[Richard]

This will probably be a multi installment read, so get a cup of joe or whatever. Hopefully you will enjoy. 

For about five years my generator engine would periodically produce black smoke, not grey, not white, black. That is usually a sign of incomplete combustion in a Diesel engine. Because our generators are oversized for most of how we use the coaches, the engine runs at partial load and has a tendency to build up carbon in the cylinders. The carbon build up can interfere with the valves sealing, or the piston rings sealing. Both conditions lead to low compression and black smoke. Also, the injector tips can foul from the carbon build up leading to poor atomization of the fuel. That too will cause black smoke. Other potential causes are air intake restrictions, exhaust restrictions, cold engine, low fuel pressure, and improper fuel injector pump timing. Can you already tell I noodled on this a long time? 

My usual fix was to load the generator as much as possible and “burn off” the carbon deposits. This worked a few times. I also tried every magic fluid out there as a fuel additive, sea foam, marvel mystery oil, acetone, ATF, and pee from an albino raccoon. None of it really helped in the long run. 

Did I say I was smoking and really wanted to stop? Did I say it was embarrassing to enter a campground and have onlookers ask if I was on fire? Did I say that this war on smoke became extremely personal the longer it continued. 

New air filters, fuel filters, turbo, and injectors ($$$$$) did not really help. As the years go by, the black smoke is no longer intermittent, now it’s all the time. The heavier the load on the gennie the worse the smoke is. 

So a deeper dive was on the list for winter projects. Someone is going to ask, how many hours on the gennie, 6000. 

A couple of other important points or clues. The engine did not use oil, not one pint in 100 hours between changes. The engine was quick to crank, even on really cold mornings. The engine did not struggle to make 50 amps on each leg. The engine ran smoothly, well as smoothly as a tractor engine runs. 

Given all that the first diagnostic is a compression test. Well this is interesting because hooking a compression tester up to a spark plug hole is not going to work. It’s even more complicated because none, including the Yanmar, commercially available testers I could find married up to the injector hole. Maybe one is out there, I just didn’t find it. So, I made one. I took one of the old injector bodies, welded ears on it to match the injector hold down screws, and gutted the inside. I hooked that up to a high pressure gauge with a check valve and pressure relief valve in the appropriate places. This first pic is of the home made compression tester. Second pic is of the modified injector to fit the injector hole. I sealed it to the bottom the injector hole with an o ring. 

The compression readings on a cold engine that had been sitting for a couple of months were around 300. Not terrible, but the Yanmar manual says it should be 350 or above. 

The next step was to remove the one way check valve on the compression apparatii, and use air pressure to do a crude leak down test on the cylinders. So I learned a few things here. One the piston of the cylinder you are going to test should be at top dead center. Ok, got that. Two, install a regulator on the air line feed. When you apply air pressure to the top of the cylinder, the crank WILL turn. It’s a 3.5 inch diameter piston, so 100 psig is 960 lbs force. It’s a bit hard to hold the breaker bar on the crankshaft for a wimp like me. Any Hoo, 100 psig was not necessary, for as soon as I applied air to the first cylinder, I heard a lot of air escaping into the exhaust manifold. Only way that happens is if the exhaust valve is not sealing. At this point, I didn’t do any more testing, because I knew the head would have to come off.

[Richard]

The exhaust manifold bolts did not seize on me as I feared, so getting the head off was pretty straight forward. Remove the intake and exhaust manifold, get the injector lines removed, take off the valve train, and let the impact wrench take the head bolts out. 

It was pretty evident upon removing the head that cylinder 1 was not burning cleanly at all. You can see a pic of the crud on the valves. Plus the exhaust port was all gunked up with a 1/4 inch of carbon.
   

When I pulled the valves out. Here are pictures of the exhaust valve and seat in cylinder 1.
       

