Having used water injection since the late '70s/early '80s, I admit a long-time fascination that began after having read about P-51 Mustang racing airplanes. Termed "detonation suppressant" in the P-51 ranks, I saw water injection as a natural to internal combustion engines exposed to stress. Especially engines whose dynamic compression ratios extend to 17:1, & beyond. Libary reading of [names-forgotten] library books showed that varied methods of application altered the results of injecting water. Allowing phase change to occur closer to the combustion process, serves more as a detonation suppressant; while futher away, earlier in the intake tract, serves more in reducing intake charge temperatures. Both yield positive benefits, but are divergent in effects on combustion, requiring the tuning approach to take this divergence into consideration.

More than one time I've read that X° of temperature reduction equates to a power increase of Y horsepower. Most often I see that a 10°F drop to intake temps equals 1 horsepower. In & of itself, increasing horsepower by reducing temps based on this formula would not yield a significant performance improvement. Perhaps more important than making visceral power is the new set of conditions that are presented to the combustion process. A cooler, denser intake charge is more conducive to controlling combustion, wherein enhanced reliability is realized.

Controlling combustion, ha! LOL, Who am I kidding? No one. Detonation, pre-ignition, audible & inaudible knock, are all terms relating to critical threshold values that a successful performance engine will respect. They are the line drawn in the sand, where life or death is decided. I hope this metaphor does not go unappreciated. Mechanisms to reduce or eliminate detonation include ignition timing, robust ignition, cam timing, fuel quantity, fuel octane, compression ratio, piston design, combustion chamber design, spark plug selection/gap, & others that don't occur to me at the moment...except for water, methanol-water, or methanol injection. Yes my friends, these near mystical methods of auxilliary injection provide one of the most tractible means of controlling the demon detonation.

Adding more ignition timing is likely the most common by-product of implementing this form of injection. Advancing base timing, & seeing the ECU still not pull back timing under high boost is a sight to behold. It is a pleasure usually reserved for the use of racing gasolines. Coincidentally, reliance on gasoline as a coolant to stave off detonation is a major benefit, thus allowing lower injector duties for a given power output. Running the fuel system at moderate injector duties serves as another reliability enhancement. That factored-in overhead combined with an -aware ECU [EG, EMS] can provide contingency-based fallback maps. Using wideband oxygen sensors, amp; EGT data, or multi-gas analyzers in either realtime on the dyno, or datalogging on the street, full advantage of the detonation-suppressant nature of injection can be taken. So much fun. AFs change with timing changes, too...so, as I may have said elsewhere many times, it can be tedious!

The mixture responsible for this dense, cool intake charge also causes lost volume in the combustion chamber, which serves the premise of combining a high octane component like methanol, or other long chain alkane alcohols. Providing a participatory component to the combustion process helps compensate for the space occupied by the water/water vapor, & a more ignition-friendly mixture is created. Additional value of alcohols is found in high latent heat absorption values. The actual BTUs of absorption are lower than water, but remain a significant contribution to intake charge cooling.

On the last dyno using plain water, AFs were found @ ~12.5:1 thru the meat of 4000-4700rpms, yet EGTs ran ~100F lower than previous runs @ ~12.3:1. Leaning AFs beyond 12.5:1, made power begin to drop, though no audible knock was noted. Likely the knock sensors in action. Keep in mind lambda values when methanol is used. The Methanol-Water Mix section of this page may provide some additional insight if you plan to use it.

Previous to the water injection, 11.8:1 was deemed the best compromise for power/safety- so there was a measurable change in AF requirement to obtain the highest output/safety margin for that day. Disclaimer: By endorsing H2O injection, I cannot stress enough the need for a properly tuned fuel curve, timing curve if you have access, or a reliable knock sensing system if you don't have timing control. It's OK to play, but decide in advance how far you want to go- like choosing your boost level. Keep the ECU/Knock sensors' max threshold[s] in mind, & AFs safe enough to not blow rings/ringlands if the injection happens to take a dump. Use a programmable ECU, & set safeguard thresholds, alternative safe-mode maps, etc.

