensitize or overload the BPL equipment on the pole. The technology was fragile in exactly the environment it was being deployed into.
The Commission opened a Notice of Inquiry in April 2003 (ET Docket 03-104), issued a Notice of Proposed Rulemaking in February 2004 (ET Docket 04-37), and adopted the BPL Report and Order — FCC 04-245 — on October 14, 2004. It was released October 28 and took effect February 7, 2005.
The rules became Subpart G of Part 15, sections 15.601 through 15.615. The definition is worth reading closely.
47 CFR § 15.3(ff)
Access BPL is defined as a carrier current system installed and operated on an electric utility service as an unintentional radiator, sending radio frequency energy on frequencies between 1.705 MHz and 80 MHz over medium voltage or low voltage lines, located on the supply side of the utility's interconnection with customer premises.
Read that again. Unintentional radiator. The Commission authorized a system whose signal, by physical necessity, radiated across the entire high-frequency spectrum — and classified it as a device that isn't supposed to radiate at all. Part 15, the rule part governing unlicensed devices that must accept interference and must not cause it, was made to carry a technology whose fundamental operating principle was pumping RF onto a mile-long antenna.
The Order did impose conditions. Access BPL systems had to be capable of frequency notching — at least 25 dB of attenuation below the applicable Part 15 limits below 30 MHz, and 10 dB above it. There were excluded bands and geographic exclusion zones: no operation on 2173.5–2190.5 kHz within a kilometer of listed coast station facilities, no use of 73.0–74.6 MHz within 47 km of the Very Large Array. Operators had to consult with public safety licensees in designated areas, list their deployments in a public database thirty days before turning on, and build in a remote shutdown capability.
On paper, a reasonable framework. In practice, the notching capability existed and the interference happened anyway — because the systems were radiating far above the limits to begin with.
ARRL Laboratory Manager Ed Hare, W1RFI, spent the better part of a decade in the field with test equipment, standing under BPL-energized power lines in Virginia, Pennsylvania, Indiana, and elsewhere. His summary, delivered at the end of the fight, was that in system after system the ARRL measured, Access BPL was "operating at levels from 15 to 40 dB greater than the FCC limits" — and still not working well.
That is the sentence that should be carved over the door. Fifteen to forty decibels over the legal limit, and still not delivering usable broadband. The systems could not simultaneously be legal and functional. The physics did not permit it. To get enough signal down a lossy, discontinuous, badly matched conductor to move data, you had to radiate. Radiating was the failure mode and the operating principle at the same time.
Hare's diagnosis was blunt: distribution power lines are simply not designed to carry broadband signals. Systems can be made to work in that hostile environment under ideal conditions. Conditions are never ideal.
Manassas, Virginia — thirty miles from Washington, in the shadow of two Civil War battlefields — became the poster child.
The city ran a field trial in mid-2002. In October 2003 it granted a ten-year franchise to expand the trial citywide. That contract was terminated in April 2004 and rebid, and in October 2005 COMTek announced the first citywide BPL service in the United States. Manassas was, for a few years, the most-cited BPL deployment in the country. FCC Chairman Michael Powell visited the site; the ARRL maintained that the visit, on the eve of the Commission's vote, was made in violation of the FCC's own rules.
Manassas hams did what hams do. George Tarnovsky and the Ole Virginia Hams documented the interference relentlessly — field monitoring, complaint letters, meetings, years of it. One operator reported to the Enforcement Bureau that he could not copy a special event station while driving through the city, with noise running from 10 dB to more than 40 dB above the background. In 2005 the ARRL formally called on the FCC to shut the system down. In 2006 the Commission tested six locations in the city and dismissed the complaints; Tarnovsky pointed out that the hams who filed those complaints were never contacted to be present for the tests.
ARRL CEO David Sumner, K1ZZ, named the fight the Third Battle of Bull Run.
It was not decided on interference. It was decided on a spreadsheet.
Subscribership never got where it needed to go. Across the system's life the customer count ran somewhere in the range of five hundred to eight hundred and fifty accounts — in a city of thirty-five thousand — and the trend was downward. COMTek got out in 2008 and the city took over the network. By 2010 Manassas was losing on the order of $166,000 a year on it. The service ran under 1 Mbps for about twenty-five dollars a month, in a city where Comcast and Verizon both already offered service and Verizon was rolling out FiOS.
