Before we left Saint Jean de Losne for the winter of 2017 and when we returned at the beginning of March in 2018 we had the boatyard H2O cut six large access hatches through the floor in Wanderlust’s living area. These holes were cut so that we could determine where the diesel fuel in her bilge was coming from. It was a drastic approach to resolve a troublesome problem.
The idea of cutting access holes in the living area was not actually H2O’s. It came instead from a conversation with the principal of a neighboring yard in Saint Jean de Losne. Though in the end the neighboring yard wasn’t interested in doing the work on Wanderlust, the discussions were still worthwhile. Indeed, in planning the repairs we learned many useful things by talking to boatyards who in the end would not do the work.
Until H2O cut access through the floor it was unclear where the diesel in Wanderlust’s bilge was coming from. The obvious source was from the integral fuel tanks welded into the hull nearby. But there were other possibilities that we could not rule out.
The most visible pool of diesel was at the stern portside end of the forward cabin bilge. There was some possibility that this was from diesel spilt in the engine compartment during a fuel filter change. We postulated that fuel in this area of the engine compartment might be working its way through small hole in the bulkhead to the forward cabin bilge. This didn’t seem likely and there shouldn’t be a hole through the bulkhead in any event. Nevertheless, before access was cut, we could not rule this possibility out.
On the starboard stern we had no idea of the either the source or the extent of the diesel in the bilge. Without access it was impossible to see all but the very stern end of this area of the bilge. As far as we knew there could have been just a small amount of diesel.
The source of the diesel on the starboard side was also mysterious. One possibility that we could not eliminate was there was fuel in the bilge from a spill during the build.
There was some reason to think that there could have been a spill. The builder claimed that Wanderlust’s tanks were filled out 50L per minute in the plant without blowback. We were there at the time. From the time stamps on the pictures we took the claimed fill rate seems about right. But how was this possible with the tank vent and fill lines configured as there were? The vent line for the white diesel tank runs below the fill line. This means that vent would always be blocked by fuel at the end of the fuel fill. And when blocked by diesel there would be no effective tank vent. In this configuration, in the absence of an alternative vent, blowback would have been inevitable the first and every other time the tank were filled. Indeed, we never succeeded in filling either the red or the white tank without blowback, even at fill rates far lower than the 50L/min claimed by the builder. So how was fuel added in the plant without blowback?
Thus we considered it possible that the inspection ports on the top of the fuel tanks had been opened as a vent when Wanderlust was first fueled up in the plant. Since the tanks were of “known” volume and fixed amount of fuel were being added, there should be no risk of liquid overflow the first time they were filled. Nevertheless, the fill pipe delivers fuel to the tank a short distance from the inspection opening. With the bends in the fill line, the diesel would likely be foaming as it enters the tank. Indeed, during a test fill at 40L/min fuel foam climbed roughly a meter into a vertically orientated wide diameter translucent hose installed at the inspection port as a test vent. If the port were open as the tank was filled, foam over into the bilge would have been inevitable. As holes were cut through Wanderlust’s floor to gain access to the bilge we had to accept the possibility that we’d find no evidence tank leaks.
Supporting the idea that fuel came from the inspection opening was what looked like a trace residue of diesel on the top of the tank. There was also the appearance that the area had been cleaned. But things like this can be very misleading. There was no way for us to tell for certain whether there was a one-time event that left fuel in the bilge or not.
At the time of the first fill of the fuel tanks in the plant it would have been easy for us to tell if the inspection ports were opened. We were there at the time, camera in hand. But we were told we couldn’t come on board due to safety reasons. In retrospect this seems odd. Even today we view this with suspicion. Does any boat owner evacuate their boat when they fill their tanks with diesel fuel? If there were concerns about diesel fumes from the filling port entering the boat it would have been simple enough to close the wheelhouse door. Nevertheless, the workers were emphatic. They did not want us on board while the tanks were being filled. We still don’t understand why.
So as several invasive holes were cut through Wanderlust’s floor we had no certainty that we would actually find a source of fuel that could be fixed. This expensive floor-altering search could have been for nothing. It could have been that the diesel was in the bilge came from a one-time spill during the build and all that would be required was a thorough cleaning. Without access to the bilge a potentially small problem becomes a bigger issue and a greater mystery.
