Fuel System Issues

[ralay]

Hey Folks, 

The winter project most on my mind is overhauling the fuel system of our Yanmar 3QM30.  Hopefully, a diagram of the fuel system can be found here.  https://drive.google.com/file/d/0B9yvhbnP3FSDNVJoLXVmS2dQTVE/view?usp=sharing

Over the course of our cruise, we experienced occasional (every ~100h or so) engine trouble in which the engine would rapidly throttle down and die.  The problem has occurred with both fuel tanks, and switching between tanks has never fixed it, so we assume the problem is not a clogged pickup, etc.  We have seen a black precipitate in the bowl of the primary filter and assumed that the most likely cause was filters clogged by bacterial or chemical fuel contamination.  Indeed, swapping the filters and bleeding the low pressure side of the system has brought the engine back to life every time. 

However, in each case, the old filters have looked pristine.  Also, I've sometimes had problems getting fuel to flow even after the filters have been swapped.  The tanks are above the filter bowls and engine AND we have an inline electric pump for easy bleeding.  In the times when I've been unable to get fuel at the bleed points after changing the filters, I've pulled off the secondary filter bowl and inspected the fuel line from the pumps.  There was not a drop of fuel coming through, even with the electric pump pumping away.  After enough wiggling, cursing, and pumping with the electric and lift pump, often the fuel would finally come through in one messy burst and then flow fine.  Also, occasionally, to rekindle the engine, we would need to leave the electric pump running while the engine ran.  Turning off the electric pump would cause it to slowly idle down and die.  Strangely, this effect would always work itself out within a day.  The next morning, for instance, the engine would start up fine without the electric pump and run another 100h or more without hiccup.  These complications have made me wonder whether the blockage is somewhere other than the filter elements and that the engine is cured once we finally apply enough suction to dislodge the obstruction.  Where the heck else would the obstruction be?  What would the obstruction be?  Of course, diagnosis is compounded by the fact that the engine runs like a champ 95% of the time.  It's been months since the last time it died, but who knows when it will chose to do so again.  I can think of three possible problems, and would be interested to hear other's insights.

1. Contaminated fuel is slowly clogging the fuel filters.
2. Contaminated fuel is obstructing some other component (ex~ one of the three way selector valves?  some internal part of the Racor filter? some part of the electric pump?)
3. The problem is an air leak rather than a clog and it's the bleeding rather than the filter changes that are fixing the problem.

I would be interested to hear anyone else's votes on what they think is most likely.  Specifically, I'd be interested if anyone has ever had/heard of an obstructed selector valve/fuel hose/inline pump/filter housing.  Our fuel hoses and fittings are 3/8" to the engine and 5/16" on the engine.  The contaminants I have drained from the filter housing have always been very fine.  It's hard to imagine a 3/8" chunk somewhere in there, but I'm prepared to be surprised. 

I have several ideas for what to do about the problem, but I'd be interested to hear what other people think.

1. I bought a vacuum gauge.  I will check for air leaks by running the engine, cutting off the fuel at the tank, letting the engine draw a vacuum and die, then see if the fuel system will hold that vacuum. 
2. I can run the engine and wait to see if the gauge is slowly (loaded filter) or rapidly (abrupt clog) drawing a vacuum.  Alternately, I can cycle the fuel back to the tank with some extra fuel hose and circulate it with the electric pump so the engine doesn't need to be run at all.  Hopefully, I'll get a clog that I can track down at the dock.
3.  I'd like to hear anyone's ideas on DIY fuel polishing or tank cleaning.  Unfortunately, our tanks don't have any access ports and are directly under our newly rebuilt deck.  We're not going to cut the deck off to get them out and we're not going to cut them up and rip them out, because they're still in good repair.  We make an effort to keep them topped off and use biocide.  It's possible the contamination is chemical precipitates from fuel sitting in the tanks while the boat was for sale or from who knows what Nicaraguan fuel dock. 
4. Pull absolutely everything apart to look for gummed up components.

Ideas?

[Norman]

I have limited experience with diesel, but have learned a few things from the repeated failures on one of them.

A fuel blockage anywhere in the system will stop fuel flow, and test the tightness of all fittings and joints from there on.