When elevated boost pressures are targeted, spark plug gaps oftentimes require reduction to ensure combustion. This is a fixed parameter decided upon by analyzing the specific results of using various gaps, as well as considering common use of cooler temperature spark plugs. Trial & error does not take long to find the right temp plug, & sweet spot gap where idle has as few bumps as possible, & boosting to 25psi exhibits no misfire. I start by gapping high to the point of misfire, then backing off a big notch. On my engine, a 0.034" gap had trouble above ~1.5 Kg/cm² on the TO-4E. A 0.026" gap was tried next, which gave back a crappy idle- sort of expected. Power was off across the board, too. A move to OEM spec of 0.031" gap gave good idle & no misfire to ~2 Kg/cm² boost, injection at full gallop.

Testing will tell you what you can get away with. Heh, I use Autolite 3923 copper spark plugs, one step cooler yet still manage OEM gap on them under full boost, without any misfire under load. No, I'm not lucky. A perfectly in-spec OEM ignition system on the 7M-GTE is an engineering marvel; so be sure to have one if you're modding the 7M-GTE, with or without injection. Spare the "but it's a wasted-spark setup," crap. It lights up just fine, BTDT...with injection.

More timing? More boost? Injection yes, but where, when, & how much? Dyno time is invaluable to determining what combination works best, given your unique engine setup. The process is tedious, yet exciting doing have gotten me doing this ~30 years later. More boost, more power. And when each engine disassembly invariably arrives, the very clean intake tract, head, & combustion chambers are exposed, & a retrospective smile will overtake me.

Water injection is relatively easy to install, & calibrate; while inexpensive to maintain. If you only use water, it's just the cost of de-ionized water at your local market. Don't us tap water, please. Methanol/alcohol will increase the cost, depending on what is chosen; but is worth the cost. Literate modders will not find this mod difficult, & will usually yield positive results without too much fuss. The early Buick GN modders used to do it with hardware store parts, while current kits offered by GN enthusiasts can be quite elaborate. The oft-mentioned ERL's 3-D mapped system is quite the techno-marvel. Given the pricepoints of their systems, they sure do think highly of them. Neat stuff, but finding a happy medium is probably the best approach, as the benefits of one system over the other are often touted, but ask yourself: Are they worth the difference in cost? How much of the 'kit' am I going to have to modify to make it work for my engine anyway?

Another price of sorts is an additional maintenance item on the 'frequent fill' list, as well as an occasional adjustment, based on anticipated macro changes to ambient temp/altitude/boost level. Running steep graded, high elevation summits to reach Death Valley in 104°F ambients, to arrive there below sea level at 114° has been done by OldSchool Supra...windows up, AC on. With lots of hill-climbing boost-on, low-pressure injection points were selected, & the engine ran so happily that any maintenance issues melted away. :-)

Final Note
Using this as a mechanism to flirt on the edge isn't recommended, as it would then become a point of failure. This is the voice in your head that keeps you alive. The voice in your head that likes to party tells you that injection works, & you better let loose on some wild tuning. Yes, it works, & I like to party, LOL. I live by two credos:

• Life is finite, so make the most of every day's 1440 minutes.
• Do that, which works for you.

A view of the parts used during the initial install back in 1998. Not seen in the photo, is the
secondary pump that was installed later on in 2000, in addition to the Spearco-provided pump;
nor the inline check valve that failed early on.
Parts in the small red bin were Spearco parts that didn't meet with approval.

This shows the polished reservoir, Air Logic pressure switch, and stainless steel braided lines for vacuum.
The Hobb's switch near the center of the photo came with the Spearco kit & went unused; it was quite cheesey.

Another view of the reservoir's top.
The reservoir was initially located right next to the passenger-side headlamp, in front of the radiator.
Removal of the windshield squirter tank in the fender, created a new home for the tank.
This move was an item in the *maximize engine cooling* project list, as well.
So, the tank found a new home in the passenger-side fender, & the additional pump [series]
was plumbed in during the relocation.

