Few of us ever look inside our engines to see how they’re faring, internally speaking. As the miles pile on, it’s common for ugly carbon deposits to build up, and they can at times get bad enough to rob the engine of performance. But how do you clean carbon deposits inside the cylinders themselves?
As it turns out, there’s an old hot rod trick for doing just that. It involves introducing water into the combustion chamber, which sounds bonkers if you’ve never tried it before.
It’s a popular shadetree hack, passed down from one wrencher to another over the years. It’s also a great way to brick your engine in seconds flat if you do it wrong. Let’s look at what’s going on here.
Water, Water, Everywhere
Let’s start with the why. The process of combustion inside an engine tends to create carbon deposits over time, either from the incomplete combustion of fuel or the burning of oil. As clean as modern cars run, most will still create some carbon deposits over time. If they build up too severely, they can impact the engine in a number of ways. They can interfere with the seating of the valves, they can reduce flow into the engine, and they can also gum up the rings such that they no longer seal properly.
To remove these deposits, the concept is simple. You want to introduce water into the combustion chambers of the engine. But how to do it? Well, carefully is the way if you don’t want to incur terminal engine damage.
Ideally, you want to introduce the water in a relatively fine mist. This can be done by spraying it into the throttle body or carburetor in the absence of an air filter. Alternatively, a vacuum hose can be pulled off to act as a point to introduce water into the intake. The water will be carried along by the air stream, with ideally some of the droplets remaining liquid until they pass the intake valves. The engine ideally run up to operating temperature before introducing the water, to give it the best chance of evaporating the injected fluid.
However, for the sake of ease, many people take a more dangerous route. You can achieve similar effects by simply pouring water from a cup into the intake. Others unhook a vacuum line and use the engine vacuum to try and suck water directly into the engine. However, go via these routes, and you risk tragedy.
See, if you pour too much water, you risk hydrolocking the engine. This is where the piston cannot reach top dead center because the cylinder has filled with too much water, which is incompressible. The water stops the upward travel piston, which is being pushed upwards with great force by the crankshaft. The usual result is a bent or snapped connecting rod and lots of collateral damage. A full engine rebuild is required in this case, if the engine is salvageable at all.
As seen in the video below from Kreosan, you can also expect a great deal of white smoke to come out of the exhaust. But that’s not smoke. It’s steam! Steam from all the water that you squirted into the engine.
Just about every major car channel has done a video on decarbonizing with water in the engine. Sadly, few actually use a borescope to verify whether their work had any effect.
The pouring method can be okay if you’re very careful about how much water you pass into the engine. You want to limit it to teaspoons at a time in most cases, as the average combustion chamber volume is on the order of a small shot glass or less. It’s okay to use a vacuum line as an injection point for water, too, as long as it is used as a port for injecting water, not a tube to suck it up wholesale from a vessel. Controlling the amount of fluid entering the engine is key.
It’s also a good idea to change your oil after this procedure. That’s because it’s likely some of the water you’ve squirted into the engine has gotten into the oil.
But what is the mechanism behind water’s cleaning action in this scenario? Filming in a running engine’s combustion chamber isn’t really possible, so it’s hard to say for sure. But the leading theory is simple. Water entering the cylinder is absorbed into the carbon deposits on the piston. In short order, the heat inside the engine causes it to quickly boil into steam. The liquid’s rapid expansion into a gas helps break up the carbon deposits, freeing them to be ejected via the exhaust.
Indeed, it’s worth noting that head gasket failures often result in a similar cleaning effect that is well known to mechanics. Such a failure often releases coolant into the cylinders, where it boils off during the combustion process. Often, upon disassembling an engine with a head gasket failure, the area of failure is obvious. This is because cylinders with coolant ingress typically appear sparkling clean and new, with minimal to no carbon buildup.
Does it really work, though? Well, hunt around on the Internet, and you’ll find results are positive but mixed. Some run the water, and borescope their cylinders, finding much of the carbon has been removed. They’ll swear by it. Others find little result and consider it unimpressive. The method definitely seems to work to some degree. The key seems to be getting the water where it needs to go. If it’s not reaching all the pistons in your engine, then it won’t be able to clean them, either.
