Fuel in Coolant, Cummins TSB170004

[Hamish]

Has anyone had fuel in their coolant and tried upgrading the O-rings on the bottom of their injectors as per Cummins TSB 170004 (for CM870)? This is a relatively new TSB, last updated 07-May-2019 and originally created in 2017. Few engine repair shops here have heard of it. I stumbled across it on the internet - but still can not find it on QuickServe, hmmm... Cummins did have the special O-rings in stock and I installed them and have absolutely no fuel in my coolant after working the truck for 30 hours or so.

Engine was de-mandated around 600,000km, inframed for burning oil at 880,000km (off-road and pulling super B's, 140000 lbs), and currently has around 430,000km on the inframe. Was a reman Cummins head installed during the inframe, aftermarket new were not as popular then.

Before the O-ring replacement I had to syphon about 1 inch of fuel off of the surge tank after each 15 hours of running. I ran it like this for a few weeks while I tried to figure out what to do. I discovered, prior to the repair, that 90% of the fuel stayed in the 3 gallon surge tank, while 10 % was spread throughout the other 10 gallons of coolant in the system, as determined by draining it all and skimming it off the top. No coolant in oil. After the repair I did drain, clean with Restore, flush with water, replace reservoir, cap, and coolant, and changing all hoses now.

Cost was a tiny fraction of what it would be to replace the head and so far appears to be a viable workaround for failed injector cups. As I understand it the cups and bottom injector O-rings both have to fail for fuel to get into coolant. It was less likely to get coolant into the fuel because the fuel destroyed the rad cap rubber and turned it into a zero-pressure cooling system, which was fortunate. First tested the repair in the shop running the coolant up to 210 DegF with 25 psi on the cooling system (via air hose/regulator). Running fine on the road now with a new rad cap.

[Signature620]

In my experience, those blue orings are only a short term repair for this problem. Certainly worth a try but be prepared to change the head in the future.

[Yauchout]

it worked for mine didn't leak for over 200K miles until I was forced to inframe due to two broken piston rings #5 and #6 like signature said its just a bandage though. i highly suggest if you are going to try it to just purchase the special tool and do it yourself because if you pay the heaps of money a Cummins shop wants and and it doesn't work you were very close to the price of a head. at least that's the case over here in the states

[Hamish]

Thanks for letting me know how long yours lasted for. Mine is a CM870, a newer model engine is modified with a special ring and tool, a different TSB, 170002 if I recall. I didn't need any special tools to do this one. Didn't take that much longer than just doing an overhead set. I think they changed the o-ring composition to something that withstands coolant as originals weren't intended to be exposed to coolant if the cups would have held out. My o-ring shop tested the composition of the original o-rings, likely Viton, and I'm guessing the blue upgraded o-rings are low-temp Viton which is more resistant to coolant, or maybe something even fancier.

[Hamish]

Quick update on the blue O-ring 'fix'. It lasted about 2000 hrs, then fuel in the coolant again. Didn't pooch any injectors this time likely due to low pressure rad cap I had retrofitted when installing the blue O-rings.

I would think that anytime coolant gets past the cup and into the cavity surrounding the bottom end of the injector it is only a matter of time till it creates a steam bomb, or hydraulic force, that blows past the bottom injector O-ring destroying it a little more each time. No o-ring can take this, so I just want back to the original green ones this time. Bottom line is that fluids have to be stopped before getting down into the cavity, not afterward on their way up at which point the cup has likely re-sealed like a one way valve.

I still only have 500k on the rebuild which was done with a reman head (before new heads were readily available) so not feeling like another inframe is due yet.

This time I choose additional o-rings to seal against the upper face of the injector cup and the wall of the bore just above the injector cup. I choose a 121 size EPDM against the lip of the cup, backed by a square cross sectioned nitrile 026. Obviously this is not an engineered O-ring gland, so compromise was made (ie. excessive stretch of the 121 to reduce cross section size). Squeeze had to be checked with each cylinder as most cups were 72mm down, but one was 73 on one side, and 72.5 on the other... yes the one that leaked the most. Squeeze was checked by monitoring how much the injector retaining bolt turned from initial contact to tight with only the additional o-rings installed - just over 3/4 of a turn. This vs. no o-rings at all and just over 1/4 of a turn. I did not want to exceed half a turn of squeeze (1.5mm between bolt threads) because this would not leave room within the 'gland' for 7% expansion at running temp. I imagine packing this area too tightly with rubber could lift the injector out of the head at temp, so definitely some risk with this injector by injector custom fitting attempt. On the offending cylinder I had to add a backup ring between the square and round o-rings and hopefully the happy medium fit works for this crooked cup during 7% expansion at temp and 3% contraction sitting outside in winter.

Worst case I lost a day and have to rebuild the engine anyway. Best case the extra o-rings solve the leaking cup problem for another 1.1 million km when the next inframe is due...

[JMBT]

Hamish, I am really curious as to whether your “custom” fit works. Please update this post with your results… thanks

[Rawze]

Injector cup repair is usually only a 50/50 shot at best in a cummins head from what most people have seen. It sounds to me at this point that you are fighting against the inevitable, unfortunately.

[marek4792]

Not meant to be an endorsement, but I've seen this floating around on the interwebs awhile back.
https://industrialinjection.com/products...or-cup-kit
I wonder if this maybe a possible solution, basically looks like a threaded injector cup opposed to a press fit. Thoughts?

[tree98]

Pretty pricey for a shot in the dark.......