water injection = Higher Efficiency?
 

hi guys,

Im trying to workout a way how water injection into the cylinder can be used to convert the waste heat energy specifically into more useable energy, easily.

try reading about the crower 6 stroke engine and you'll see the angle im trying to get at.

basically, ICE's lose 30% of energy as waste heat out the tailpipes and radiator. how can we get that 30% to do some work?

water expands ~ x1800 by vol when it absorbs heat energy, its latent heat of phase changes is comparatively high. meaning, so long as is doesnt hamper the combustion process itself, it "should" be able to absorb the waste heat energy in the combustion cycle and provide additional pressure energy to the piston similar to a forced induction setup.

an easy way i see of implementing this is trying water injection(NOT METHANOL) into a nicely setup lean burn engine(where all the LB bugs have been worked out).

what do you guys think?

I've been thinking about this for a while, and instead of putting down the money for a kit with pump, injector, and controller, I was just thinking of doing this:

http://www.youtube.com/watch?v=k5n7d...06A1zzkcLwA%3D

The only thing I'm waiting for is the time to get started on all my projects.

This isn't as good as mist injection, but this more closely resembles driving on a foggy night which has shown to increase fuel economy. Plus, this setup is practically free.

I think it'll work, because unlike a lot of MPG enhancers, the science behind this idea is sound. If you try it out, be sure to let us know how it works.

If you read my wiki and bothered to check my links, water injection on a gasoline engine doesn't appear to boost FE.

Water injection - EcoModder
http://www.files.thinksitout.com/Alt...e%20Engine.pdf

I think you are better off looking for other ways to boost fuel economy on a gasser.
You can always try.

Water injection has proven its self enough for me to where this weekend or next I am going to try and fit a 2003 duramax 40 gallon diesel fuel side tank on my suburban for holding water for the water injection system.

water injection has been on my todo list also.

I think a big key is going to be to get the water vaporized to steam, and I think you need a LOT of water.

Maybe enough water to where you don't need a throttle plate or you get rid of your pumping losses?

...+1, see my earlier posting on this exact subject: http://ecomodder.com/forum/180860-post10.html

I tried this on a Ford back in the 80's.

From what I remember you only see gains on a very high compression engine. The water helps prevent detonation and allows for greater horsepower. Don't know what it does for FE.

Some radial engines in WW2 used water injection to increase horsepower.

I'm adding your single post to my WMI wiki.

Actually the WW2 engines used water injection for emergency power situations, where boost was a high as two atmospheres or more. I think it went to 35 PSI boost, which at sea level meant the engine would blow up in a matter of minutes, if not seconds.
War Emergency Power settings were for emergencies only, like in a B17 where you had two engines shot out and you needed to clear the coastline in order to ditch.
Heard once about a B17 that skipped itself over the channel and just cleared the cliffs of Dover on one engine.
You can get a lot of power out of a car engine. The old SOHC Nissan V6s (2 valve) were used in Rutans Pond Racer. They were boosted up to 1000 HP but engine life expectancy was measured in hours or less.


regards
Mech

water injection also enables you to run advanced timing, which should help f.e.

Water injection by itself on a normally aspirated engine is not going to return any FE or power gains. It does do a fabulous job of cleaning combustion chamber and piston of deposits that may have formed. If you raise the compression/expasion ratio along with water injection then you will realize FE and power gains.

im not sure if thats constructive critiscism or sarcasm.

either way, i figure this was the forum to throw this idea out. the science behind it, especially the thermodynamics side of it agrees.

i do however admit i didnt do a forum search before i posted.

i still think a High Compression + Lean Burn + Advanced timing + Water Injection setup should create a noticeable increase in FE over just a Lean Burn + Advanced Timing setup.

I could see water injection helping in 3 ways.
1. Reducing pumping losses. Add enough hot water/steam, and vacuum should be less, leading directly to less pumping losses.
2. More power. Add enough water, and the water turning to steam increases cylinder pressures after combustion, giving more power for a given amount of fuel.
3. You could add more compression, directly leading to more efficiency.

#2 is factually incorrect.

Any time you add a liquid into the combustion chamber, the act of vaporizing that liquid will reduce the peak temperature of the mixture captured inside the combustion chamber. Due to Boyle's law, this will also reduce the peak pressure, and cause the process of extraction of mechanical energy to become less efficient.

And if you add enough water to the incoming mixture, you'll just quench whatever combustion does occur.

