Sunday, April 21, 2019

Vector takes a lightning strike

We are offshore in the Atlantic Ocean, roughly abreast of Melbourne, Florida. We are headed either for Port Canaveral this evening, or else somewhere between Jacksonville and Brunswick tomorrow afternoon, depending on conditions and forecast as we approach Canaveral. Or possibly depending on how well the electronics hold up after being zapped.

We left our anchorage in Fort Lauderdale Thursday morning. Conditions outside were lumpy and so we bit the bullet and slogged up the inside. That meant lots and lots of waiting for bridges, on the most bridge-intensive stretch of ICW. We got off to an inauspicious start; we started weighing anchor in time to make the 9am opening at the Sunrise bridge, but the snubber chain hook was fouled. By the time we cleared it we had missed the opening and we sat on a short hook for another twenty minutes and finished weighing for the 9:30 instead.

Top of our flybridge VHF antenna after  the lightning strike.

Even with all the bridges, we made it to our usual anchor spot in Palm Beach (map) by 3:30. It's a good thing we were not planning to go any further, because the Flagler Bridge was on a limited opening schedule for yet another presidential golf outing at Mar-a-Lago. Trump had not yet landed, so at least we did not have to run the ICW security zone gauntlet, or get caught at Southern by the motorcade.

We dropped the tender and went ashore for dinner, strolling Clematis Street to the end of the business district and back, ending up right back at Grease, which turns out to have very nice salads and tacos in addition to great burgers. A cover band was playing at the park as part of the city's very nice Clematis By Night program, and we enjoyed music of our era as we ate on the sidewalk.

Expecting a rainy and windy day on Friday and possibly just staying put for another night, we left the tender in the water overnight. But around mid-morning we got a window of nicer weather till mid-afternoon, and we decided to make headway either to North Palm Beach or maybe all the way past Jupiter Inlet to Hobe Sound before the bulk of the storm hit in the afternoon and evening. We decked the tender, weighed anchor without incident, and cleared through Flagler bridge before the daily lockdown.

Thursday night music in the park off Clematis.

Winds were high enough that we could not afford to wait too close to any bridges, and we very nearly stopped in North Lake Worth just to avoid it. But instead we timed the Parker Bridge by lying ahull in a wide section of the lake near Munyon Island for nearly 20 minutes, arriving at the bridge right on schedule. We had fair tide all the way to Jupiter and dropped the hook in Hobe Sound (map), right where we anchored on our way south, around 2:15, just as the first drops of rain were hitting. We buttoned up all the windows and hatches and settled in. Unlike last time, when we had to squeeze in among several sailboats, this time we had the anchorage all to ourselves, which would prove to be a liability.

The first wave of the storm gave the boat a nice rinse and dropped the outside temperature and humidity into a very comfortable range. We had cocktail hour in the saloon, but by chance, the rain let up and the weather cleared long enough for us to enjoy dinner in very pleasant conditions on the aft deck. I had brought the chair cushions inside earlier to keep them dry. The dry spell was short-lived, and after a few minutes of "let's watch the lightning show from the deck" the wind picked up, and driving rain, well, drove us right back into the saloon. We put a handheld VHF and a chart-equipped cell phone into the microwave oven for EMP protection during the storm.

And there we sat in our comfy chairs, working on our computers, as the thunder and lightning grew steadily closer. I had most of a glass of red wine next to me, and I was just wrapping up editing the last photos for my generator sound shield blog post when "BAM," a loud thunderclap simultaneous with a very bright flash of light, seemingly from all directions. We'd been counting seconds between flash and clap all evening; this one had no separation. Louise shrieked.

Video from inside our pilothouse as lightning strikes (18:45:54).

One of us said "that was close," and it took me a second or two before I processed that we might have been hit. It did not seem violent enough; our lights were still on, computers still working, Internet connected. We smelled no ozone, felt no hairs stand up on our bodies, and saw no immediate evidence around us in the saloon. But instinctively I got up and headed to the pilothouse to check on things.

