Fire! Lost my Beneteau. Cause?!?

[Captain Scott]

It was a sad day earlier this year when my Beneteau 43 caught fire and sank. Luckily we were near shore and everyone got off it quick and easy fashion. After going through every possible cause (the boat had just come out of the yard and the only electrical work was the replacement of a bilge pump) the only I reason I can see is a short in electrical right at the autohelm. But is this not fused? There is nothing else back there except a liferaft at the transom. The first part of the deck that started to melt away was at the emergency rudder post cap and surrounding area. I just started thinking it could have been a lituim battery someone brough on board and stored behind the access panel in one of the two aft cabins. I REALLY want to get to the bottom of this the best I can. When you think everything is good and then it goes total wrong, it is sole wrenching. Thanks in advance for any insight.

[Garrettw]

Sorry to hear but glad all are safe. How is your fuel system routed? For example, my Jeanneau has an aft fuel tank, aft genset and aft fuel/water separator. One time I pinched an o-ring in my fuel water separator and flooded a hidden bilge area under the aft berth with several gallons of hydrocarbons. Consider that the metal fittings and other appurtenances related to your steering gear may have conducted heat from another area.

[SailingHarmonie]

I would not be surprised if most electrical fires on boats occur in properly fused circuits. The reason is that most circuits on most boats are properly fused, or at least close to it. The causes are things that fuses can not stop.

What causes fires in fused circuits are connections with high resistance passing high current. The amount of current can be totally within the wire's rating to carry it, and well below the fuse rating, and completely within normal operating parameters, but if there is a loose and/or corroded connection that current passing through high resistance can generate a LOT of heat. Unfortunately, because of motion and salt water loose and corroded connections are not terribly rare on boats. In high power equipment, the bad connection can be internal just as easily as it can be external.

Let's assume your autopilot pulls 20 Amps when the drive is working hard. If you have a bad connection with a resistance of just 1 ohm, that will generate (20*20)*1 = 400 Watts of heat right at that connection. Think about how how a 400 watt heater gets. If you are lucky, the wire melts and breaks the connection before something catches fire...

Of course this might have nothing to do with it, but you need to think about causes other than simple short circuits.

[Lake-Effect]

Originally Posted by SailingHarmonie

I would not be surprised if most electrical fires on boats occur in properly fused circuits.

Let's assume your autopilot pulls 20 Amps when the drive is working hard. If you have a bad connection with a resistance of just 1 ohm, that will generate (20*20)*1 = 400 Watts of heat right at that connection. Think about how how a 400 watt heater gets. If you are lucky, the wire melts and breaks the connection before something catches fire...

12v cannot push 20A through 1 ohm. Ohm's Law.

If the autopilot drive pulls 20 A max, then we can pretend that it's a load of V/I or 12/20 or 0.6 ohms, and if said circuit then developed a high-resistance connection fault, that extra resistance is in series with the autopilot... so the total current would no longer be 20A. In your example, the total load would be 0.6 + 1 ohm, which means total current would be 12/1.6 or 7.5 A, and the power in that 1 ohm bad connection would be I*I*R or about 56W. Sure it's still enough to possibly get to a ignition temperature, but way less than your 400W.

In real life, I'd expect that if such a 1 ohm fault developed in the autopilot circuit, its presence would be immediately apparent in the autopilot going weak and unable to steer the boat in heavy seas.

This is the long way around to say that I don't think that properly-fused circuits cause the majority of electrical fires on boats.

[Statistical]

Sorry for your loss and glad nobody was hurt.

"But is this not fused?"

The challenge is fire destroys the evidence. The autopilot 'should' be fused and it should be fused at the source not the autopilot itself. Was that done? If you don't know unfortunately the evidence is likely now gone.

This is one reason I am not a big fan of inline fuses. It isn't obvious at the busbar that the circuit is protected and it provides no protection "upstream" from the inline fuse. Some autopilots also have an internal automotive type fuse which can lead some installers with the wrong conclusion that "it is already protected". The autopilot is but the wiring isn't.

Did anything of the boat survive to do a post mortem?

[tanglewood]

I'd start with what else was in the area where you think the fire started. It sounds like some autopilot components. What about engine controls, thruster controls, or any other electric stuff? There might also be wiring passing through, even if no electrical devices are there.


Also, what other work was done on the boat? And especially any work done in the area where you think the fire started?

[Captain Scott]

Great answers! Thanks to all. Now... More background.

And I will answer all the question to this point from others as well in this response. First, a more in depth overview:.

The fire started aft of the engine compartment and both Stb and Pt aft cabins. In the space there is only the autohelm a stern light wire and a wire going to the propane tank. All of these were off at the time. The auto and (on occasion (shown "Low Voltage) when there was NOT low voltage. I am not an electrician, but after reading some of the more indepth electrical scenarios, I though to incude this

One piece of information I did not include was the observations of the captain and crew as the fire progressed.