In the case of a multi-cylinder engine, this can be more difficult than you might think. If you’re simply pouring water into a random vacuum port, you might be delivering more water to certain cylinders than others. Some might not be getting any at all. This is where an atomized spray of water into the throttle body can be more effective, as the intake after the throttle body is naturally designed to feed air to all cylinders. Still, in some cases, flow effects inside the intake and inside the cylinder might stop denser water particles from getting everywhere you need.
It’s also worth noting that if you’re spraying water in too fine a mist, most of it might be evaporating in the intake charge before the liquid water can get to the carbon deposits. Then it wouldn’t be doing a whole lot to clean off carbon at all.
It’s also worth noting that this method won’t do much for carbon buildup on the intake valves. This is a common problem with direct injection engines. Unlike port injection engines before them, they don’t have a regular squirt of fuel on the back of the valve to wash off any buildup. Thus, they can get covered in sticky carbon which causes all sorts of problems. In any case, the liquid water method doesn’t seem to do much to remove buildup on the intake valves They’re not exposed to the higher temperatures and pressures inside the combustion chamber which helps turn the liquid water to steam.
If your engine is running fine, it’s hard to imagine this being worth the bother, carbon deposits or not. If you’ve got an older car that’s maybe running a little poorly and choking itself up, you might be considering trying this. If so, it’s worth getting a borescope so you can poke it down the spark plug holes and see if it’s actually working for you. Most of all, though, you’ll want to be careful. Slip up with the water, and you can kill your engine in seconds. Maybe best not to fiddle, hmm?
Image credits: via YouTube screenshot
I once owned a BMW E39 528i. Wonderful car…except for one flaw in the construction:
The brake booster sits below the cowl below the windscreen. if the drains are plugged, the booster goes diving…and then the engine will start sucking water from it via the vacuum line…
To make it short: it sucked in water, the engine started blowing lots of smoke and then stopped working. I tried to get oout as much water as possible from the air intake…and it started running again. Still smoking and initially just in limp mode…but a day later it ran like new.
Having ridden a Suzuki RG250WE-1 2-stroke road bike as my first bike, the 2-stroke equivalent of this is that carbon deposits gradually build up in the expansion chamber in the exhaust and start to make the engine run rough because the ‘reversion effect’ is not working 100% (2 strokes with expansion chambers use the rapid narrowing of the outlet end of the expansion chamber to bounce the pressure pulses of the exhaust back to the exhaust port in time to ‘trap’ the next incoming charge of air and fuel in the cylinder ready for the spark).
The ‘fix’ for this is to take the bike out on a freeway and run it at high speed and load for 10 minutes or so to get the exhaust hot enough to loosen and burn off the carbon – the first time I did this at night I got a fright when I looked over my shoulder and saw 6 foot long jets of burning carbon sparks pouring out of the exhausts!
The only problem with this is that the symptoms of the exhausts ‘carboning up’ are identical to the symptoms of the oil injection system failing (in my case when the cable controlling the metering valve broke and the pump was only injecting enough oil to lube the engine at idle speeds), and a high speed run to clear the exhaust resulted in the engine seizing at 135km/h on the freeway, leaving a long skid mark before I grabbed the clutch lever (on a 2 stroke you should ALWAYS be prepared with your hand on the clutch lever!)
I could never bring myself to trust oil injection in 2 strokes so I always just ran premix. It wasn’t like I ever took them on road trips or anything.
My buddies thought I was anal for premixing my rx7 in the 70’s. A mechanic friend recommended two stroke oil. Never had any seal issues over 220k miles of abuse/ use. I did the same with my Rd350 with top ends lasting 15k, an improvement of 5k miles.
There is some science behind the ‘Italian tune up’.
As an aside, if you inject the correct amount of water you can get more power because the water expands into steam more than dry air, and with modern control systems in the abstract you could get much better mpg because you can replace some or all of the cooling provided by the cooling system. A few years ago, there were a couple of projects that were trying to do this, have not heard much in the last decade, but this was once promising avenue to improve efficiency.
There has been a BMW model with water injection (from the factory), but if I’m not mistaken, they only did it during a few years before giving up on it (it was probably somewhere between 2005-2015). I remember Top Gear (Clarkson) testing the car, they did not find it particularly useful.