The first thing I noticed was that the anchor watch circle on the chart plotter was gone. I thought perhaps a nearby strike might have affected the GPS "mushroom" on the mast; I restarted the plotter program and then rebooted the computer with no change. No other instruments appeared, at first glance, to be down. With no anchor alarm and 30-35kt winds, Louise brought her computer to the pilothouse and connected it to the backup GPS/AIS receiver at her station. That wasn't working either.

The chart on our stand-alone iPad showed we were not moving, and so we turned our attention to assessing other damage and figuring out why the chartplotter computer was not working. I wanted to go up on deck to look, but the wind and the rain precluded that for the short term. It only took a short while to figure out that the plotter was down because the four-port USB-to-RS232 adapter that connected all the sensors was dead. We plugged it in to two other computers, neither of which recognized it. I learned much later that the USB port on the computer where it had been connected was also fried.

As seen out the forward windows. Impressive.

I grabbed the single-port USB adapter that came with the boat for programming the sat dish and was able to get the plotter working by connecting it directly to the Garmin GPS mushroom. That, amazingly, was working fine. We now had our anchor alarm back, which was reassuring. We started powering up and testing all the other systems in the pilothouse, and everything looked like it was working fine, even the radar.

I decided to try to get AIS input to the plotter back by re-wiring the fancy AIS/GPS-equipped VHF radio to the one working serial port, instead of the Garmin. After an hour of frustration I finally realized that both NMEA (serial) outputs from the VHF were fried. The radio and it's built-in AIS display was otherwise working fine, and the Coast Guard later responded to my emergency comm check to let me know we were loud and clear.

We tested both VHF radios against one of the handhelds, since we could not raise the nearby CR-707 bridge on either. They both appeared to be working. And after we'd finished in the pilothouse, I even fired up the TV and sat dish; both the sat system and the mast-mounted amplified TV antenna appeared to be working fine. We started the generator and ran the battery charger, not wanting to find out in the morning that either was damaged just when we would need it. And Louise checked the really critical systems: the coffee maker, beer fridge, and sewing machine.

After the rain let up I went topside with the handheld spotlight. I saw no damage at all on the mast. But when I shined the light up at the separate VHF antenna clamped to the forward end of the flybridge top, it was clear we had taken a direct strike and that antenna was the entry point. The top of the antenna had exploded, sending fiberglass shards everywhere, and the lead-in cable blew out and arced right by the clamp. This is the antenna for the flybridge radio, which was powered down at the breaker when the strike occurred. Nevertheless I am amazed the radio still works and that it was even able to transmit and receive on the heavily damaged antenna (now that we know, we are not using that radio for transmit).

Fried lead-in cable.

The huge amount of current flowing through the flybridge top and down through the cabin sides and into the hull induced enough current in all the delicate NMEA-0183/RS232 serial links to blow the driver chips in all the equipment. But inside the boat we felt nothing, as the current flowed harmlessly around us and into the water through the metal hull. Lightning strikes on fiberglass boats can be much worse, because all the current flows to the water through very much small metal bits like through-hulls, which can blow right out of the boat and sink it.

After Louise retired for the night, and still amped up on residual adrenaline, I finished up my generator blog post and then went to our video camera system to see what it might have captured. The camera that faces out the forward pilothouse windows quite clearly captured the lightning bolt. Stepping through the video frame-by-frame I could see the arc moving laterally everywhere except right where it met the boat.

Video inside the pilothouse clearly shows multiple flashes. This is the only feed with audio and you can hear the bang, and the screams from the saloon. Watching the plotter screen shows the anchor circle disappear a few seconds later; the plotter has some hysteresis after loss of signal. Shortly after that we come into the pilothouse to respond. Well, some bald dude does, anyway.

As seen from the aft deck.

The aft deck camera is less interesting and only shows the flash. I also ran the video from the engine room and tiller flat cameras and was relieved to see no flashes or arcing of any kind. Our engine (and generator motor) is entirely mechanical; we know several folks who had tens of thousands in engine computer and control replacements after lightning strikes. The three cameras that caught it all told me exactly what time we were hit: 7:45:54.5 EDT (the time stamps on the video are in EST, so they are an hour off).