A valve or fitting that is fuel tight is not necessarily tight for the smaller air molecule.

When the flow stops, the powerful suction of the injector pump pulls a vacuum, causing air leaks into the system, and when the engine finally shuts down, changing sources of fuel does not recover engine function because air is now in the pump and injectors.

It would seem probable that you have a few large floating pieces of debris or organic matter that floats to the fuel line, covers it, and is sucked tight.  When you do your filter drill, the back and forth sloshing of fuel washes the debris away, and you are good to go.

We could see into the fuel tank thru a 1.5 inch pipe, and at night with a strong, small flash light, saw half inch ?lily pads? of black organic growth floating around, mostly by luck, as there was not much of it there.  We thrashed it with a long flexible bristle brush to break it up, connected an electric pump and a filter with a clear bowl to the fuel line after the fuel valve, and circulated fuel.  When the clear bowl had a fair amount of debris, we stopped and drained the bowl into a large glass jar with a lid.  For your boat, this would have to be done on each tank.  We had 4 or 5 failures in 200 miles, but as we were positioning for a major race, then returning to home port after the race, effective trouble shooting did not occur until we were in home port with access to tools and equipment suitable for the problem.  We also loaded the fuel with biocide after the cleanup.  The owner had previously used biocide, but not until he started having trouble, and that was too late.  Biocide prevents, but does not remove growths.  We properly disposed of our collection of dirty fuel at a gas station.

If you cannot ?thrash? the tanks, and alternative would be to use a large diameter suction hose down the filler, up to a suitable electric pump, filter with a clear bowl, and then back down the filler thru a much smaller hose.  This would work best with a modest amount of fuel, so swirling motion of the returning fuel would bring any debris to the pickup hose sooner.

A simpler temporary solution would be install a 5 gallon ?day tank? before the electric fuel pump, with your vacuum gauge on it.  Check it frequently for developing vacuum, and switch tanks if any pressure drop is seen.  The likelihood of any debris being in place on the other tank with no flow out is near zero!  Counter intuitively, a gallon of air in the day tank would be a good thing, so fuel could be drawn from it for several gallons as the air expanded, before the vacuum would be likely to pull air into the system elsewhere.  4 gallons would be useable, and if you are properly monitoring the vacuum gauge, none of your intentional air would reach the injector system.

I hope that this is helpful, and you enjoy returning to cruising!

Norman

[SeaHusky]

Could there be a filter on the end of the fuel line inside the tank?
I have read about this being a problem on old diesel tanks that there is a sieve type filter inside the tank that clogges up after many years and eludes inspection.

[Godot]

Intermittent problems are the worst. They are difficult to diagnose and leave you feeling perpetually uneasy. I'm not a diesel mechanic; but having dealt with the fuel system on my Yanmar 2GM extensively on this summer's cruise perhaps I have a shot at it...

My first thought would be the fuel pickup screen (if installed); but it seems unlikely with two tanks that an intermittent problem here would also exist while switching tanks.

However, if a tank contaminate gets sucked up into the valve before the primary fuel filters, it seems like there is a possibility it could get stuck there.  If a clog exists at that valve, and somehow survives exercising the valve, it may be possible that switching primaries would not help. Maybe a way to test for this is to have each tank go to a separate Primary and eliminate the selector valve. The system gets a bit simplified as you will have completely redundant fuel systems up to the valve before the electric fuel pump. At the least it could be useful for diagnostics.

I guess it is possible that there is an air leak; but I wouldn't expect the symptom: "After enough wiggling, cursing, and pumping with the electric and lift pump, often the fuel would finally come through in one messy burst and then flow fine." This really does make it sound like an obstruction and not an air leak to me.

How old are the fuel lines? It might be worth replacing them if they are more than a few years old. It is at least conceivable (although probably not likely) that they are cracking inside and periodically dropping pieces of rubber into your fuel stream.