The entire idea behind water injection is to cool the charge mixture enough to prevent detonation and/or preignition. This makes it possible to safely force in more charge mixture than would otherwise be prudent. Yes, it also means that water injection also allows a higher compression of the existing charge mixture than would otherwise be prudent.

how does an old school steam engine work in a locomotive????

Well, the steam engine is an example of the thermodynamic process known as the Rankine cycle. Basically, you have a coal fueled fire that heats up a pressure vessel that contains water. The water boils to become steam, which is then further heated so as to put the steam into superheat. This is necessary, because the steam has to be above atmospheric pressure to effectively push the locomotive forward.

The superheated steam is then admitted to one side of the piston via a reciprocating valve mechanism. As the steam enters the piston, the steam pressure causes the steam to expand and push against the piston. Since the reciprocating valve is still open, we can consider the steam pressure to be relatively constant, so we can model the steam expansion as an isobaric process.

The reciprocating valve then closes, while the steam further pushes against the piston. At this point, the steam pressure inside the cylinder drops as the piston travels. Due to Boyle's law, the steam temperature also drops. This can be modeled as an adiabatic process. The piston travels to its distant end.

The reciprocating valve then opens up in the opposing direction, and steam is admitted to the other end of the piston, beginning the process described above to push the piston back to its starting point.

It is important to note that the Rankine cycle is substantially different than the Otto cycle, in terms of execution.

Seems this topic morphed or vascillated from adding water as a means of enhancing the expansion cycle to using water as a facilitator (detonation suppressant) in order to enjoy the benefits of higher compression and/or leaner mixtures. Personally I think the latter is more likely to be useful.

Here are some good questions for those seeking to use water in an attempt to recover waste heat:

What is the total heat energy of one mole of exhaust gas at 600 C and 150 kPa? You may assume, for purposes of this question, that exhaust gas is 80% diatomic nitrogen, 11% carbon dioxide, and 9% water.

What is the total heat energy of a 1/10th of a mole of liquid water at 30 C?

Mix the two together. What is the final temperature of the resulting mixture? What percentage is liquid? What is the final pressure? Is this pressure higher or lower than the pressure of the original exhaust gas?

imagine the combustion chamber at 600C full of hot gasses at a pressure of 500 psi.

Add excess liquid water. My GUESS is the pressure will go up slightly, the temperature will go down significantly.

Now expand the volume by letting the piston go down. As the piston goes down, the pressure normally would go down, but the liquid water is rapidly turning to steam.

Overall, on average, the resultant temperature WILL be significantly lower over the entire stroke of the piston as the latent heat of vaporization for water is very high.

Will the average pressure be higher or lower with water?

Thought experiment two. Imagine a 4 or 2 stroke engine, except it has superheated piston, and we have direct injection of liquid water. We inject the water at TDC.

Will this engine "run"????

I think some of the confusion in this thread comes from two different topics being discussed:

A) What is traditionally referred to as water injection - injecting water with the air/fuel mixture on a 4 stroke engine to help cope with high compression or high boost.

B) Add 2 strokes to a 4 stroke engine. At TDC during the 4th stroke (exhaust) inject a small amount of water, which will immediately vaporize due to the heat of the cylinder and piston, creating a second power stroke. This is the concept of the Crower 6 stroke engine. As a result of excess heat being carried out of the combustion chamber during the 5th/6th stroke, more aggressive timing/boost/compression can be used during the traditional 4-stroke cycle.

Points are not awarded for guessing. If you add even a small amount of low temperature water (low temperature meaning that the water temperature is such that it's still a liquid) to a set amount of high temperature gas, the temperature of the gas will drop. Because the temperature of the gas drops, so does its pressure. Since water needs much more heat energy to raise its temperature by 1 degree C, as compared to the gas in question, it's going to lower the gas temperature that much more. Finally, that water has to absorb even more heat energy from the gas if it's going to vaporize (latent heat of vaporization, right?), which is going to further drop the temperature of the gas. The gas temperature will drop significantly. You're not going to get the results you've envisioned.

You apparently are making the mistake that exhaust gas has the same specific heat capacity as water. This is not true.

You really need to work the math out yourself. This statement is factually incorrect.

With water? Lower. Much lower.

Not nearly as well as the standalone engine without direct water injection, if it runs at all.

I'll construct a graph later on, that graphically shows what happens when water is added to exhaust gas.

when water turns to steam it expands.

it expands a fair amount.

