November Six Oscar Lima

Author: n6ol (Page 1 of 2)

Finally, a Real Update from PG&E

Though there’s still no damned excuse for PG&E taking more than ten months to send someone out to investigate the noise source at the PG&E Poplar substation near my QTH, I did get a call last week after my second letter to PG&E’s CEO and my case being escalated to a supervisor in the customer call center, who then raised the issue with a handful of other supervisors.  It shouldn’t be this hard to get someone to react at a public utility, but at least they finally did.

It turned out there had been some activity at the substation in an effort to resolve the problem between March 31 when the source of the trouble was confirmed, and when I finally got a call from someone who knew what was going on.  Evidently PG&E has a fairly terrible communications problem when it comes to information moving between the people who actually work on the issue and the customer-facing call center and ticket tracking system.

The trouble has been isolated to the vicinity of a 230kV bus at the substation, though in the intervening weeks they’ve also power washed and cleaned up some 115kV systems.  Evidently the trouble is that they cannot simply power-wash around the 230kV bus; to do so would risk a catastrophic arc flashover because of the sheer potential difference involved.  (230kV is a LOT of electrical potential.)  From the explanation I was given, it sounds like PG&E does not have a redundant system for this 230kV bus either.  They can’t work on it without shutting it down, and they also can’t just shut it down to work on it– to do so would shut down power at the substation for everyone it serves.

Given the lack of redundancy, I asked how they do routine maintenance on this system just in general, and the answer could be summarized as “rarely, and at considerable expense.”  The systems themselves are designed to have a very long MTBF, but on the occasions that they do need to shut them down, they have to plan it out months in advance, bring in a mobile unit of some kind (presumably either power generation or a busbar on a flatbed maybe?) to keep things working while the 230kV system is shut down for maintenance.  This apparently happens only once every 4-5 years and can be a multi-million-dollar operation.

That said, all hope is not lost: they’re trying to book their infrared team to come out and image the system while it is under load to see if they can spot any “hot spots” which might be addressable without something as inelegant as a power-wash, and they’re also looking to work with their high-tech services team to see if there are other ways they might suppress the RFI and keep it from being a nuisance in our neighborhood.

But lets be very clear about one thing: what PG&E is doing at the substation, though it may be very inconvenient to fix, is still a violation of FCC rules.  And it doesn’t matter if it inconveniences a handful of Ham Radio operators, shortwave listeners, AM broadcast enthusiasts, or a thousand people’s cell phones; the rules are the rules and they don’t distinguish by how many people are affected or how troublesome it may be to correct the root cause.  (That said, I do believe I’ve observed a correlation between noise emanating from the substation and reduced signal quality of KRCB-DT on ATSC channel 5, but I need more data to confirm it.)

The PG&E supervisor promised to keep me better informed going forward as they continue to work on the problem, which is a very different attitude than what I got from the rest of the company between May 16, 2024 and May 16, 2025 (it was in fact one full year to the day between my opening of a case about the substation and receiving a call from a supervisor on Friday).

That’s what I know for now.  More to come, I hope.

Real-World Example of PG&E / Electrical Utility RFI

In case you ever wondered what PG&E’s (or really an utility company’s) electrical arc noise sounds like on an actual radio, here’s a short video in which you can hear how that awful 120 pulse-per-second racket actually manifests in a real-world scenario:

I start out tuned to 3.560 MHz, which happens to be about where my 80/40/30/20m fan dipole antenna is resonant on 80m.  I’m using AM mode at the request of the folks at the ARRL for the purpose of documentation. Though this would normally be a lower-sideband segment of the band, it helps to hear the noise over the background static to decode as AM.

Toward the end of the video, I tune to WWV on 15 MHz, which is a bit above where my 20m antenna is resonant, so the reception isn’t quite as good.  When PG&E is not partying like it’s 1899 with their unintentional spark-gap transmitter, WWV is still clearly audible…  not so much when they are producing this RFI though.

It’s worth noting that this broadband noise doesn’t affect only the amateur radio spectrum.  I’ve charted it as low as 1600 kHz (AM Broadcast band) and all the way up to 450 MHz (UHF) and everywhere in between.  There’s a reason spark-gap transmitters are forbidden by international treaty, and this is it: electrical arcs produce tremendously wide signals and horrific broadband noise.  This noise affects AM and FM commercial broadcasts, Shortwave radio, government and public safety, VHF and UHF television, and even cellular phones.