Once the floor was opened to the bilge the access was immediately enlightening. Through the floor cuts we could see that there was far more fuel in the bilge on the starboard side than we had imagined. A yard owner in France, an ally of the builder, had told us that Wanderlust had small inconsequential “weeps” from the tank leaving small amounts of diesel in the bilge. That did not turn out to be the case. There was far more diesel than we had ever imagined on the starboard side of the bilge. Though we now knew there was a lot of fuel in the bilge we still couldn’t tell the source.
It was time to hunt for the origins of the diesel. The first step was to clean the bilge and the sides of the tanks as much as possible. Next the tanks were filled to brim with diesel, a process that allowed us to test an alternative venting arrangement. After the tanks were topped we waited looking for diesel to appear in the bilge.
And diesel did appear.
We could see diesel coming into the bilge on both sides of the white tank. All told we counted ten different points where weeping diesel was coming from the weld seams. The leaks were all from the white fuel/propulsion tank, confirming what we had suspected based on the color of the fuel in the bilge. Though we could see around 60% of the white tank, we had better access around the red tank. The red tank containing the diesel for heating and the generator showed no signs of any leaks. That was good news.
Why was the propulsion diesel tank leaking?
By and large the fuel leaks from the white tank occurred in places where the welding access on the outside of the tank was difficult. It was difficult for us to see the leaking welds, even with the new access through the floor, just like it was difficult for the welder to make and inspect the welds in the first place when the boat was built. Still, even with the difficult access, the presence of so many fuel leaks was a surprise. We expected that the tanks would have been welded more easily from the inside, as the junctions of the steel plates would have been easily accessible. A good solid interior weld should have been more than sufficient to seal the tank. But for whatever reason it wasn’t enough.
For a couple of weeks we thought that the fuel leaks were all from faulty tank welds. We were wrong.
On further inspection we found that there was another source of the bilge diesel unrelated to the tank. There was fuel coming through a hole in the “sealed” bulkhead on the starboard side. Though we had considered the possibility of fuel coming through a hole in the bulkhead on the port side, we had not given much thought to the notion that there could have been a similar leak on the starboard side as there was no obvious source of fuel on the inside of the engine compartment.
The diesel we saw was coming through the bulkhead where the vent and fill pipes passed through. Tracing the fuel back we identified the source: Wanderlust’s white fuel fill pipe was leaking at a welded joint well into the engine compartment. This fuel dripped to the bottom of the boat, rolled forward, and pooled at the engine wall. Once the diesel level was high enough it worked its way through the holes cut in the bulkhead to allow the pipes to pass to the tank.
In terms of volume the fuel pipe leak was far more significant than the others. Indeed, when the fuel level was high enough in the tank to keep this portion of the fill line partially full, liters of fuel spilled into the bilge out per day. Because the leak was high up relative to the tank level it was self-limiting. If it were lower down it would have drained the entire contents of the fuel tank into the bilge in weeks.
Closer examination revealed that the fill pipe leak came from the bottom of a welded joint connection two sections of pipe. Sitting only a few millimeters above the bilge bottom, access to make this weld during the build and for subsequent visual inspection after was very difficult at best. And in this case there was no possibility of a secondary weld on the inside of a pipe, unlike the fuel tanks. The pipe weld had to be perfect. But it wasn’t. And once the main engine was mounted and the fuel tank was filled it was difficult to even see that this pipe joint was leaking.
The discovery of this last leak established clearly that Wanderlust’s fuel system had not been properly pressue tested. Before the fill pipe leak was discovered it was mathematically possible that a pneumatic tank test on a 1000L tank would not have been sensitive enough to show the small leaks around the edges of the tank. But this fuel pipe joint leak was not a small leak. It should have been readily apparent if pressure test was performed properly as required by law under the Recreational Craft Directive. Further, given the size of this last leak, fuel in the bilge would have been apparent in the plant and at launch if any sort of inspection focusing on the potential weak points in the system were done.