The AirLogic pressure switch, installed on the inner fender using an aluminum bracket that was massaged by Dremel tool.
The slotted adjustment screw is visible on the right side of the switch. What's neat about this model
of switch, is the fact it can be configured as normally open, or normally closed. Great for NOS deactuation.
The -04AN junction you see below the switch, is the post tank/filter line headed toward the injection nozzle.
Note: This switch will be replaced by the AEM EMS for actuating the system.

---

View of the 0.4 gallon tank installed where the windshield squirter tank used to reside.
Heh, I just filled it, so look for the water marks, & brass cap.

---

View of nozzle installed in the Intercooler pipe.

---

View of nozzle in relation to throttle body.

---

View of insulation wrapped -03AN pressure port.
Crappy pic, but that's where it is...right above the strut tower bar, where
the word "wrapped" is positioned in the previous sentence.

---

View of 0.031" nozzle installed in the IC pipe.
The VPC temp probe can almost be seen behind it, above to the right.

---

View of injection activation ON LED, on the right side of the ashtray slot.
The HKS Type-1 Turbo Timer, & OldSchool AVC-R can be seen to the left of the LED.
The edge of the Halmeter AF30 can be seen on the far left.
Note: The turbo timer is gone, & the AVC-R will be removed at some point in time as well;
to be replaced by the AEM Engine Management System [EMS] including their capabilities.

---

   50/50 Methanol-Water Mix

Mix?

50/50 Methanol/H2O was tried as the initial injection medium, back in '98, then again in 2000 when I added the aux pump & did disassembly/inspection/maintenance. Once dialed-in, performance was phenomenol, & lowering AFs [per Motec], while still enjoying reduced EGTs [60mm HKS P-H-W in deg °F] by ~100°F. Reading AFs by Oxygen sensor while using combustibles other than gasoline require utilizing lambda comparisons, rather than straight readout values. If you use alternative fuels, & desire more information in this regard, please search the Internet, or visit your local library. Just know that methanol is CH3OH, & be certain that combustion will make that freed atom of Oxygen combine with some free N, C, & H; invariably affecting your AF readings. If my chemistry is wrong is of no consequence, just know that AFs are affected.

So, what were the differences on the dyno? Heh, oops. Time doing direct comparos with/without injection alone wasn't made. NOR between injection mediums. Don't get me wrong, plenty of dyno tuning was/has/& always be done, but primarily to get done, and only some comparisons will be logged in the Dyno/Performance section of this Site. Usually, bulk-mods are done between sessions, so please understand. See the Evaluation section of this page for some data points to peruse.
Qualitative observations, including engine behavior, dyno curve quality, & SOTP all point to a positive experience. Using it on the TO-4E [0.60AR/60-trim - 0.70AR/P-trim] was inconclusive, since it was on pure water, other problems manifested, like IC hoses popping off trying to hit 25psi. More work will be done using methanol-water mix once a brand of alcohol is found that doesn't corrode the aluminum like the product pictured above. I obviously bought the wrong stuff, because the alcohol I bought back in the '80s certainly didn't do what this stuff does!

Yes, that blue-canned stuff with distilled water left a white powdery residue on all exposed aluminum & brass surfaces. Initially, the 0.025" nozzle was used, & a disassembly/inspection of the system after only one tank revealed some of the aforementioned residue. Not much, just a hint. Reassembled using the 0.031" nozzle, a few tanks went by, & aftr alittle less than 3 months, the quantity of residue found during disassembly was alarming, The filter was semi-clogged, as was the injection nozzle. The cast alumnium of the filter was the most corroded component discovered in the system.

Spark plug inspection revealed spotless electrodes/ceramic, while the perimeter of the threaded barrels had some carbon, zero speckling. Piston tops were clean. 4K mile Mobil One 10W-30 oil analysis, revealed normal FE, Cr, Zn, B, no water, 25% Oxidation; but Al was 18ppm, basically indicating Al corrosion. Still, overall, valve-train, rings [et al], & oil were basically happy. I haven't done anothr oil analysis since I began running plain water.