My Letter to the FCC Regarding ATSC 3.0 and the Sunsetting of ATSC 1.0

It’s far from the most elegant thing I’ve ever written, but I think it gets my main points across reasonably well.  The due date for public comment is coming up on May 7!  If you want to leave a public comment:

While I sympathize with content providers’ desire to mitigate copyright infringement, I strongly disagree with their behavior thusfar when it comes to DRM-encrypted ATSC 3.0 broadcasts. The A3SA and NAB appear to be taking advantage of this crossroads in technology to reverse the status quo which has been in place since the advent of the VCR and the decision in Sony Corp. of America v. Universal City Studios, Inc., 464 U.S. 417 (1984), and in the process they’re reducing the utility of broadcast television as a public service over the public airwaves.

As it stands today, despite many channels in the San Francisco Bay Area broadcasting from the ATSC 3.0 “lighthouse” on ATSC channel 7, I cannot tune most of these channels at all– DRM has been taken to its logical conclusion and the content cannot be received, despite my receiver being ATSC 3.0 capable. The reason? I, like countless other consumers, use a home-networked tuner, and the A3SA has failed to work with manufacturers of network tuners to ensure that they can receive their DRM-protected content.

So not only can I not record broadcasts for time-shifted viewing, like I’ve been able to do since the 1980’s, I can’t even watch them live via the more-reliable ATSC 3.0 signal.

The courts have already ruled that time shifting for one’s personal viewing qualifies as “fair use” and now the NAB & A3SA are trying to use new technologies to eliminate fair use for their own benefit with no regard for the viewing public, who own the airwaves.

Many claims have been made about the necessity of DRM-protecting these signals, and yet, the recording of live television, both on analog and digital media, has been a common practice for more than 40 years. What’s changed is an opportunity arising for content providers to wrestle back control from the viewing public. But in placing so many restrictions on protected content, they’ve rendered their own broadcasts unviewable by a large percentage of the public.

I have additional concerns about their plans to disseminate encryption keys over the Internet. There are many reasonable privacy concerns for why consumers may not wish to have their televisions or tuners connected to the Internet 24×7 or even temporarily to retrieve an encryption key (which reveals the intent of the requester to watch a particular stream); moreover, in the face of an Internet outage, this could render it impossible for consumers to watch television at all. In an emergency situation, where broadcast television might be the only way for the public to be informed, for example in the face of a natural disaster, terrorist attack, or regional power disruption, losing the ability to decrypt an encrypted signal could expose consumers to unnecessary risk of harm or loss of life.

Lastly, on a technical note, the methodology for streaming ATSC 3.0 video causes it to take several seconds to either tune a channel or to recover in the event of transient interference. The strange choice of ISOBMFF fMP4, a reliable delivery and storage format, over MPEG-TS, an unreliable delivery streaming format, means that synchronization can take several seconds in the event of even transient interference and upon every channel change, leading to a very poor user experience, especially for those living in areas of challenging reception conditions (as we do, living less than 5 miles from a major airport).

These decisions taken in aggregate give the impression that A3SA and NAB are entirely concerned about protecting their own interests and not at all concerned about serving the public over the public airwaves. To sunset ATSC 1.0 without addressing issues which affect the viewing public in ways unseen before in broadcast history would do them a great disservice.

Powerline Noise Ever Worsening

Not only has PG&E failed to address the noise emanating from their substation, the overall problem has now gotten worse, as a new noise source has appeared.

I noticed in checking the charts yesterday that the noise floor has jumped from the typical -70dB (or so) on 80 meters to a much higher level which precisely tracks the noise source that I had previously tracked to the Poplar substation:

My means of calculating the noise floor is admittedly a little naïve: I use the same algorithm I use to findthe noise peaks but look for the worst-fit level rather than the best-fit of the pulse noise.  This generally works very well when there’s a single noise source, or multiple noise sources which are in-phase.  But as we know, power utilities handle electricity in multiple phases, commonly three of them at utility scale, spaced 120 degrees apart.  In the time domain, at 8000 samples per second, that puts the peaks about 1333 samples apart (because there are two peaks per cycle), and depending on how badly a particular piece of equipment is arcing, each peak can in turn last for hundreds of samples.  With enough noise generation, the entire time domain can be saturated with noise, which is what appears to be happening now.

I have, of course, already tracked down the source of the noise.  It’s a pole about 530 feet from my house.  It may have a small vegetation problem:

My usual method of noise hunting worked nicely here.  I got my 220 MHz, 3-element Yagi and my handheld radio in AM mode tuned to match the Yagi, started at my house, and aimed for the direction the signal seemed to be coming from.  Then I walked a short distance away and aimed again.  It seemed to be coming from behind a row of houses.  So I took a walk, stopping every so-often to take another reading.  This led me to a nearby park.  I then kept going to the other side of the park to make sure the signal now appeared to be moving “backward” (ie, that I had walked past its location) and it did.  From here I could see the probable source: a lone utility pole in a marshy area between the park and the rear of a row of houses.  I then got as close as I could to the pole, which was making a pretty obvious racket, but confirmed it anyway by rotating the Yagi 45° to put the null in the cardioid pattern in line with the probable source to see that the noise dropped off.   It did.