Yesterday was a clear sunny day and we went out on deck to inspect the damage. Bits of antenna are everywhere, and I was puzzled by the very large chunk that I found on the aft settee of the flybridge, underneath the cover. It had shot through the canvas top like a bullet, leaving an oblong "entry wound." This is the canvas that we had repaired last year in Charleston, and is due for replacement. We will probably patch the hole with tape in the interim.

"Bullet hole" in the top.

After our walk-around we fired up all the instruments to get under way, determining that we had enough items working or jury-rigged to continue to make progress until we could get somewhere to receive parts. Four-port USB converters, high-end VHF radios, and vintage Furuno gear are not things you can just pick up at the next town. We set out sights on Vero Beach, where we have friends.

Getting under way revealed more damage. When I pulled up the data display on the radar/plotter to get our starting mileage it was scrambled. I got the mileage from another screen, so the scrambling turns out to just be the display format. No heading information was getting through to the AIS. And the stabilizers would not come online for lack of a speed signal, which also comes from the radar/plotter suggesting that some of its serial ports, too, are fried. We have the speed signal bypassed to keep the stabilizers working at the moment.

Other than a windy cruise we otherwise had no issues running. on the jury-rigged plotter. With only one working serial port I had no way to let the plotter drive, so I used heading mode on the autopilot all day. When we knew we could make it to Vero I reached out to our good friends Alyse and Chris, but they turned out to be out of town. We decided to continue to Vero anyway, where there is a nice anchorage with access to shore and a couple of restaurants; I thought I might even be able to pick up a second single-port serial adapter to improve our life a bit.

A beautiful anchorage. And the tallest thing in it..

The vagaries of tide and current put us abreast the Fort Pierce Inlet twenty minutes too early for the opening of the North Causeway Bridge. Faced with station-keeping for that long, we decided to just call it a day and drop the hook in the little anchorage south of the bridge (map), where we've stayed before and where we knew there was a decent beach-bar restaurant where we could drown our troubles in draft beer.

We were just about to order a USB adapter on Amazon Prime for delivery to Titusville when it occurred to us that we were right next to an inlet, and these storm-driven westerlies would make for good ocean conditions. And so before dinner Louise pulled down the forecasts and we made a tentative decision to go outside today, possibly all the way to our next stop, in the Savannah area.

A big heavy meal and a couple of drafts does not do wonders for one's chartplotting abilities, but after dinner I was able to run numbers and conclude that we could only make Savannah in the daylight with a late departure from Fort Pierce. A late departure, however, would mean we would lose the very best part of the weather window, and also mean that a bail-out to Port Canaveral would have to be well past sunset and possibly past the time when the lock closes, leaving us fewer stopping options there.

A bit of burnt material around the antenna clamp.

We decided instead on a dawn departure, which would make a nice day to Port Canaveral if that's what we decide. That means that we can make it only as far as St. Simons Sound, where we have a familiar anchorage near Jekyll Island, by a reasonable hour tomorrow evening. And that's only if we maintain a fairly high speed made good. Quite possibly we will be stopping sooner than that, either at St. Mary's Inlet, or at the St. John's river. We won't really know until mid-day tomorrow.

As you might imagine, when not driving the boat, I have been busy every waking minute since the strike, effecting repairs. I've spent hours squeezed under the helm staring at NMEA junction blocks and re-routing signals to recover functionality. Every NMEA-0183 connection has two ends, and a non-working link could be down due to the transmitter end or the receiver end or both, and I've done a lot of parts-swapping and troubleshooting.

Yesterday I was able to get another serial port working on the chartplotter computer, which I robbed from the now-useless system that fed Louise's computer at her station. She's using a stand-alone GPS "puck" to keep her plotter running, albeit without AIS input. That required some Windows driver legerdemain, because it is an older, unsupported converter chip. The second port meant I could connect the computer to the autopilot to drive the boat, virtually a requirement for an offshore passage.

Inside the four-port USB converter. No obvious damage but the driver chip is dead. The power LED comes on, though.

I replaced the dead flybridge depth display with a spare that I bought used some time ago. And just before bedtime last night I replaced the autopilot with the fully-configured spare that I had taken out some time ago, to restore our heading information. That was a calculated risk; replacing an autopilot just before a passage can be a bad idea, but I knew we could just turn around and continue up the inside if we had problems with it.

