top of page

My Life-Sucking Gray Swan..

Updated: Jul 10

This blog entry is an approximate timeline and digest of lessons related to what in the end was a simple, but uncommon leak triggered by a designed-in electrical fault. It’s long because I know there will be discussion of it at my end because there’s always such discussion and such tomes allow me to save time when talking with others. Most boat owners don’t need to read most of this but if you have an NT42 with an Onan MDKAW generator, or some other setup as described below, you should at least consult the highlighted text in the last section – Final Comments and Advice.

2nd note - I've had several requests for certain documents I created in working up the repair. I can't do that. Will happily share why in person.

I decided to enthusiastically pursue ex-boater status while in an ambulance in Springfield with a potentially serious (drop dead serious according to the paramedic) issue from my boating life a day or two earlier… call that the “the straw”. I won’t get into the full range of things that motivated me to decide to gain distance from boating/ownership and back to more familiar life-long passions. They don’t matter here. The way things unfolded after that decision in the ambulance was determined by three  things… One was a spontaneous, atypical and utterly factory-avoidable oil leak at the generator.. untriggered in any sense present in my background. Leaks that just show up after ~6 weeks of hands off and move all your oil to outside the engine always have a trigger and often warn you they're coming. Except this time. This one was a designed-in electrical fault resulting in a total loss of the engine’s oil. Not an everyday thing in my world. This will be the star of this writeup.


The second thing was a huge (to me) change in status of my body from a bit better than most guys my age early April 2023.. strong and able to do/enjoy any flavor of $hit physical work for long hours.. to much less than that, and ultimately unable to get into any of the positions or do the basic maneuvering in the engine room to diagnose the leak.. or at the end, even climb the ladder. The third thing was I was forced to rely on the pool of marine mechanics. And sadly, my history and that of multiple others was that pool has trouble with basic technical and business skills, and ethics. Despite my direct personal experience and that of too many others I know, I genuinely believe there MUST be skilled, ethical marine mechanics out there because the Gaussian distribution doesn’t just predict the existence of the crappy ones (left side) or mediocre ones (middle) BUT it also predicts the excellent ones (over on the right). These two issues certainly affected the response to the failure by stretching it way out.. but they didn’t create the problem. Low/no-effort design of certain elements created it. From here forward, I’m going to try to avoid saying much about these two things and instead try and discuss the path from first observed symptom to final solution.


After the ambulance/hospital fun, I promptly engaged a Nordic Tug specialist (ok… correction to above.. he is definitely a good and conscientious mechanic and I recommend Tom enthusiastically) to get a second set of eyes on things, and he also decommissioned the generator. Brought CW back to my home marina a few weeks later, decommissioned the remaining systems while in the water, and hauled it where I prepped the hull for storage (primed ~2 sq-feet of bare hull and prep the metal) up on the hard. The interior IMHO was perfect (except for the $480 carpet stain, in case I haven't mentioned that earlier). See my final departure video in the photos section.. at least I was happy.

I “do” decision in the presence of opaque stochastic risk for a living, so the plan from the start was to put her up for potential long-term storage, just in case. Like, maybe some unforeseen thing will bring the entire world to a full stop for a year or two or something... or maybe, inconceivably, my brilliant yuppie friends will be proven wrong and my family might actually someday need a few of those 350 N95's and gallons of Purell I had sitting on a shelf for >15 years.

Up on the hard, the day I finished the hull stuff, I believed that I was done, and for the last time. As I cleaned up my work area talking with my friend, I believe what I then said was something like “Well… last time I’m going to do any of these things.. let’s go get some dinner.”


Introducing the Problem


But, sometimes, at the height of our revelries, when our joy is at its zenith, etc. I got up on CW to do my usual pre-departure check, and found a river of oil in the port fairway of the engine room.


Insert my idiot initial response here -  In ~1980, I received some “leak test training” in conjunction with a bunch of (real) non-destructive test (NDT) training as I had to take over a family business when my dad passed away. All the other NDT training made sense for the various processes where I’d be technical lead (Level III Inspector). But not the leak testing or safety wiring. I utterly wasted my time trying out what I was able to remember from the leak training 40 yr earlier..  mostly trying to estimate the amount of involuntarily released fluid (reminiscent of Taleb’s reverse process mentioned in The Black Swan on the difficulty of precisely back-predicting what an ice cube looked like before it melted by examining the puddle), and adding UV tracer dye to the oil and using a UV source to find where it’s coming out. I went thru two rounds of dye testing.. to leave no doubt about my idiot nature. Fill up the engine with dyed oil... watched it for 5 or 6 hours... and NOT SEE A GD DROP either time. Then 5-6 d later observe it had all leaked out. Right then, I should have known it was 1. a very small pin hole or two, and 2. Located as low on the engine as possible. That it was NOT a "water-main break" type of flow.