On April 5, 2010, the Manassas City Council voted unanimously to shut it down effective July 1. Councilman Jonathan Way had put it plainly the year before: the city needed to get out of BPL forthwith, it was not a good product, and the business had never been financially sound.
Sumner's verdict was drier and more damning than any interference report — Manassas had demonstrated that there was no business case for BPL as a consumer broadband medium.
While Manassas was bleeding money, the ARRL was in court.
The League filed its petition for review in the U.S. Court of Appeals for the District of Columbia Circuit on October 10, 2006. Oral argument was October 23, 2007. The opinion came down April 25, 2008 — American Radio Relay League, Inc. v. Federal Communications Commission, 524 F.3d 227 — and the mandate issued June 13.
Judge Rogers, writing for a panel of Rogers, Tatel, and Kavanaugh, found for the League on two grounds.
First, the redactions. The FCC had relied on five internal staff studies in adopting the BPL rules, and released them to the ARRL — under the Freedom of Information Act, after two requests — with portions blacked out. Among the withheld material were sections headed "New Information Arguing for Caution on HF BPL" and "BPL Spectrum Tradeoffs." The Commission had, in other words, redacted the parts of its own engineering that cut against its conclusion. The Court held this violated the notice-and-comment requirements of the Administrative Procedure Act, and said no precedent sanctions such a "hide and seek" application of the APA. An agency may not cherry-pick a study it has chosen to rely on in part.
Second, the extrapolation factor. This is the technical crux, and it deserves an explanation.
You cannot measure BPL emissions at every distance. You measure at one distance and extrapolate, using an assumed rate at which field strength falls off with range. The FCC used 40 dB per decade — meaning the signal drops 40 dB for every tenfold increase in distance. The ARRL's analysis, supported by studies from Ofcom, the FCC's British counterpart, put the real-world figure closer to 20 dB per decade.
The difference is enormous. A slower rolloff means the interference reaches much farther than the Commission's math predicted.
The part that should stop you
The FCC's own Laboratory had recommended a 20 dB per decade factor. The Commission overrode its own engineers, and then — in the Court's words — summarily dismissed the contrary data in a manner that could not substitute for a reasoned explanation. The Court further noted that an NTIA study already in the record cast doubt on the Commission's decision.
The panel remanded. The FCC was ordered to put the unredacted studies into the record, take comment on them, and either justify 40 dB per decade or adopt another factor and explain itself.
Judge Kavanaugh dissented in part. He believed the Commission had adequately explained its reasoning on the extrapolation factor, and joined the rest of the opinion.
Ten months later, ARRL General Counsel Chris Imlay, W3KD, wrote to the Commission to observe that literally nothing had been done to comply. The FCC eventually issued a Further Notice in July 2009 and a second Report and Order in October 2011 — by which time the technology it governed was essentially extinct.
It is tempting for hams to tell this story as a victory. It is more honest, and more interesting, to admit that the lawsuit did not kill BPL.
Arithmetic killed BPL.
The cost per home passed never came down. Every distribution transformer is effectively a low-pass filter to an HF signal, so every transformer needed a coupler or a bypass — a piece of hardware on a pole, installed by a bucket truck. Throughput was shared and slow: Manassas delivered under a megabit while cable was moving to DOCSIS 3.0 and Verizon was pulling fiber to the house. And utilities, for the most part, did not want to be internet service providers, and had no appetite for a fight with the telcos and cable operators whose equipment already hung on their poles.
Meanwhile the rural argument — the last redoubt, the case for BPL as the answer where nothing else reached — got taken away by everyone else. Federal broadband stimulus money went to fiber, not BPL. Wireless ISPs were already serving those communities. Cellular data got good.
IBEC — International Broadband Electric Communications — was the last serious American operator: an ambitious rural play headquartered in Huntsville, Alabama, with a $9.6 million agreement with IBM to build BPL on electric cooperative lines across seven states, aimed at a couple hundred thousand rural homes. It closed its doors in January 2012, citing financial damage from the April 27, 2011 tornado outbreak that tore through its Alabama service territory. Outside observers were less charitable and probably more accurate: the tornadoes were the proximate cause, but the business model had been failing for years, and investors and lenders had stopped believing.
When IBEC went dark, Ed Hare noted the practical consequence for amateurs — no remaining BPL system in the United States used the amateur bands.
The fight was over because there was nothing left to fight.
A noncommercial hobby reference compiled by N6JET, gathered from public sources and shared freely for anyone interested in amateur radio history.