As far as polishing the fuel goes, I think it would be faster, easier, and probably cheaper to just pump the tanks dry. I've never done this, and haven't read much on it so I might be talking out my posterior; but, if you have a spare electric pump (or if you don't mind borrowing your system booster pump), run your tanks down to just a few gallons (no need to throw away money, after all), and use the pump to cycle the fuel in the tank (suck it up through the intake, and return the fuel to the tank at first) to mix up any contaminants. If you can get a dowel or something into the tank to help stir it up all the better. After doing this for awhile, suck the fuel into a jerry jug or something. I'd be tempted to add several gallons of fresh diesel to the tank and do the mix it up and pump it out process again. If contaminates keep coming out I would keep repeating the process. Even though this will entail abandoning some fuel; I suspect it is much cheaper and easier than building a polishing system.

[ralay]

Since everyone is pining for content, I figured I'd post an update of the troubleshooting I've tried to do with our fuel system over the winter.

Since we suspected we might have a fuel contamination issue, I decided to add a vacuum gauge to our fuel system and polish our tanks to see if I could catch a clog or load up our filters.  Since our engine doesn't have any lines returned to the tank, I bought a length of fuel hose that would stretch from the engine room out the companionway and into both deck fills.  I attached my spare hose to the secondary fuel filter and used our inline electric pump to circulate the fuel through the primary and secondary filters and vacuum gauge and then return it to the tank from whence it came.  I ran this set up ~8h on each of our 30 gal tanks.  I wasn't lucky enough to get any clogs I could diagnose and the vacuum gauge was still near its lowest reading even though the primary filter has been in use for at least 100h.  So we topped off and treated the tanks.

Next I dismantled the entire fuel system from the tank on/off valves to the engine.  I disassembled both our Racor housings entirely.  Both had a good bit of chocolate pudding in the bowls, though nothing that seemed close to causing an obstruction.  The bowl residue was thick enough that they couldn't be drained through the petcocks, so the cleaning was still important.  Now that they are squeaky clean, we can do a better job of monitoring and draining any accumulating water.  I also dismantled all the fittings on the selector valves in the system.  I found one that had been incorrectly fitting with a flare adapter rather than an NPT adapter and fit loosely.  I replaced it with a correctly threaded piece and retaped and tightened all the threaded fittings.  I also disassembled the engine's fuel pump, including taking out the incredibly tiny one way valves for cleaning and inspecting the diaphragm.  Once reassembled, fuel would feed freely via gravity from the tanks all the way to the secondary filter at the high point of the engine. 

I was still worried that we had gunk in the tanks that wasn't being picked up by our the slow flow of my DIY polishing set up.  I decided to pump a sample directly from the bottom of each tank.  I used my spare hose and a electric inline pump to get the sample.  The hose was so flexible, I had a hard time telling whether the tip was at the very bottom of the tank, so I wrapped it with wire to stiffen it and allow me to hear the metal wire tap tapping on the bottom of the aluminum tanks.  Dragging the hose along the tank bottoms, I wasn't able to get any exciting chunks at all, just a few drops of brownish water on the bottom of my sample jar.  It was a great demonstration of why one should never put an electric pump before a water separating filter.  The pump broke the water into infinitesimally tiny drops and the samples looked pink, opaque emulsion.  It took hours for the tiny amount of water in the sample to settle to the bottom.  No Racor filter would have been able to separate that water once it had been fragmented.  If one was retrofitting an electric pump into an existing fuel system, it would be easy to make such a mistake with possibly terrible consequences at the injectors.

So the tanks (or at least the section I sampled) seem clean except for what is probably some condensation.  I trimmed the ends of all the hoses which are reanchored with tight hose clamps.  I also ordered new crush washers for all the banjo nuts.  Lastly, I read that electric pumps themselves can be a source of clogs.  Supposedly, most models contain an uncleanable screen and tiny bypass passages for when the pump is not running.  It's possible that the inline pump itself was clogging.  Some folks in other forums suggested never installing one without a bypass route.  We're going without it for now.  While I was at it, I also ordered a replacement for a nearly stripped bleed screw.  I've also made a habit of keeping an "oh sh*t" bag next to the engine room which contains two filters and the appropriate wrenches for the bleed points.  Wouldn't be a bad idea to put a spare impeller and the right size allen wrench for the pump cover in as well.  That's what loose tool bins in pawn shops are good for.  ;D

All I can say is, the next time the engine dies, I'm really going to be out of ideas.