This reminds me of a hilarious story of a destroyed engine.
My uncle and cousin were doing some sort of fix/tuning on a carbureted car. In the process they had been toying with dumping gas down the carb and using started fluid to help get the car to fire.
The car finally turns over, but the timing is off or something so it has a backfire (with no air filter on). Some of the gas/starter fluid that had been splattered around the engine bay caught fire which would probably be harmless if you just let the droplets burn out.
My uncle, being a wise experienced mechanic, immediately started to turn around and tell the guy in the driver seat to turn the engine off so they could let the flames burn out or get a blanket to smother them.
My cousin was less wise and more panicked, grabbed his gatorade bottle and started dumping it on the flames to try and douse them. A decent amount went right into the intake and hydrolocked the engine. So in a few seconds “a quick carb tune” job turned into “a not so quick engine rebuild/replacement”
I was still in high school and helping my little brother change the fuel pump on his late 70s beat-up 4-door Chevy Nova. We were almost done, it got late and the sun was going down while I tried to get the fuel hose re-attached, so I went inside to grab a cold one and take a leak. I told him to move the car under the street light. Came back out to absolute chaos because he didn’t push it, he got in and turned the ignition and the unattached fuel hose sprayed the entire engine compartment with gas, which promptly ignited. Shenanigans, and the next project became replacing everything plastic/rubber that melted.
That’s similar to how Jesse Pinkman accidentally destroyed the generator when it caught fire (in the desert) ????
“Brawndo has Electrolytes! It’s got what engines crave.”
Shoulda used metric water
Used Seafoam in my ’84 Subaru once. Got on the highway after and the resulting exhaust cloud cleared out all traffic behind me. Legit couldn’t see past the rear glass for maybe 5 miles. Seemed to idle a little smoother after.
Shit, now that I think more about it, I’m pretty sure I put it in the gas tank instead of the intake, though maybe both? IDK, this was about 30 years ago.
Both are legitimate ways to apply seafoam, the tank method is usually more for if you want to clean the carb too.
I tried that method one time in an old Datsun and while it seemed to help the carb, I started getting way more smoke/blowby in the engine. My theory is there was actually some carbon helping seal the pistons/cylinders and the “fix” actually kinda made the engine worse.
A variant of this that I encountered in the 80s was using brake fluid instead of water. It did produce an amazing amount of smoke. Or steam. No idea, really. But the “theory”, as it were, was presented the same.
Man I can only imagine how “healthy” it is to inhale burning brake fluid smoke haha
The proper way to do it is with an old glass Coke bottle. Any other tool is inferior ;).
The way we used to do it water would not make it to the cylinder in liquid form. You only do it on a engine that is up to temp and with an old school V8 that had a exhaust cross over passage meant that the water vaporized in the intake, both due to the temp and the vacuum which lowers the boiling point.
Just ordered water injection and a continental kit from J C Whitney.
I did a DIY water injection kit based on, I think a turbo Mirage website in 2005 or so on my JDM swapped MR2. It was boost referenced, used a high pressure pump and accumulator. I certainly felt high tech AF back in the day.
“water, which is incompressible.”
Water is much more compressible than aluminium, but much, much less so than air.
We need a word that means “as close to incompressible as make no difference in this situation”. Blurting out that “oil is incompressible” in front of other engineers is still one of the most embarrassing moments of my career, especially as I’d spent a good chunk of that career calculating how much I’d compressed bits of steel and aluminium.
Apologies for this trauma induced pedantry.
How about minimally compressible?
Works for me, not getting a flashback or anything.
Thank you.
Was about to say the same thing. Everything is compressible, the only question is by how much, relative to the other materials in question? Water is roughly three orders of magnitude less compressible than air, but two orders of magnitude more compressible than steel. This happens to nicely illustrate what happens when you hydrolock the cylinder- your slug of water is only compressing about 1/1000th the amount your air is, and your steel walls of the cylinder only 1/100th of even that, so to effective approximation the piston is hitting a solid wall, with predictably disastrous consequences.