I have of course reported this new noise source to PG&E with the pole number and coordinates and a note that there appears to be a good bit of vegetation growing up around the pole (despite the photo, though, as best I could tell there wasn’t any tree parts actually touching the high-voltage sections of the pole; however it is behind a fence in a protected wetland area, so I was unable to confirm this for certain).

Perhaps they’ll ignore this arcing pole for less than a year before they do something about it?

PG&E almost did something.

The stupid comedy with PG&E continues.

After my polite-but-assertive letter to PG&E’s CEO, I did hear from a telecom tech at PG&E who was ready and willing to help, but that was the last I heard of anything actually being done.  Unfortunately we had a few days back-to-back during which time whatever equipment they have causing interference was silent.  It then kicked up again as it usually does, but by then it seems that PG&E lost interest.

That was two months ago.

A month ago I called for an update, and as there were no notes about the activity after my letter, the representative opted to open a new case with a promise that I’d get a call back in a day or two.

Eight days later, I received a call from a blocked number from a maintenance supervisor who left me a voicemail with no callback number.  I tried to return the call via the main PG&E number (the only number I have) only to learn that the case opened by the representative had simply been closed after this supervisor made that one impotent attempt to reach me.  This exchange resulted in yet another new case being opened, this time with an admonition to stop closing the case without resolving the problem.

It has now been 390 days since my initial case was opened with PG&E.  They could have fixed this problem 100 times by now.

My Letter to PG&E’s CEO

It’s now been 10 months since I originally opened a ticket with PG&E regarding their RFI in my neighborhood, and though you’d never guess it by looking at Southern California right now, wildfire season is no longer a viable excuse for their inaction.  Moreover, this winter storm season has been tragically very mild, with almost no rain to speak of in January and none on the horizon, so winter storms are not an excuse for inaction either.

My contact within PG&E is ghosting me.  I haven’t heard back from the FCC or the ARRL in months either.  So I decided to drop a letter in the postal mail to PG&E’s CEO outlining succinctly what the problem is, why it’s a problem, and what they really need to do about it.  Here it is for your reading pleasure:

Ms Poppe:

Since March 12, 2024, I have been trying to get PG&E to resolve Radio Frequency Interference (RFI, also known as RTVI) that PG&E equipment is producing in the vicinity of my Amateur Radio station, beginning with the still-open case #5——–0.

PG&E does not appear to have a process for resolving this kind of problem. For example, it took until May 14, 2024 to resolve the closest source of interference to my station after the case was first ignored completely until May 9. Unfortunately, once that interference source was cleared, it came to light that there are several additional sources of interference nearby.

I have, on my own time and at my own expense, located these interference sources myself with my own equipment, and I’ve written monitoring software so your technicians can easily tell when the intermittent interference is active.  Sadly, despite my efforts, nothing has been done.

PG&E is obligated to resolve RFI issues arising from its equipment under FCC rules under Title 47, CFR sections 15.5, 15.13, and 15.15.

Additionally, it is in PG&E’s best interest to locate and resolve these problems, as they can be an early indicator of an unsafe condition, such as current arcing across an insulator.

Current that is producing RFI is current that is not being delivered to customers and current that could be generating heat or sparks and thus may be a potential fire risk. Ignoring RFI is unwise.

I would like PG&E to meet its obligations both to safety and under federal law and fix these interference sources.  Further, I strongly suggest that PG&E develop and implement consistent processes for this type of problem to ensure straightforward, timely resolution in the future.

I don’t think I’m being unreasonable, and I think 10 months is plenty of time to track down a handful of RFI sources and get them fixed, especially as I have already done all the legwork of figuring out (mostly) exactly where they are.

PG&E Serendipity

A little over two weeks ago, I fractured my humerus in a cycling accident, and since that time I’ve been doing a lot of walking to try to minimize how much weight I put on while I’m stuck in a sling and unable to run, bike, or swim.  So it happened on Tuesday after I got my sutures removed downtown that I decided to walk home (about 1.5 miles).

As I was nearing home, I heard a familiar sound: the sound of arcing 60Hz power.  I looked around, and sure enough, I was standing underneath a PG&E power pole.  It was arcing so loudly I could hear it over the traffic of the nearby 101 freeway.