If you [Mr/Ms Reader] are aware of a product that contains methanol, & is compatible in a 50-50 mixture with distilled water, please let me know. Thank you kindly!.....
Methanol product info for Larry link.

BTW: I ran a 50/50 Methanol/H2O mix on Celicas [early '80s, 18R/20R motors] all the time. Run thru Weber downdrafts, & side-drafts, there was no weird residue; perhaps due in part to a plastic tank, & minimal exposure to metals. Whatever. Plain water in the Celicas ran like shi+, that's all I know. The change from the 50-50 mix was enormous. I'm old, & it was long ago, but I still remember shaking my head at the improvement the mix had made. Any myth status the 50-50 mix may have had with regard to efficacy was removed in my mind's eye. If not for the corrosion, I'd definitely run the mix in the Supra; however, until I locate a non-corroding alternative to my mixture, it will be plain water on the Supra. Fortunately, plain water has not shown itself to be quite the detriment it was on the Celica. I know of some guys out in the world who've gone to 100% methanol that usually includes an upper-valve area lube additive in the injection medium. I do not deny the additional potential for power, but I prefer running a a less-volatile concoction. Besides, water was my inspiration for the entire idea, & a wannabe-tribute to the P-51 Mustang. I can think of no other propeller-driven plane I actually yearn for a ride in. But I digress.

---

   Evaluation

Test activation points in Kg/cm²

• 0.53
• 0.74
• 0.86
• 0.99
• 1.21

Three [3] runs were done at each activation point. Boost was set to ~1.4 Kg/cm², & was solid on all runs.

At ~0.5 Kg/cm², an occasional bog occurred, depending on ambient temps. The cooler it was, the more noticeable the bog. At ~0.7 Kg/cm² & above, there remained a mild surge [not turbo associated]- more a change in characteristic, rather than a negative symptom. ~1.40 Kg/cm² setting on the AVC-R using a 0.58 valve actuator duty setting, yielded 1450°F pre-turbo EGTs, while the AVC-R reported injector duty of 60%, at a 23psi [vac-line off] reference FP; 74°F ambient. OEM 10deg btdc base timing. This combo provides decent AFs, a 1psi lower fuel pressure than spec'd by Toyota, while not exceeding 1550°F EGTs from non-water injected operation. SIXTY PERCENT Injector Duty, at 23psi reference fuel pressure is crazy.

Confirming a measurable ECU-induced timing benefit was never done. This is the currently held premise, as a series of changes to base timing was done to observe engine behavior. Yes, minor adjustments to the GCC were required along the way. Sorry I can't explain it, but after so many years of doing this, I just know when you've got good timing. Call it a feeling, call it VooDoo, call it ESP, but don't call it arrogance, thanks. I also used to balance dual Webers by ear, & still do lace/true racing bicycle wheels, so I'm not normal. Further, I modified the vacuum advance distributors to full mechanical advance on both 18R & 20R motors [back in the day], then proceeded to run an extra 6° of base advance, using rpm-switched[!] water-methanol injection. The extra timing was mechanical for Pete's sake, & scoping the motor allowed plotting across most of the usable rev range, but it worked. With no compensation for load back then, methanol-water injection was a God-send for spots along the curve that would normally induce detonation. Oops, got off the Supra-track there & went full Celica.

To conclude, intake charge dilution was a concern, but revving a peaky NA 4 cylinder motor to upwards of 7500 rpms meant intake velocities that dispersed the injection medium quite effectively through the 'wet' manifolds of yore. SOTP only, no known dynos around back then. Anyway, this oldschool modding [with known timing/curve], illustrates what feels like a parallel experience with the Supra. Although verification by TechTom was never done, I can only surmise on the current ECU-controlled setup. Running the car without the water injection at 1 Kg/cm² boost, 72°F ambient- as boost surpassed ~0.8 Kg/cm², the car lurched [albeit gently], and speed increase from there was not nearly as crisp as it had been with water injection operational. Putting the water injection back into service gave back an amazing rush.

Thank you for reading. If you have any comments, questions, corrections, or flames, use the e_Mail link in the Methanol-Water Mix section.

---

Back to Engine

---