You’ve at least helped me crystallize the notion that there are different levels to pedantry! (I mean this kindly, truly.) I was reading an article about graphene sheets and there was at least one knucklehead popping up to bitch that it was described as two dimensional when *obviously* atoms are three dimensional. Like come on.
Someone tried to snipe me about using the saying “even a broken clock is right twice a day” by saying “No, it’s wrong an infinite number of other times. Math”. It was dumb to have to break it down, explaining that “the saying isn’t “a broken clock is practically accurate the duration of two Planck times per day and mathematically inaccurate an infinite number of times per day.” We have sayings and axioms and shortcuts for a reason
I’d say it’s a reasonable if imperfect shorthand to say water is incompressible rather than “in its liquid phase at terrestrial conditions water is extremely difficult to compress, but it is not incompressible.”
In the UK we have some horrible not-really-legal numberplate options, and for decades one of them has been a 3D effect font. For a few years now people have been selling plates which have actual 3D letters glued on them, but as “3D plates” were already a thing they are calling them…
…4D plates.
I’d like to thank you for how much I hate this.
Germans have to have a word for it!
I bet it’s a really long word.
perhaps “not compressible by a standard ICE engine?”
But it is compressible by an ICE engine. It’s compressible by any force at all. All water, all matter other than wacky high-end physics stuff, is basically springy.
Perhaps the compressive forces required exceeds the Euler’s critical force for the connecting rod?
I did this on my old Buick with a 3800. It seemed to help that old school engine. I never did it to my old Cruze since I was able to run Top Tier gas in it almost exclusively the last 5-6 years I had it.
I had a Grand Am that blew its lower intake manifold gasket. Ended up dumping the entire contents of the cooling system through a cylinder in about 15 seconds. No hydraulic lock but a lot of steam. Seems it takes quite a bit of liquid to lock it up, at least if you’re on the throttle at all.
GM V6’s and intake manifold gaskets, a match made in heaven. At least those cockroaches don’t really give a shit if they drink some coolant, slap a new gasket on there any hit the road again.
when it’s as catastrophic as what I encountered a warped head becomes an issue. But, otherwise, it didn’t seem to care.
1) I’m addicted to Project Farm
2) That combustion chamber looks like balls
Very impressive!
Barry Zuckercorn, is that you?
Back in the mid-80’s, I purchased an Edelbrock kit that wasn’t much more than a washer fluid container with 2 “ports” – one connected to a vacuum line, the other vented a fish tank aerator into the water (washer fluid was recommended, so it wouldn’t freeze in the winter). When under vacuum, the aerator would create bubbles, with the vapor being sucked into the engine.
Only thing I ever noticed, was that the engine ran cooler, and since the radiator was pretty clogged, I felt like I accomplished something
This is why we only pour Pepsi (no Coke) into my engines. YMMV
Project farm is an awesome channel cannot recommend him enough. And as soon as I saw this article I was just like why wouldn’t you just get spray seafoam? So much better then water and also much less chance of killing your engine.
Just as easy to hydrolock you engine with Seafoam as it is with water, the big difference however is that water is much much cheaper.
True I guess if the spray can some how fails or you spray improperly.
We regularly did this at International Navistar in the MaxxForce engine days. They ran an insanely high duty cycle (80%+) on the EGR in attempts to meet EPA standards sans DEF.
Well, as you might expect, between the closed loop crankcase ventilation and the EGR, the turbos and intakes would gum up insanely bad.
The solution? Replace the intake elbow with one that has a nozzle, hook up a water/cleaning solution sprayer that’s pressurized at 60psi, run the engine up to around 1800-2000rpm, and then feed FIVE GALLONS of liquid into the engine over the course of a few minutes.
It was a factory-approved solution, but I was always scared shitless doing it. I was just waiting to hear the sounds of the incompressible trying to be compressed.
When I was young, working at a quick-lube in the 80s, we offered a “fuel injection cleaning”. We hooked a little bottle of “stuff” (I think it was a petroleum distillate) to a vacuum hose and slowly opened the valve. Gray smoke would pour out of the exhaust in such quantities that we often had to explain that this was a good thing.
Was it a BS service invented just to fleece customers? Probably.