Finding a leak in an oil pan should be trivial if you can see what you need to see, which I could not. It turns out a NT42 engine room is the IDEAL trash compactor for eating problem knees/hips/back/neck by staying wadded up in a little ball the whole time you’re trying to operate [2023 in fact was the most medical year of my life with muscle and joint failures triggered more each time I tried to work down there]. I hung in trying to do that job too long and decided, while standing in a giant pile of shame and swearing constantly for days afterward, to try and find a marine mechanic to do the work.


Let me shorten this part up… the mechanics “didn’t work out”. At all. 3 of them. Identical failures consistent with all my prior attempts.


Diagnosis - Despite being a pretty terrible person, I keep on attracting exceedingly incredible friends. My friend Doug is one of the most mechanically adept/experienced and intellectually curious people I've been fortunate enough to know, and just decent and interesting to be around... mostly I like he fights me when I try to pay for dinner. He just kept getting on my boat and studying/working the situation. I don't like how it makes me look that I quietly considered he was trying to establish a squatter's claim... I don't usually do that to friends:) The engine area was a royal mess from my stupid runs at applying Air Force leak testing (hmm.. only now wondering.. do they even have boats?? See..... I'm an idiot).. he fab'd up a tool to facilitate cleaning up the dyed oil/insulation remnants in the twists and turns of the spill tray under the engine. People who can't let a problem go and work to apply actual deep experience and thinking to observations are THE most interesting to be around. Despite being “a few months” older than me, he soon produced a perfect clear perfectly lit photo of the cleaned up oil pan wound. It was bottom/center of the aft vertical surface of the oil pan. Very obviously created by two copper braided bonding cables routed exceedingly loosely through mid-air from hard-points on the drip pan to the upper midline of the engine block. The carelessly draped cables were reminiscent of typical last mile power distribution in India. Repeated cable contact burned 2 pinholes of ~2mm diameter through the pan wall. 


First photo showing leak as 1 or more pinholes.

Fails to show how much lies between the photographer and the wound.

Also.. I firmly palpated this wound before I executed the coating to make sure it wasn't delicate. Plenty firm.

Shows the aggressor bonding cables. Numbered sites on photo facilitated collaboration with a friend at the boat to rule out features that visually merit specific verification, hands on and with probes. The only pinholes were 2 located inside site 6.

Attack Plan - A mechanic in the fall said the hole was 1/2 - 3/4". To be fair, that was hard to see before the cabinet was fully removed, so I do not fault him for that. At the time, because of the size, I reflexively placed the repair into the mandatory effort-festival of the hoist-and-swap category and kept it there until Feb. You don’t have the room to do that. You don’t have a hatch over it. You might do it with air bags and cribbing which has a role in removing 4-cylinder motorcycle engines from their frames, but what do you do with the cribbing once you get it high enough? And you probably have to separate the engine and the alternator to make it easier in one sense, but then you have a potential new pushup of aligning them again on reassembly.. maybe you don’t, but then, maybe you do. And it’s just too unsafe to me risking a 700# engine collapsing onto me or the main engine or a fuel tank.


When Doug confirmed pin holes, and then seeing it in his excellent photos, I was instantly freed from the constant disgust with unnecessarily poor design decisions (which bought the deciders literally nothing) that screwed up my world for too long….. to extremely excited for the in-situ, low-drama repair that was now possible.. no hoisting… and even more excited for the guaranteed serotonin flow of working out the absolute best practices I could, and executing them precisely. I’m broken and LOVE solving new problems.  AND… not having to be at the mercy of intellectually and morally diverse assistance.


Forming/Memorializing the Repair Rules/Process - My time away from Clock Work became 100% about researching this repair instead of searching for a marine mechanic that can do super-human things like call back, or do what they say you’re going to do. Don't shoot the messenger. The performance mode would be to operate to the same level of professional learned rigor as was (note past tense) common across companies with strong curated design cultures decades ago, like the ones that pushed out the W123 Mercedes, DEC PDP 11/70, and the M1 rifle [pardon my toxic masculinity]). The next several weeks or so, I poured myself in to an epoxy-based repair, gathering success (multiple dozens) and failure (none) cases from my severely mechanical closest friends.. who range from race car and off-road motorcycling maintenance.. in-service plant machinery repair... oil field equipment repairs.. aircraft mechanics... and multiple others. All reported use of epoxy for machinery repairs and they are intrinsically conscientious and detail oriented. Further, discussed and cleared my own open questions about success factors with apps engineers at three epoxy manufacturers under consideration (see note *). In two of those cases it spanned multiple conversations. Apps engineers who take pride in knowing the performance envelopes and best practices for their products actually like the rare opportunity to go deep (ditto coatings engineers BTW). I was on the lookout for any possible failure triggers like thermal and vibration for the materials being considered, and, of course, success factors. I didn't encounter one single failure case.