Sure enough, as soon as I engaged the autopilot this morning, it veered off well to starboard before sounding an alarm; I ended up hand-steering out of the inlet. The last time I replaced the autopilot I had to run the calibration procedure, and you need to be in the ocean for that anyway, or at least a very large body of water with no traffic, because you have to drive the boat around in circles.

Reasoning that the problem might just be a calibration issue, I hand-steered while Louise checked or set all the autopilot configuration parameters, and we ran the calibration procedure. That didn't fix the problem and neither did changing a half dozen settings. We made a nice loop though, which added a few minutes to our trip, as did stopping the boat to run rudder calibrations.

Top view of the "bullet hole."

To make a long story shorter, after a few more tests we realized that the rudder was turning to starboard when the autopilot commanded either a starboard or a port turn. That turned out to be a wiring problem, either a loose or shorted wire where the hydraulic solenoid connects to the autopilot. Cutting the ends off and re-terminating the wires fixed it, and it turns out we probably never needed to touch the settings to begin with.

All is well now, and in the time it has taken me to type all this, we're abreast of Patrick Air Force Base and have made the decision to round Canaveral and continue north. Our immediate goal is Fort McAllister on the Ogeechee River, where a painter will quote us on some touch-up work. We're also hoping to catch up with good friends John and Laura Lee; John is the former master of this ship, then called Steel Magnolia, and recommended the painter to us as someone who'd worked on this boat before to good effect. If we don't get a window for another outside hop, we will work our way up the inside.

Sunrise this morning as I hand-steered out of Fort Pierce.

While a lot of the foregoing sounds like we've had a miserable couple of days, I want to be clear that we consider ourselves very, very lucky. Not just in the "first world boat problem" sense that you've seen me write here before, but in the sense that our lightning damage is minimal. We know several folks who were dead in the water after a lightning strike and spent months in the yard and tens of thousands (or, in one case, hundreds of thousands) of dollars on repairs. We also know someone who lost their boat.

If I called a marine electronics company and asked them to fix all this, it might approach $10k, and that's far enough above our insurance deductible to file a claim. We'd also get some newer gear out of the deal. But I'll replace all this stuff from the used market and Amazon for less than $2k (plus my own sweat equity), which is under our deductible. Mostly we just need a place to receive the equipment, at this point most likely Fort McAllister.

We were also lucky to have our tall SSB antennas lowered, which we had done for the gauntlet of bridges. We have not yet raised them and tested the SSB, but for sure if lightning had hit one of those antennas it would be damaged, and that's a more expensive piece of equipment than anything that was lost.


  1. I don't want to play detective here, but I distinctly hear two screams on the video, one lower pitched than the other. I was impressed by how quickly Louise reacted,demonstrating that her senses have most certainly not "lost a step". In fact, my timing has the whole thing from flash to end of scream is just over 1 second. I am elated that you guys got out of this light, and in one piece. I know your mind is already fast apace on a scheme to make it even less of an issue next time. Safe travels.

  2. Ouch! You guys were way more cool than 4 of us guys in the cockpit of a Western Airlines 727 landing at Portland in a T-storm. We got struck twice in 2 or 3 minutes. The first stroke hit just outside the First Officer's sliding side window (he was in control at the time) and the second hit on a screw on the radome. We had our harnesses really tight. When the first stroke hit I would swear that the CAPT's head almost hit the overhead panel. They immediately ran through the emergency checklist in very high-pitched voices then the Second Officer (Flight Engineer) picked up the cabin microphone and, in a calm deep voice, explained to the passengers that all was OK. Seconds after he replaced the mic we were hit again. When we got on the ground the 4 of us broke out laughing to relieve the tension. When the S.O. opened the cockpit door and we looked into the cabin, I've never seen such frightened looks on anyone's faces.

    BTW, Happy upcoming anniversary!

    Pat and Nancy


Share your comments on this post! We currently allow anyone to comment without registering. If you choose to use the "anonymous" option, please add your name or nickname to the bottom of your comment, within the main comment box. Getting feedback signed simply "anonymous" is kind of like having strangers shout things at us on the street: a bit disconcerting. Thanks!