Water injection is a thing. Helps cool down the intake charge, and effectively “increase” octane (by making the charge harder to detonate).
I wonder if engines running water/methanol(or alcohol) injection are cleaner than the average.
I’ve always wondered about this one. German engineers used methanol/water injection (MW50) as well as nitrous oxide injection (GM-1) on high-performance aircraft engines during the latter years of WWII. I’ve always wondered about the impact on engine longevity, but given the service lifespan of German fighter aircraft at that time, I suppose that wasn’t much of a concern.
USAAF anticipated combat lifespans were rated in the low hours from what I’ve read. Not sure about Luftwaffe, but it’s probably about the same. Of course, that factors in combat attrition, so it’s kind of like life expectancy from centuries ago being low due to the factoring on of high child mortality. There was also a WEP (Wartime Emergency Power) setting that required breaking a small throttle stop cable (at least on the P-47 and I imagine other pursuit aircraft were similar) that raised power to (hopefully) get one home at the expense of the engine as I’m pretty sure that was effectively a rebuild or scrap setting depending on engine and how much the WEP raised power (usually water or water and methanol injection on USAAF and Luftwaffe, plus NOS, higher supercharger pressure for RAF). Whatever the case, they were not expected to last.
IIRC, the average lifespan for a Merlin in a P51 used in combat conditions was a mere 7 hours. Replacement engines made up a very significant portion of the logistics chain for any air group.
Interestingly the 2016 BMW M4 GTS includes water injection from the factory.
In general, yes, water/methanol injection does keep the engine cleaner for exactly the reasons detailed in the article. However, since water/methanol injection is pretty much the sole domain of older but still insanely high performance aero engines (eg. Rolls-Royce Merlin and it’s other WWII cousins) who’s service intervals between full rebuilds or outright scraps could be measured in single-digit number of flight cycles, the improved cleaning was not really a big benefit.
if only it worked on the valve tops, on Direct injection only engines.
I did this when I was young and dumb. I had a 94′ Thunderbird in highschool with the V6. Pulled a vacuum line and dunked it in a gallon of distilled water. Didn’t smoke much and luckily didn’t hydro lock. A few months later it blew a head gasket. I doubt they are related, but could have been.
I’ve also done it with Seafoam on a couple of vehicles, just pouring a little into the vacuum line at a time. I was older and wiser then and wasn’t nearly as aggressive with the amount that was getting into the engine.
That gen V6 is notorious for blowing head gaskets, so probably unrelated.
I’ve done this with Seafoam instead of water. Unhooked the brake booster line and suck a little bit up with the engine hot, let it soak a few min, then start it back up and blow the crap out. I certainly wouldn’t try this on a newer car. I always wondered about the gunk clogging up the cat on its way out, though. Also, that lawnmower engine combustion chamber is…interesting.
I do the same thing, vacuum line into seafoam until the engine stalls. Let it sit for 5 minutes. Pour 1/3 of the can into the crankcase, run for a couple hundred miles and do an oil change, put the last 1/3 in the fuel tank with about 1/4 tank and don’t fill up to near E.
But how does Pepsi work for this?
It’s pretty sweet, but you have to use flat Pepsi.
You’re trying to clean the carbon out of the engine, right? But Pepsi is carbonated, so if you just poured it in you would actually be ADDING carbon to the combustion chambers and that’s no bueno. Be sure to shake the soda a lot first so you can get all of that carbonation out of there.
COTD lol
Yep, that’s our top COTD for today. LOL
Well, I think I’m going to save myself some trouble and just get some really sugary sweet tea. That’ll avoid the pesky carbonation. I’ll probably slather a little extra sugar in there just to be safe, since I know colas have pretty high sugar content.
Don’t forget to add some honey.
Need to remember to squeeze a fresh lemon too, or it won’t work as well.
Maybe diet pepsi since the sugar in regular pepsi is carbon based 🙂 and less caloric so better for weight savings for your car.
Its the the caffeine that really wakes an engine up.
I really want my car to have a nice energy boost, so I’m gonna try to give it some cocaine.
And Red Bull
Would adding red bull make it a plane?
I think it’d turn my car into this: https://www.manx.life/cars-263