The sole repair failure triggers I found were: poor surface texture prep, surface contamination, air/substrate temps/humidity out of band and significant post-cure vibration which was a non starter. It’s quite a bit like applying high-performance aerospace coatings which was a big part of my technical life in the early 80’s.

Every possible detail.. every success or failure factor.. was collected into a ~13 page “project document”/bible. Project documents are a decades old habit that provides me with the following advantages:

1.    Prevent useful knowledge from falling through the cracks

2.    Prevent bit rot - thoughts and reasoning can become corrupted over time

3.    Prevent anchoring on the last thing that was has been said/seen/pondered as the best best/central idea


"How can you ruin a perfectly good $150 oil pan?"

That's a corrupt version of asking why I did it... so here:

TO LEARN (the unknown unknowns)

To prevent a time-intensive do-over.

To master kneeding the putty

To see how the putty feels as you initially kneed it during your 3m/15s crush time and how distributing it feels in your 2 min application window.

To determine the optimum crush time.

To see what the temp of the putty is when it's ready to distribute and when you better be done screwing around which I found useful.

To DISCOVER other things that might be important to the job, like being prompted to find a method to keep the putty out of the interior, below.

I promise, on my deathbed, I will not say "Man, I really wish I didn't F up that oil pan and instead returned it!"

I also repurposed the unused replacement oil pan I bought in 2023 to be a test fixture for site prep, application method, and validating repair durability. The oil pan had been transformed into a test fixture with built-in holes precisely sized and positioned to simulate the faults in the installed oil pan. I think it was 9 separate experiments were conducted on that fixture and the takeaways from those tests definitely affected the repair process. Sorta-sidebar: I made a goalie mask for myself back in the 70’s which I wore for maybe 8 or so years. Actually tested it with a .22 (not actually while on me).. it passed. I wished that whole time I used it that the one for me was my second.. multiple success/failure factors were learned that would have improved fabrication and use.. e.g. somewhere in this room is a small piece of plaster of paris with a large number of my eye lashes from my teen years sticking out of it…. always use lots of vasoline. It would have been nice to learn that on another goalie’s eye lashes instead. Ditto on this test fixture.


That document was circulated and discussed. Best/required practices for the particular repair tack were then built into a repair script. I worked full time at discovering everything I could to uncover rare failure triggers and best practices before I finalize the repair script. None found, but I know from past projects there’s enormous value in discovering even one failure factor on the test fixture instead of the real machine/system. Basic rule from the product design/build cycle.. the earlier you discover that in that cycle, the cheaper it is to fix. The old value of the savings approximated as one order of magnitude per stage of that cycle.



Fortifying the Solution


The initial repair will very likely outlast the boat, based upon the numerous collected cases.  But sadly, I have design optimization/overkill OCD and after the initial repair, spent about 2 weeks obsessively pondering IF it were to fail, how might it fail. I didn’t have a single reasonable mechanism I’d offer given the plethora and wide range of success cases. But I did have an overkill belt/suspenders idea in the back of my mind for which I could see no down side.

It was an epoxy collar that extends an inch or so from that perimeter line.. the addition of a reinforcing layer to the perimeter of the first repair where the thickness of the already applied epoxy is minimum and blends to zero. This collar further extends how far any oil that got past the initial repair would have to propagate to comprise a failure, and increases the total tensile strength of the overall repair (more area). It was applied to extend a bit more than an inch from both sides of the perimeter of the first repair. As I said, I see no downside and it appears reasonable this layering of defenses will push the MTTF further out. The collar was executed with the two part original JBWeld. The original repair (JB Weld SteelStik putty) had a tensile strength of 4000psi. The 2-part original had 5020psi.


Discussion – Regarding the Choice of JBWeld


This is worth discussing. I've had some questioning looks/voices on this. How can an $8 tube of silly putty be "good enough for such a fine yacht". I can understand to a degree questioning the choice of a solution you can probably get out of a vending machine some place. But the people who reacted as I described weren't the actually mechanically fluent people whose advice I sought. They mostly came across as "forum smart"... never did X but they saw a 20-something's tic tok video or something.

I had used JBWeld twice, both times some years ago. One was on a dirtbike ride up in VT when one of my riding partners knocked a chunk of his engine case out on a rock, returning some of his oil back to mother earth. We laid his bike down, wiped the repair site off with his t-shirt and epoxied it back with the putty… 6 min cure time, but wait 20. Put some of my oil in his and back to the truck. If I recall, it lasted more than a year or so when he got rid of it.


Unfortunately, with two wonderful experiences under my belt, I failed to be impressed even slightly how a durable repair could be made so simply. I carry it in my truck. On my dirt bike. I have 4 or 5 of them hanging near my fuel oil tank in the basement. I used some to reinforce a riving knife I added to my table saw. And I never became impressed until I collected that huge range of success stories.


