Lern me: Water injection

15 years after I was forced to sell my beloved Evo VIII, I'm back behind the wheel of another. This time, it's a Japanese-market GSR Evo II.

According to the FIA homologation papers for the Evo II and III, they ran 2 bar (29 PSI) with the stock TD05H-16G turbo. They achieved this "with water injection device."

I'm not clear on whether they mean intercooler spray or water injection into the intake. I'm assuming it's the latter, because I can't imagine intercooler spray would make such a huge difference in power. 

Which brings me to my question: is it crazy to consider water injection? I know very little about this, other than reading about some systems from AEM and Aquamist years ago.

Thoughts?

I just watched an episode of engine masters where they tested water/methanol injection on a turbo LS.  It was pretty informative about the subject.  

My turbo motorcycle has water injection. It's a Mr. turbo kit, the tank only holds around 20oz of 50/50 water alcohol mix. The tank is metal, pressurized with intake boost. From the tank it goes thru a check valve, then a jet in a spray nozzle mounted before the throttle butterfly. The engine has lower compression pistons, from the stock 10 to 1, makes about 15psi boost.

Water injection is very useful especially if you are limited on intercooler size or no intercooler. I built a very useful system for my e350 powerstroke due to no room for an intercooler on an e350. I use -32* washer fluid due to the methanol content & much cheaper than the dedicated meth/water. I can just put water if that's all I have.  1 gph at 4psi & 5 more gph at 19psi.  It's much more forgiving in a diesel & not really any tuning required & it drops the egts nicely. 
The issue I see with gas cars is the tuning. You would probably need to design the system, then tune for the system, make sure you have a solid system &  limit boost or even no boost if you lose pressure or it runs dry.  
On the Powerstroke on a cross country trip when a hose broke & the tank low level led came on I just stopped & pulled the fuse to the pump. 
I never added it to any of the turbo cars I had- Supra, XR4Ti, MR2, Hondas, due to it just being a bit more tuning intense IMO.

Windshield washer fluid  which here in Minnesota comes  in 10-20 and 40 below  ranges. -40 is 50/50 methanol and distilled water.  Can be simple or complex.  
     My original Twin turbo charged Jaguar V12 use the original windshield washers system with each spray nozzle going into the plenum on each side of the V12.  It was triggered by a boost switch that came on at 5 psi. 
  The whole turbo system was kludged up since it was built to compete in the original Chumpcar series with a $500 total cost.  The earliest EFI  system was basically 3 VW rabbit ECU's  which was slightly adjustable for mixture strength. By running full rich I used Gold palladium spark plugs to keep from fouling.   At 3 psi the cold start injectors  were triggered and at 5 Psi the Windshield washer system was triggered.   
remarkably the whole system still works. 

Find Hugh MacInnes' book "Turbocharging".  There is a whole chapter dedicated to it, including a lot of the primary research done by Ricardo in the 1930s.  Yeah, the book is nearing 40 years old, but engines is engines, most of the difference between, say, a 4G63 and a 1930s engine is detail work 

Apparently they were unable to find knock limits on their test engine up to the limit of their dynamometer, running lean of stoich and up to a 2:1 water:fuel ratio.  Specifically, they found the knock limit with an "economical" mixture, then added fuel and boost until adding more fuel actually increased detonation, then they started adding water and pulled fuel back out.

Methanol adds a pre combustion cooling effect.  Pure water actually interferes with the detonation process.  As pointed out in the book, methanol, being a fuel, can detonate too, while water does not.

Practically speaking for the kind of thing you can do on the street, a mixture works as a good compromise.  I suspect that for FIA purposes they HAD to run pure water because methanol would be an illegal fuel.

More because I do not feel like editing: GM's turbo V8s in the early 60s did not have intercooler, or even low compression.  10:1 compression and a draw-through turbo.  They relied on water injection (well, a mixture called "Turbo Rocket Fluid" *gack*).  Worked well enough when the system worked, and the driver remembered to keep the tank full.