Because arcing is a dangerous condition, I noted the address and pole number, and called it in to PG&E’s emergency line.  They promised to send someone out to check on it.  If indeed they did send anyone out, they did nothing.  I went back over to the same pole yesterday, this time armed with my handheld Yagi and my HT, and confirmed both that the pole is still arcing, and that the arcing is definitely causing RFI.

I called it in to my PG&E “customer relations” specialist, who told me that the department responsible for investigating this kind of thing has been working on changing their processes so they don’t just ignore reports of arcing and RFI for months, but rather do something about them instead.  We’ll see.  I expect this pole to still be arcing two months from now, as is their usual custom.

It has now been 142 days since my initial reports of really bad RFI in my neighborhood to PG&E, and though they have fixed a few of the problematic poles, they still haven’t fixed them all.  When you hear PG&E talking a big game about all the safety inspections they’re performing, know that it’s performative BS.  If they really cared about powerline safety, they wouldn’t wait two months to investigate any report of RFI or power pole arcing.

Some PG&E Progress, though not enough yet

Three days after they decided to send someone out again (finally), I got a call from PG&E’s RTVI hunter guy that he’d found a couple spots in the neighborhood, but unfortunately he called while I was in a meeting, so he got my voicemail.  He was calling to check to see if the locations he found aligned with where I had picked up the noise myself, and strangely enough, no, they did not!

What this means is he found two more locations that I didn’t even know about.

The following night the noise was unusually bad and choppy, so I decided to go out for a walk with my Yagi and my HT; this time I found yet another location, unfortunately behind someone’s house, but where the same type of sagging insulator was used that has been implicated in several other noisy poles.

And tonight, I decided to see if I could track down what was making the occasional snapping, crackling, or popping sound, and no it wasn’t a bowl of Rice Crispies.  This turned out to be another location entirely, and what’s particularly disturbing about this location is that it’s off the levee and completely surrounded by dry vegetation and people’s homes.  And once again, the same type of sagging insulators are implicated, this time on two poles which are connected.

Both of these two new locations only cause trouble when the wind is blowing; there seems to be some kind of fault that is triggered by the insulators moving slightly in the wind, which I could see from the ground.  I have of course relayed all of this information to PG&E.  Maybe some day they’ll act on it.

This brings the grand total of problematic utility poles in my neighborhood up to 6 now, including the original one that was close by, and that’s just the ones we know about.Two utility poles situated over a great deal of dry vegetation

PG&E’s Resistance to Fixing QRM is Getting Old

It should come as no surprise to anyone that after insisting I open a new case for the secondary source of QRM from PG&E, nothing further has been done.  I opened that case, as requested, back on May 16, and the QRM continues to follow a fairly predictable pattern, so much so that I can say with pretty high confidence that someone will hear it around 8 AM, 2 PM, and 8 PM most every day, more when it’s hot and dry.

A few things have happened though.  Last week PG&E finally responded to the CPUC regarding my original complaint that they were dragging their heels and not doing their job to ensure their equipment was maintained properly and safely.  I know it’ll come as a tremendous shock that PG&E misrepresented or omitted key facts in their response to the CPUC.  Naturally I couldn’t allow a factual misrepresentation in their response stand, so I took advantage of the CPUC’s “one and only one opportunity” (how they put it) to set the record straight.

(You read that right: the CPUC will only allow you to respond once after the utility sends their response, and they warn you that unless you provide them with some new and compelling evidence, they’ll kick you to the curb.  They’re not even vaguely nice about it.)

I corrected the record thusly:

  • My initial trouble report was made to PG&E on March 12.
  • PG&E assigned someone to work on this case only AFTER my informal complaint to the CPUC was forwarded to them on April 12. Until that point, they had completely ignored my trouble report.  My first contact from PG&E regarding my case was on April 16, and it still took them almost 30 days, until May 9, to assign someone to look into a safety issue that could have started a fire.  “Complain to the CPUC just to get PG&E’s attention” is not a scaleable process.
  • Upon assigning telecom techs to the issue, PG&E found exactly what I said they would find, exactly where I said they would find it: high-voltage power was arcing across a failing insulator a short distance from our house.
  • The problem was severe enough that their techs told me they could actually hear the arcing when standing below the pole. This is rather different from PG&E’s initial characterization that there was nothing wrong.
  • On May 14, PG&E replaced failing insulators on the pole in question.  There is no dispute that PG&E equipment WAS at faultIt took them more than 60 days from my initial problem report to get to this point.
  • However, a problem is still unresolved and threatens safety: there remains a secondary source of interference of exactly the same type, apparently originating within the Poplar substation (based on radio direction-finding I’ve undertaken on my own time and at my own expense).
  • PG&E insisted that I open a NEW case for the secondary source of interference, which persists in an unresolved state to this day.  The new case number is [redacted], opened on May 16.  It has not been addressed at all.
  • Given the interference is of exactly the same type, it is reasonable to suspect the cause is likely similar: high voltage electrical arcing, the type which can start fires, and which has started fires in the past.
  • There remains a threat to public safety any time PG&E’s infrastructure is experiencing electrical arcing.  PG&E should not be ignoring these early warning signs.