Is it good enough for this job? The job is to hold 4” of unpressurized warm fluid inside the unstressed oil pan of a non-critical system. The most mechanically sophisticated people in my life unanimously say yes. Beyond anecdotes, what about numbers? Ok… versus brazing… Brazed steel has a tensile strength of 40-65,000psi. Max tensile strength among the JBW products is just over 5000psi. But then look closer... A brazed repair for the pinholes in this fix might cover ½ sq-in. The combined footprint of the implemented JBW repair was maybe 5 sq-in. So a total grip of the braze might be 20-32,000#, and the implemented repair 25,000#.  Maybe a bit more with the layering of the belt and suspenders repair? Why doesn’t that work even on such a fine yacht? And I didn’t have to bring my acetylene tank into a closed space with a body no longer made for escape.


By the way, last night I was fixing a heating subsystem in some kind of yuppie kitchen appliance for a friend. When I looked close, I could see a ceramic insulator for one of the wires coming off the heating element had split into 2 pieces and several more very small chips. I temporarily joined the two big ones with cyanoacrylate and built it back into its final shape with JBW, which is non-conductive. Fast easy fix. I plan to repair the skid plate on my TY-250 and on the heat shield on my truck.. the mounting bolts pulled through on both. That’ll be easy too. I would now call the entire epoxy category amazing products.

Final Comments and Advice


My boat should have gone to a new home several years ago, and I didn't want to sell it with an unusable generator. This bizzaro failure ate a pile of my heartbeats and likely 8 months of wall clock time. This was an uncommon leak, progressing to full-tilt failure unstimulated 6 weeks after going on the hard.


Whoever decided to DESIGN IN those two completely exposed welding cables jumping around unrestrained in mid air just screwed me for no good reason at all, and I didn’t appreciate it. Lazy design. My NT 42’s Onan is an MDKAW. Now both cables live in sleeves of heavy duty heat shrink tubing. How did this ever get out the door like that?  If you have a 42 with an MDKAW on her, check your bonding cables behind the engine. If they’re swinging free, I suggest you put them in separate 1” HD heat shrink sleeves (see below). Personally, I only heat-shrank mine at the very top as I’m not in love with running my butane torch in an enclosed space with difficult exiting for me longer than I need. If you have other kinds of boats with other kinds of generators, I would still suggest you look for the pathological condition of free swinging uninsulated bonding cables and if present, same answer. 


Philosophical point - Given the line of friends and others who made remarks like “You’re overthinking it” or “I can’t believe you destroyed a brand new $150 oil pan”, all I can say is over 55 years and 2M miles I have built up some f'ing great habits and philosophies that have done nothing but serve me and others well. The collateral learning from the dozens of other test fixtures before it, and this one, taught me vastly more than I could ever get by paying $150, or $1500, to someone to teach me, if such a person existed. I was ready to pitch, and knew what pitches I was going to throw the second I took the mound. My second piece of advice would be that if you were going to do this repair, building a test fixture will serve you well.  Doesn’t have to be a new pan, or the same one, to learn useful and valuable things. I know I made adjustments to how I applied the epoxy.. timing, pressure, etc. based on feedback from the fixture.


As long as we’re speaking of design, that row of zip-ties along the head-liner and cut off to mimic the morphology of stilettos or shark’s teeth penetrated my skull a couple of dozen times during this adventure. My friend Doug had some blood coming out of his head a week or two ago too. War crime. Zip ties are brutally simple things.. might as well put them in smart instead of the way they were put on mine. You won’t get good at the big stuff if you just phone in the little stuff.


If those swinging welding cables in the back of the gen had turned out to produce much more than a pinhole issue, I’d have been more screwed. An access hatch where you could hoist the motor enough to get the pan off or the complete gen out might cut a few years off some poor guy’s trip to hell. Boat makers can do vastly better than dropping all the big things into the hull and then gluing the deck/superstructure on.


BTW, a standard hatch above the starboard fairway would be smart too. Of course, I’m assuming they’re selling more of these things to old guys than Chinese acrobats.


Also.. it’s not just the designers.. I blew enough things too. The knowledge this was just a pin hole was delayed substantially by my carelessness. I SHOULD have known.. slow onset failure (>6hr to start to flow) and TOTAL release within a few days screams pinholes. I wasted 8 months on just this one intellectual choke.


Further, I initially concluded the only way to fix a ~1 inch hole, or more, was to hoist and swap. I’ve done a pile of experiments the last month and a half and would absolutely and confidently first try a JBW solution to a 1 or more inch hole on an unstressed component, most likely a composite approach with fine mesh screen.

Ready to go


207 views0 comments


Commenting has been turned off.
bottom of page