It was good news that we had identified the origins of the diesel in the bilge. But there was also bad news. We could only see 60% or so of the weld seam on the white tank. Though we counted ten leaks from the tank itself so far, there could well be more. Indeed, we had good reason to suspect that there would be more. Clearly the welder, machine or person, was having a bad day when this tank was built. The chances that the issues were confined only to the places that we could see were small.
In the end this is why pressure tests on fuel tanks are required under the regulations. We’ve been told that confident builders skip these tests. But how confident can you be about difficult welds that you can’t at the very least inspect visually with assurance? A pressure test or full inspection access around the tanks was necessary for the most basic quality insurance. Anything less was asking for problems.
Now it came to figuring out how to make the repairs.
Even with the newly cut hatches allowing access to the bilge, fixing the outside welds on the tank would be near impossible. If it were difficult for the builder to weld the outside tank seams during the build, it would be near impossible for a contortionist French welder to repair the tank’s welds. At the least a good chunk of the interior fittings, including more floor and possibly part of a bathroom, would have to be removed. Even then it wasn’t clear that effective repairs could be made from the inside of the boat.
The head of the boatyard in France suggested an alternative approach: Instead of trying to fix the tank from the interior of the barge, they would cut access through the hull’s 13 mm plate steel from the outside with the tank empty and Wanderlust was out of the water. Once there was access, the inside seam of the white tank would be entirely over welded. We had to assume, probably correctly, that all of the welds inside the tank were bad; there was little point in trying hunting down and repairing just the points that were clearly defective.
Fixing the leaking fuel fill pipe in the engine compartment required another though smaller hole through the hull for access. Otherwise this was going to be a blind weld in a super tight space. During the build, without the engine in place, this was a difficult weld. Afterwards, with the engine in the way, it would be much harder.
Cutting holes through the bottom of a basically new boat is not what one expects to have to do. But so it was.
The boatyard’s plan was a seemingly simple way to repair the fuel leaks. We figured it would take about a week to execute and then Wanderlust would be moved to the paint shed to be grit blasted and to have her bottom recoated as planned.
Once the holes were cut through the bottom of the boat we could see inside the tank. It was clear that the interior weld was defective at several points. These faults were easy to see; they should have been easily identified on inspection during the build. But they weren’t found and repaired.
Now the genesis of the “weeping” interior welds was clear. The holes in the defective interior weld allowed fuel to penetrate into the gap between the steel plates. Once between the plates, the fuel worked its way around the edges of the steel until it found its way out into the bilge at the weak points where the exterior welds were difficult to make due to lack of access. The welds on this tank were defective on the inside and out.
As with all things related to Wanderlust the repairs took much more time to execute than expected. In the end it took five weeks to fix the fuel tank.
There was some additional time added at the end of the repair so that Belzona could be applied over the freshly welded tank seams. Given time and effort involved in gaining access to the inside of the tank we were going to do everything to make sure that tank was fixed for good. Sealing the inside with Belzona, a product recommended by our paint surveyor, seemed like good insurance.
After the interior seams were over-welded and sealed, the edges of the plates of steel cut from Wanderlust’s bottom were beveled, lifted into place, and welded in. Once bottom holes in the boat were closed the white diesel tank was filled to the brim. After the fill we monitored the bilge for diesel. Months later there was no evidence of further diesel coming into the bilge. The leaks were finally fixed nearly three years after we first reported that there was diesel in the bilge to the builder. For the first time ever Wanderlust’s white tank did not allow fuel into the bilge.
After the tank repairs were completed Wanderlust headed into the paint shed to have her bottom redone. Because of the widespread underwater coating failure her bottom was going to be grit blasted anyway. Blasting the new welds from the tank repair was a minor addition.
All told Wanderlust was out of the water for 12 weeks during the summer of 2018. We lived on board for all but a few weeks of the time, leaving for short weekend trips to the region and a longer tour while she was in the paint shed.
Living on the hard in a construction zone is not ideal. It’s even less ideal when the black tank filled up, which meant the nearest functioning toilet was a half-kilometer away. It was unpleasantly inconvenient for sure. We just hoped that at the end we’d have Wanderlust’s problems fixed once and for all.