I have installed AI systems, the modern ones have safety features like flow monitoring and low tank level switches that can be wired to trigger whatever controls your boost to go into low boost mode.  This can be as simple as tripping a relay that cuts power to the boost control solenoid.

In reply to LanEvo :

I know back in my DSM/GVR4 days people would run AEM injection when they wanted to push the td05h frames to the max. I had a buddy who would run all the boost on his td05h 16g. 

Can you pm me about your early Evo purchase? I always wanted one. 

I tune two relatively serious cars with it. They make similar power as e85 does without needing to go hunt for it. Instead they are always stocking up on cold weather washer fluid. 

First of all, sweet Evo! 
 

On a b16g with plenty of fuel, intercooler , and a tune you should be able to push that 20+ psi without water injection. Especially with e85 these days like others have mentioned.

if you're going to do it though, use the name brand stuff and make more power.

I'm considering this for my Galant VR4. It currently runs 20psi on a TD06 20G no problem on 91, but I'd like a bigger cushion before I up the boost any farther. I'm leaning towards water injection because my car currently runs like stock off boost and passes CA smog checks in it's current configuration. I think would be more difficult with a bigger fuel pump and injectors for E85. I figure head studs and water injection will get help me get the most out of my current setup. 

Calvin Nelson runs 30 PSI ( even 40 at one point) on a stock bottom end  engine with original 175,000 mile head gasket and studs. On E85.  Part of the secret is E85  alcohol has 100 octane and a much softer flame front without the pressure spike gasoline can produce. 

  There is a good probability you won't need new studs. If yours are good now.   If the head gaskets yields  all you've lost is a head gasket and studs you were going to replace anyway.  
     Watch his video's on U tube ( nivlac57 ) he's amazing the power he makes and the testing he does to justify his thoughts.  And a nice guy to boot) 

   What I really appreciate is how he takes Junkyard  motors and runs 9's with them in ordinary cars. 

crankwalk (Forum Supporter) said:

First of all, sweet Evo! 

On a b16g with plenty of fuel, intercooler , and a tune you should be able to push that 20+ psi without water injection. Especially with e85 these days like others have mentioned.

Thanks! The previous owner of my car upgraded the intercooler (and a bunch of other stuff) so that the car runs 21 psi with no boost creep. It's tuned for the 94 octane available at PetroCan. The best fuel I can find around here is 93, but the car seems to be running well with that. There is no E85 available locally.

I'm forced to replace my turbo (long story) and have a brand new TD05H-16G on order. I've been reading up on CE9A Evos and this whole water injection thing seemed pretty reasonable to me. If I can up the boost another 7-8 PSI, then that should be good for another 50 or so WHP. Not that I really need it, but if I can do it safely... then why not?

In my experience I wouldn't run more than 22-23 psi on a b16g without it being the evo III frame. I made 330 awhp in my galant on my b16g and went to an external dump pipe to control creep at 23 psi. That was on 93 octane and DSMlink. I went to a forced performance fp green that was internally wastegated and didn't look back. If you want 30 psi+ I would look at bigger turbos. I'd much rather have a 50 Trim at 20psi than 16g at 30. 

crankwalk (Forum Supporter) said:

In my experience I wouldn't run more than 22-23 psi on a b16g without it being the evo III frame. 

The turbo I ordered is the MHI Evo III unit. The exhaust manifold and O2 housing are also stock Evo III, converted to open dump. I'm also running DSMlink, but I have a MoTeC M4 with complete Evo V wiring harness that I can use if that makes more sense.

Part of my motivation is to keep the car close to factory Group A Rally specifications. For a track/hillclimb car, I really like the quick spooling of the Evo III big 16G. I'm afriad I'd lose some of that responsiveness with a bigger turbo.

LanEvo said:

crankwalk (Forum Supporter) said:

First of all, sweet Evo! 

On a b16g with plenty of fuel, intercooler , and a tune you should be able to push that 20+ psi without water injection. Especially with e85 these days like others have mentioned.

Thanks! The previous owner of my car upgraded the intercooler (and a bunch of other stuff) so that the car runs 21 psi with no boost creep. It's tuned for the 94 octane available at PetroCan. The best fuel I can find around here is 93, but the car seems to be running well with that. There is no E85 available locally.