The CPUC took about a week to respond, and when they finally did, they seemed to accept my “new evidence” and forwarded it on to PG&E.

The very next day someone from PG&E actually showed up at my door to ask about the issue; I explained that it’s been an ongoing problem for well over 100 days now (112 days when he turned up), that it’s a 120 pulse-per-second interference that appears to be originating at the nearby substation, and that it is possible to discern it now that the nearby pole has been fixed.  He indicated that he’d arrange to have the telecom guys come out again and hunt for the source of interference.

We’re now up to 117 days since my original case was opened, and 52 days since my most recent case was opened regarding the substation QRM source.

To an outside observer, it appears that PG&E’s policy is to ignore these problems entirely for at least 2 months before taking any action, and I have no idea if they’d have ever taken action at all if I hadn’t opened a complaint with the CPUC.  Even after a CPUC complaint, it can still take 3 weeks before someone is finally scheduled to come out and hunt for the arcing.  These data fly in the face of PG&E’s ongoing PR effort in which they want to appear as though they place a high priority on safety.

If they really cared about safety, and someone was telling them they had equipment that was arcing near a field full of dry grass, you’d sure think they’d put a priority on finding the source and fixing it as quickly as possible.

Fixing problems costs money, though, and things that cost money cut into profits.  For-profit corporations will never do the right thing voluntarily if the right thing costs them money.  They can only be forced to do so by a regulatory body.

Buzz-B-Gone!

Last week, PG&E finally sent a team of RFI hunters out, who agreed with my assessment regarding which pole seemed to be the trouble, and this week they had a crew out to replace the high voltage section of the pole:

The top of a utility pole featuring antiquated, sagging insulators.

The “before” picture- now it’s obvious how the insulators were sagging.

The offending crossbar and insulators removed by a crew.

Placing the new crossbar and insulators.

Job done!

The “after” picture; note how much further away from the pole these insulators keep the wiring

The RFI crew, once PG&E management finally relented in sending them, was very friendly, personable, and professional.  They did a great job in confirming my suspicions and told me one of them was even able to hear arcing coming from that pole while standing underneath it.  They reported back their findings, and a crew arrived this Tuesday to actually rebuild the top of the pole.  This crew was fascinating to watch- so efficient, and they did such a nice, clean job of reworking that pole.

Here we can visually see the buzzing come to an end when power was shut down and the top of the pole was replaced:

Just after 9 AM, it’s gone, hopefully for good, and with also a reduced risk of fire, damage to property, or someone getting hurt.  Unfortunately, this exercise also revealed a secondary source of interference, which is further away and picked up in the early afternoon.  This one is proving more difficult to track down; from the bearings I’ve been able to get so far, it appears to originate inside the PG&E Poplar Creek substation.  It’s also more intermittent, so I suspect it’s going to be more tricky to trace:

Data collected so far would seem to indicate that this is more of a mid-morning and early evening noise.  Because it’s further away, it’s a little more difficult to get a fix on its location, but I was able to get a collection of points yesterday evening:

The area immediately around the Poplar substation is inaccessible to the public, so the best I could do was take a few readings from the closest I could get on public property.  Note that there’s considerable uncertainty around the real origin of the noise; my antenna is directional, but the main lobe is still up to 30 degrees wide, and my headings are approximate, provided by a tool on my phone and an approximation of the direction my antenna was facing- they could all be off by a few degrees.  Thus the real noise source could plausibly be anywhere in a 20 meter radius around these intersection points.

Unfortunately, PG&E wants me to start the entire process over again for the secondary source, so I’ll have to close out the original case, and open a new one now, which means probably another 60 days of dealing with this secondary noise before they send someone out again.  At least being further away and less persistent, it’s somewhat less bothersome.

It’s kind of astounding how this utility doesn’t monitor these things on its own and drags its feet so aggressively in dealing with them when they’re reported.  Once they actually relented in getting a crew out to investigate, the first problem was identified and resolved very quickly.  It would stand to reason that they might want to proactively look for things arcing inside a substation…

« Older posts

© 2025 N6OL

Theme by Anders NorenUp ↑