It’s pretty evident, that those two mating surfaces are not going to seal. The other exhaust valves were a little pitted but nothing like cylinder 1. There are lots of opinions on the interweb about what causes the valves to pit like that, but the leading contender is carbon deposits that build up on the sealing surfaces and get hammered into the seal each time the valve closes. I started by lapping the valves and seats. It worked just fine to restore the valve and seat on the intakes but I would have been lapping for a week to fix the exhaust valves. So I did a very light grind to remove the pits, and followed up with lapping. I was unsure if this would work or if I would need new exhaust valves and seats. Luckily, I didn’t have to remove too much metal, and I was right on the edge of Yanmar specification for seat width and valve recession. Here are a couple of pics of the valve and seat from above after some TLC.
       

[Richard]

At this point, the temptation was to button it back up. But………

I bought, $100 who knew?, a test stand for checking the injectors. It applies high pressure diesel to the injector so you can see the spray pattern, measure the “pop” pressure, and see if the internals are leaking. All important things to a clean burning engine. I did make some minor adjustments to the pop pressure to ensure it was in Yanmar spec, 2800 psi. By the way don’t put body parts next to the injector while doing the test or you will inject diesel under your skin. Not a good idea. 

Ok the injectors are good. But…….. the head is off, and although the cross hatching in the cylinder walls looks pretty good, I decide to pull the pistons. After all, rings not sealing correctly can also be the cause of the low compression. I did run a quick and dirty compression test again after lapping the valves but before pulling the pistons. I got about 320 psi, better than before, but not the 350 spec in the manual. I did the classic test for leaking rings by squirting oil in the cylinder and retesting. That brought the pressure up to 350. So I decided to pull the pistons, rehone the cylinder, and install new rings. 

For grins, here are two pics of the cross hatch in the cylinder. The first is the as found after 6000 hours on the motor. The second is after I rehoned the cylinder.

       

The red tape was to prevent honing debris from entering the oil and coolant passages in the block. 

The next step was the scary one for me, and ended up creating a small problem which I will detail later. To remove the honing debris, the instructions are to scrub vigorously with Dawn and hot water. Since I did this work as an in frame, the crank is still in place. I did as best I could to protect the crank and clean the journals before reassembly. But pouring hot water into an engine makes me queasy. 

All right, button it all back up. Compression is 350, but with brand new rings and cross hatching, with no time to develop a good seal. A quick squirt of oil into the cylinder brings the compression up to 400, so happy dance all around. Put the injectors in as a final step, bleed the fuel system, and it fires on third try. Woo Hoo. And NO SMOKE. After a few minutes, start loading the generator with Aquahot and water pump, and stove, and dryer till we get to over 40 amps on each leg. Still no smoke. Woo Hoo.

[Richard]

So my very best helper, @"RussWhite", leaves for the day, and 5 minutes later, white not black smoke starts pouring out of the stack. 

OH NO! Well maybe I expressed it a bit differently. But nonetheless. 

It’s almost dark, so I have to quit for the night. Needless to say my noodle was working OT to try to figure this out. 

Cranked it again in the morning with the same pattern. Engine exhaust is crystal clear until the engine runs about 15 minutes and then LOTS of white smoke is coming from the exhaust. 

My diagnostics before cranking the engine a second time were:
Turn the engine with breaker bar to see if I feel or hear anything unusual. Nothing
Open coolant cap to look for oil or diesel indicating blown or leaking head gasket. Nothing. 
Drain a sample of oil into clear container to see if coolant is in the oil. Have to wait for that to settle over nite. 
Hmmm, if it were an injector or compression problem it would manifest itself when the engine was first cranked, so I kind of eliminated that. 
Lot’s of internet chatter on what to lube new rings with to ensure good break in. The only consensus is DON’T use assembly lube like I did. Lovely. 
Now that I am confident, that nothing catastrophic is going on, I decide to let it run for a couple of hours to see if the mystery smoke will clear. After all it could be all the junk built up in the exhaust system from years of smoking. 
While it’s running, I take a ladder to the roof and let the exhaust run onto a white rag for 30 minutes to sniff it and feel it. Maybe this will give a clue. Curious enough the white smoke does not have a distinct odor. It does not smell like diesel. It does not smell like burned diesel. It does not smell like burning engine oil. In fact I don’t smell anything. And not much of anything except flecks of carbon are collecting on the white rag. But it is still smoking like crazy. And it doesn’t show any signs of getting better after two hours. 