I'm forced to replace my turbo (long story) and have a brand new TD05H-16G on order. I've been reading up on CE9A Evos and this whole water injection thing seemed pretty reasonable to me. If I can up the boost another 7-8 PSI, then that should be good for another 50 or so WHP. Not that I really need it, but if I can do it safely... then why not?

What's it like driving from the wrong side of the car? 

Hey, guys. I design and provide water/methanol injection systems every day. -been at it for the last 20 years or so and have seen these kits go from very crude to exceptionally sophisticated. Most systems use the same pumps (with different decals). The big distinction among suppliers is the nozzles selection, type of fittings used, and the controllers offered.

The two dominating kits on the market are AEM (entry level) and Snow (middle tier). AEM provides just 3 different jet sizes. -small (250cc), medium (500cc), and large (750cc). One nozzle body is supplied along with all three modular jets. Thus, there is no customization possible. One kit is exactly the same as the next.

Compare that to Snow Performance. They've got ten nozzle sizes or more. Your means to actually customize a setup depends on if you're buying from eBay or a stocking dealer, etc.

Also important are the fittings. AEM uses old school push-to-connect fittings. Snow offers excellent 1/4" compression fittings. My own company does, too. Plus, we also provide smaller 3/16" fittings and tubing (for sharper control response when multiple nozzles are installed). Here's how compression fittings work.



AEM's basic controller is entirely adequate. Snow has an equivalent and then a pair that are significantly more advanced. The best currently on the market is the Cortex EBC. As the name implies, it's actually boost controller. It also does incredible water/meth with advanced 3D mapping.

By far, the most critical variable is whether or not you're working with a mere kit seller or a system designer. Just as with other aspects of tuning, water/meth injection is a technology with plenty of variables to master and nuances to get just right. You're very curious about how all this works? Desire a real consultation from a racing industry pro? I'm your guy. Shoot me an email or drop me a line: 856.456.3335.

That 8 port setup is pretty sweet.

I have never worked with a direct port AI system, but really want to.  The only ones I have done had a single nozzle aimed at the throttle body or supercharger inlet, and EFI intake manifolds are not designed for wet flow distribution.

I ran an old school DIY setup, actually, based on old DSM directions. I think I got from Turbomirage.com, long gone. Water pump, accumulator, pressure solenoid, single nozzle in the intact tract. I "tuned" by watching EGTs. They've obviously gotten more sophisticated. I was running this on an MR2, and was more interested in the cooling rather than fueling effects from a meth blend, as Mr2s are notorious for heat soaking the IC and I had a JDM ecu with more aggressive ignition timing and was more of a CYA kind of thing. I still think it's fairly underrated and inexpensive to implement. With flex fuel/E85, that seems to be the easy button. 

https://web.archive.org/web/20200218172917/http://www.turbomirage.com/water-alcohol-injection-parts-list.html 

If you're running a J-57, it makes for very impressive takeoffs.

Some reading from the GRM site that might help:

Is a Water/Methanol Injection Kit the Tuning Aid You Need?

Is Water Injection a Cheap Substitute for Race Gas?

This was my DIY setup about 20 years ago.

A Shurflo (I also used in tank fuel pumps with no issue) pump fed the block which has a boost referenced OEM fuel pressure regulator, which in turn fed one or two OEM fuel injectors depending on the application. I used both standard and SS injectors, both worked fine. The injector driver board is just a 555 timer based extra injector controller I built. It used a tach signal for the pulse and you could vary the width with the potentiometer. This is a really simple but effective setup that allows for a lot of adjustability. I'm surprised you don't see more systems like it. Or maybe I'm missing something.

i was told that EFI injectors would rust the pintles to the body if exposed to water.  Hm..

FWIW, I did use them with windshield washer fluid, and at one point used (what I was told was) a single SS injector, but I can't recall what the OE application was for the SS injector.

The standard Suzuki injectors I used worked for a few years in regular use with no issues. I did check them occasionally for that exact reason.