THe autohelm was not on at the time. Engine had been idling in gear for about 6 minutes as the main was being hoisted.

The first thing was faint black smoke from the pedestal under the center cockpit table. There was and access plate it that pedestal that was off at the time and that is were the the whisps of smoke first came out. Went below and nothing was wrong. Engine compartment was check and no fire there. Came back up with a fire extenguisher and emptied it into that port with little effect as the smoke was increasing. Went back below and there was now smoke in the salon. Came back up and opened the Port cockpit locker and flames came out. (Engine controls are on the Stb Side). Shut that lid then everyone got off the boat and into the dinghy (being towed at the time).

?Fuel Tank/Connections a possible cause?
Fuel Tank was under the aft stb bunk. The Racor, in the engine compartment

?Fused?
It had a breaker switch on the panel. Nothing on the boat survived for inspection.

[Montanan]

Glad to hear that you safely evacuated, sorry to hear of the loss.



Hope you were insured.

What was housed in the locker from which flames were apparent when you opened the access?

Alternatives:

Electrical origination, e.g., over current resistance heating, or alternatively of spark ignition of nearby flammable materials.

A prospective cause:

Spontaneous combustion, runaway exothermic chemical reaction.

E.g., a linseed oiled rag.

[Been there done that twice in the same evening after spending the day preparing the exterior of our home for painting; one rag placed in the pocket of my wife's painting smock which was hung on an entry coat hall tree indoors, and another outside in a trash bag that I had placed outside on a stone patio.]

[Statistical]

Originally Posted by Captain Scott

THe autohelm was not on at the time. Engine had been idling in gear for about 6 minutes as the main was being hoisted.

...

Sometimes autopilot being off doesn't mean it isn't energized if we are talking about the display. The on/off on the autopilot display is simply and soft on. It is possible for the autopilot to be on an always on branch or even directly connected to the battery. However ...

It had a breaker switch on the panel. Nothing on the boat survived for inspection.

Since the autopilot had its own breaker which is also a switch if it was switched off then the entire circuit would be dead. If you are sure it was off at the breaker panel then whatever started it it is unlikely to be the autopilot.

[donradcliffe]

Give us a list of everything the yard worked on.

[CharlieJ]

@LakeEffect #4:

If the autopilot drive pulls 20 A max, then we can pretend that it's a load of V/I or 12/20 or 0.6 ohms, and if said circuit then developed a high-resistance connection fault, that extra resistance is in series with the autopilot... so the total current would no longer be 20A.

You are on the right track until the bold section. The connection fault in series would not change the current in the circuit.

[Lake-Effect]

Originally Posted by CharlieJ

@LakeEffect #4:

You are on the right track until the bold section. The connection fault in series would not change the current in the circuit.

Sorry, what? if you have 12v feeding a load resistance of 0.6 ohms, and then a resistance of 1.0 ohms (the proposed resistance of the new fault) is added in series into the load circuit... (total load is now 1.6 ohms)... the current MUST come down. No?

[Statistical]

Originally Posted by Lake-Effect

Sorry, what? if you have 12v feeding a load resistance of 0.6 ohms, and then a resistance of 1.0 ohms (the proposed resistance of the new fault) is added in series into the load circuit... (total load is now 1.6 ohms)... the current MUST come down. No?

You would be correct if he autopilot was a static resistive load but it is't one. We can model it as a resistor but it likely has a switching powersupply. If there is a high resistance connection that will result in a voltage drop. The power needs of the AP haven't changed so when the input voltage drops it will compensate with an increase in current at least until until OCP in the autopilot trips.

If the OP however is correct that the AP had its own breaker on the panelboard and it was off then the entire circuit would be cold. While the circuit breaker could (unlikely but possile) fail in a fault condition itis very unlikely that it wouldn't open if opened under normal conditions. That would mean the AP is very unlikely to the source and that includes any wiring, connectors, and OCP for that circuit.

[Lake-Effect]

Originally Posted by Statistical

You would be correct if he autopilot was a static resistive load but it is't one. We can model it as a resistor but it likely has a switching powersupply. If there is a high resistance connection that will result in a voltage drop. The power needs of the AP haven't changed so when the input voltage drops it will compensate with an increase in current at least until until OCP in the autopilot trips.

This is of course the problem with oversimplified hypotheticals. Given the proposed 1 ohm series fault, the available current from 12v can not exceed ... 12 A.


Anyway, OP tells us that the AP was off. Moving on...

[Statistical]

Originally Posted by Lake-Effect

This is of course the problem with oversimplified hypotheticals. Given the proposed 1 ohm series fault, the available current from 12v can not exceed ... 12 A.


Anyway, OP tells us that the AP was off. Moving on...

Derp yeah missed that. You are right simplified examples comes with their own problems.