So either I have an unknown contaminant, or the new rings are not seating. 

Sigh. 

I then look at the oil sample collected a couple of hours ago. Something has settled to the bottom. But the sample is small, and the settled liquid sample is too small to really evaluate. It is coolant, is it water, is it kryptonite? Who knows. 

But this gets me thinking. It could be water. After all, you did pour gallons of water into the open engine block, perhaps it hid somewhere and has contaminated the oil. If there is water in oil, it will emulsify, and water is being smeared on the cylinder walls with the lubricating oil. The film of water is turning into steam. A ha, that’s why it has no smell. 

I had planned on doing an oil change a few hours after break in anyway, so I went ahead and change oil and filter. I collected a much larger sample of the old oil this time. The first pic is a shot of the clear container while holding it at angle to produce a low spot. The second pic is a closeup of the same thing. Looks like about a teaspoon of water has settled out of the oil.
       

Cranked the gen with the fresh oil. NO SMOKE. Loaded it up with amperage. NO SMOKE.

What an adventure, but I won’t be embarrassed in campgrounds about the coach being on fire. 

I would like to thank @bestgenman for his excellent guidance and sage wisdom on this project. 
I would like to thank @Russwhite for helping me put the thing back together, there were times when five hands were needed. 

So, what should you take away from this, besides that dude is bonkers. 
Take the time, maybe a couple of times per month if you are using the coach regularly, to turn on everything till you pull some serious amps, and let the motor burn the carbon out of it. It’s nice that Newell puts a monster genset in there, but the light use we give it is not good for the Diesel engine. 

And should you need a compression tester for the Yanmar, or a Flexhone for the cylinders, I know where one can be obtained.

[Fulltiming]

You are amazing Richard. I don't think there is anything you won't undertake once it gets to the point where it is bothering you!!!

Good for you!

[folivier]

Like we used to say in the 70's: Run it hard to burn the carbon out!
Great job.

[BusNit]

Great Job! You had me laughing at Albino raccoon pee. You sound like me. If something is bothering you, there is no stopping a runaway freight train of a mind from trying to figure it out. Interesting that the little bit of water was enough to smoke so bad. Look at the bright side. You steam cleaned the internals!

[hbens]

Thanks for posting Richard. And more so, thanks for making this a riveting read!

[whatsnewell]

Richard, as always, a great read, and a lot of useful info. You, my friend, are fearless!!

[Richard]

I was privately and politely asked about the water theory, and how that could have happened. I felt it was a good challenge, for after all that is a zebra not a horse explanation of the smoke. Some of you may have the same questions that were asked privately.

How do you know it was water? I did two things that I did not write about in the novella installation series. The first is that I put the large oil sample in the freezer overnight. The clear liquid that settled froze hard. I also took the other two possible contaminants, Permatex assembly lube and Lubriplate assembly grease, and dissolved them into clean motor oil to see if they would separate over night. They did not precipitate.

Where was all that water hiding? This is an idea, not proven. On the bottom of the engine block is a 1 inch thick casting with webs that is bolted to the bottom of the block. The casting actually obscures the main bearings. IF, and that’s and IF, the tops of the connecting webs are open, it would be a nice collection point for the water from above.

So how does water in the oil result in steam coming out the exhaust? This is where my Mama’s hard earned money that sent me to Chem Engineering school comes into play. The first step in cleaning up the bores was to wash them down with a Dawn/water mix. As all of us dish washers know that Dawn cuts grease and oil. So, if the water left in the engine has any Dawn left in the catch pockets, then the Dawn helps the water emulsify into the oil. Water doesn’t actually dissolve into oil, it breaks into very small droplets and is suspended. The more surfactant, Dawn, in the water will result in smaller and smaller suspended droplets. When the piston moves up and down, a small amount of oil is smeared onto the cylinder walls. If there is water in that oil, some of it could remain on the cylinder wall. As the engine reaches operating temperature, the water will evaporate into steam. Without the Dawn factor, I don’t think this would have happened.

Now, all of that may be Zebra hooey. For sure, the oil I drained was thinned by something, the viscosity is thinner than new oil, much thinner. A change of oil, changed the smoking behavior instantly.