Faeflora

iTrader: (2)

This is a good paper that discusses pre-compressor water injection AKA "wet compression". It's oriented around industrial turbines but the same benefits and risks apply to our small turbos.

Link to google cache of paper

Paper is also attached to this post.


Summary paragraphs:

Wet compression is a process in which small water droplets are injected into the compressor inlet air in a proportion higher than that required to fully saturate the air. As the air gets heated during work of compression, the “excess” moisture is absorbed by the air in subsequent compressor stages. Since it takes less energy to compress relatively cooler air, there is reduction in compressor work. A reduction in compressor work directly translates to increase in net turbine output since one- half to two-thirds of a turbine output is used to drive the compressor.

Another important contributor to increasing the turbine capacity as a result of wet compression is the ability to fire more fuel in the combustor, without raising the firing temperature. Because of the evaporative process in the compressor stages there is a reduction in the compressor discharge temperature. A lower entering temperature in the combustor allows more fuel to be added without raising the firing temperature. This is the reason why wet compression has also been occasionally utilized to reduce the firing temperature in cases where firing temperatures of units without wet compression systems in service have caused concerns with turbine component life and increased maintenance.

The third factor that contributes to increase in turbine capacity is due to increase in mass flow rate as a result of water spray and increased fuel flow, since turbine output is directly proportional to the mass flow rate.

ZX-Tex

iTrader: (15)

Very interesting read. From the article

This is a really interesting observation of the erosion patterns.


Also from the article:

Braineack

iTrader: (62)

thats a lot of water, i bet that could dut you in half.

mythbusters?

hornetball

iTrader: (4)

Entirely expected for a multi-stage axial flow compressor. Not really applicable to us as we run single-stage centrifugal compressors.

Theory is sound though. Vaporizing liquid water to absorb the latent heat of vaporization is a powerful mechanism, whether it's done pre- or post- compressor.

ZX-Tex

iTrader: (15)

Agreed on both points. The article is generally interesting but there are of course several significant deviations in design, among which are axial versus radial, and single versus multistage, not to mention the blade tip velocities, blade materials, and the general scale of things.

Oh yeah and FWIW I am trying out pre and post water injection on the Vortech centrifugal compressor in my M3. 5-7 psi, no intercooler. I will be going in soon to see if I can see any erosion on the blades. The compressor geometry is very similar to a turbo compressor but the RPM is not as high.

hornetball

iTrader: (4)

Interesting. Are you running an MS or other programmable ECU on the M3 so you can take advantage of the cooling with additional timing? Also, I've done quite a bit of automatic failsafe work on my build.

Faeflora

iTrader: (2)

ZX: how are you going to position the pre compressor nozzle?

I am strongly considering doing direct port plus pre compressor WI. I have 4x .5mm direct port nozzles and am thinking about using one .1 or .2mm nozzle for the compressor for lwsser flow.

PatrickB

iTrader: (1)

I'm going to go back to the dyno soon, currently my pre turbo water injection comes on at 8psi. I am going to lower it some do some runs with it on and with it off , I'll get a print out manifold pressure. This will show if it effects spool. This won't clearly show however that it's the most effective way to do WI... So I am also going to add more injection pre throttle body so I will do a test of that on turbo side off and them both on. I think we will see that the turbo is more efficient when the water sprays into the compressor. I also think that it will be most efficient to have both. I am also going to test different water flow rates into the turbo. I have not seen any erosion of my turbo blades and it's been running a few weeks now and daily driven all this week.

Also I will be switching to 50% meth mix instead of straight water.

Faeflora

iTrader: (2)

Can you do a data log that shows psi and RPM with pre comp enabled and then disabled?? That would easily show if it is helping spool without going to the dyno.

That said, any R&D you share will be greatly appreciated. I plan to do such myself. Actually, maybe I will set up pre-comp WI this weekend and log too since I have the plumbing to do so at hand.

ZX-Tex

iTrader: (15)

No unfortunately the BMW world from what I have seen so far is not nearly as accommodating for grassroots tuning as the Miata world. The predominant tuning method seems to be 'remote tuning' via flashed ECU chips. There is not nearly as much megasquirt action. There is a programmable map based system called the WAR Chip but it does not seem to have a broad user base.

But I digress... The system is controlled with just a simple on-off Hobbs pressure switch trigger.

I think the failsafes are a good idea. What have you done?

ZX-Tex

iTrader: (15)

The nozzle is already in place. It is about 6-8 inches upstream of the compressor intake in the tube that runs from the MAF sensor. It is a small 1 gph DO nozzle running at about 200 psi for good atomization. The post-compressor nozzle is a DO 3 gph.

hornetball

iTrader: (4)

This stuff is over in my build thread, but here's the gist:

hornetball

iTrader: (4)

Why?

This is one of those wierd things in the WI world. Everyone loves to run a water-methanol mix because they think when the methanol burns, they get more power or something. Also, the WI companies sell these mixes, so they encourage the urban legend.

If you go back to the originial WI work, you'll find that there was only one reason to mix methanol into the injection media. Namely, to prevent it from freezing as you flew your B-29 bomber at 30,000 feet.

The real effect we're after with water injection is cooling of a gas mixture that is being heated by some mechanism (compression or combustion) by vaporization of a liquid. This is what relieves the pressure and allows your compressor to spool more easily with pre-compressor injection. In a large industrial gas turbine, this is what cools the combustion process and allows the introduction of more heat energy via more fuel leading to more power and higher efficiency before reaching critical turbine inlet temperatures. In our piston engines, the combustion process can also be cooled yielding similar benefits. Actually, it's a lot more complicated, but that's the gist.

Given the above, what should we vaporize? Here are the facts:

Acetone 223 BTU/LB
Benzene 168 BTU/LB (This would be a stand-in for gasoline and shows why running rich is a poor substitute for water injection)
Ethanol 364 BTU/LB
Methanol 473 BTU/LB
Water 970.4 BTU/LB (Yep, you read that right)

Water is a miracle substance. It gives us life and horsepower. Don't dilute it. Especially for pre-turbo injection. Unless, of course, you're freezing.

ZX-Tex

iTrader: (15)

There is some more interesting WI information in this SAE paper
http://www.ethanolboost.com/SAE-08FF...r%20review.pdf

Faeflora

iTrader: (2)

There is one benefit for pre-turbo methanol injection though--- methanol evaporates more quickly which means less risk to your wheel.

ZX-Tex

iTrader: (15)

I like it. That is very similar to what I was doing, except you also have the indicator LEDs. Plus I really like that test switch setup you added. Good idea. That is easier than pulling out nozzles for a visual test and probably just as effective.

hornetball

iTrader: (4)

Interesting paper. I would have loved to see a further analysis for DI of water, especially since they maintain that most of the knock resistance comes from the absorption of heat by vaporizing the liquid alcohol. I'm guessing there are long-term corrosion concerns with DI of water though.

hornetball

iTrader: (4)

I looked for a long time to find something foolproof and simple to detect nozzle clogging. I didn't like anything that I found (all complicated, prone to failure and expensive).

A manual inspection switch does the trick though. I plan to push it every time I drive the car. I'm probably going to add another switch to manually disable EBC in case the clog test fails (or in case my teenage daughter wants to drive the Miata). We do things like this all the time with safety critical systems in aviation where we can't assure automated detection within certain probabilities.

hornetball

iTrader: (4)

That's true, some of the methanol will vaporize pre-compressor. In fact, from a design standpoint, you should assume that none of the water will vaporize because you could easily be driving with saturated air (think New Orleans or Houston). This is normally why you inject post-compressor and post-intercooler -- because otherwise you run the risk of the water not vaporizing at all or even condensing.

However, the goal (as I understand it) with pre-compressor injection is to injest a mixture of liquid and gas so that pressure build-up during compression is relieved by a temperature drop due to vaporization. Vaporizing prior to the compressor doesn't achieve this particular goal -- although it's still good for the overall cycle. Note that if the goal is simply to vaporize liquid in the overall cycle, then pre-TB or port injection is more effective with less risk of condensation.

ZX-Tex

iTrader: (15)

HA! I was thinking the same thing. Why not just use water? I suspect like you that it is corrosion in the injection system they are using to inject the methanol. The other thing they mention is that the methanol evaporates very quickly which is important for DI (what they are using).