I haven't done the math on the heat, and it sounds like you haven't either.

Sounds like we are both guessing.

You must first account for the fact that your water injection scheme must allow for your liquid water to rise in temperature to water's boiling point. This is something you are neglecting.

The specific heat of liquid water is about 4.18 J/gram-K. In other words, you need to add 4.18 joules thermal energy to a gram of liquid water in order to heat it up by 1 degree K.

The specific heat of steam is about 2 J/gram-K. In other words, you need to add about 2 joules of thermal energy to a gram of steam before it rises 1 degree K.

The specific heat of air is about 1 J/gram-K. In other words, you need to add about 1 joules of thermal energy to a gram of air before it rises 1 degree K.

Now, for purposes of this little "thought experiment," as you would call it, we can consider that exhaust gas (that stuff that we're going to inject with water, remember?) will have a specific heat of about 1.05 J/gram-K. Do the math yourself. So, taking 28.86 grams (or one mole) of exhaust gas at 600 C and 150 kPa, we then spray 1.8 grams (or 0.1 mole) of liquid water at 30 C into it.

Now, in order for a gram of water to rise in temperature by 1 degree K, it needs to take about 4.18 J from a gram of exhaust gas. That means that for every 1 degree K per gram rise in liquid water temperature, the exhaust gas must drop 4 degrees per gram. You do remember, of course, that in order for liquid water to become steam, it must first reach the temperature at which it would turn to steam.

So, to heat that 1.8 grams of liquid water from 30 C to 100 C, we need about 527 joules from the 28.86 grams of exhaust gas we have. That will cause the exhaust gas to cool off by about 499 K. So, now, instead of exhaust gas at 600 C, we now have exhaust gas at 101 C. Hm... that's about the same temperature as the water we just injected!

Applying the ideal gas law, we find that the pressure dropped from 150 kPa to 64 kPa, which was below atmospheric pressure last time I checked.

Oh, but wait! We need 2260 more joules per gram of water to transform it from a liquid to a gas (which I like to call "steam"). You read that right - we need approximately 541 times as much heat energy to turn the liquid water into steam, than we did just to heat up that water by 1 degree C! Whereever are we going to get that from the exhaust gas we just got done cooling off?

So, instead of coming up with some clever new way of pushing a piston that somehow eluded the best minds of the 20th century, we merely came up with a novel way of cooling off exhaust.

You're about as ignorant of thermodynamics as you are of aerodynamics.

This much is obvious.

My guesses are backed up by college-level thermodynamics, supplemented by Naval nuclear training. What are your guesses backed up by?

Well, to be fair, you are cooling more than the exhaust. Both the Crower Six-Stroke (2006) and the Dyer Six-Stroke (1915) were able to run without a cooling system as well :)

In reference to your above argument, the energy that converts the water to steam comes from the piston/valves/chamber/etc, not just the exhaust air. The problem is that extracting the heat from there just increases the delta-T between the combustion and chamber walls (and therefore heat transfer) on the next power stroke, effectively weakening it.

There is probably a reason Crower never finished his patent application, and why Dyer's engine never went anywhere 100 years ago.

True that. I just wanted a simple enough presentation to show the absurdity of trying to recover waste heat by spraying water into a combustion chamber full of spent exhaust gas.

It's much the same as with HHO or magnets. Despite hard empirical proof to the contrary, some true believers will continue to push their version of how things should be. Might as well harness the power of all them midiclorions in the cells of their bodies, for all the good it'll do.

how many grams of exhaust gas in the cylinder?

what is the expansion ratio of water to steam?

seems like if we are running "efficient" we are somewhere in the neighborhood of 14:1 air to fuel.

If we run twice as much water as fuel, then our final ratio will be something like 7:1 water to exhaust gas (rounding a bit).

one of the interesting things is our pumping losses go towards zero - if we get the water up close to 200 or so before we inject it, the "vacuum" of the intake will cause it to boil in the partial vacuum.

gotta love it when the engineers start spouting degrees and years of service to the guvment. fwiw, I'm pretty sure i've got you covered, but my schooling was more theoretical and less train driver driven.

What difference does it make? You completely missed the point.

Look, even if you injected 1% by mass of liquid water into the exhaust inside a combustion chamber when the piston is TDC, you're not going to get the result you think you're going to get, even if the water's near 100 C anyway.

That pesky latent heat of vaporization, that you keep neglecting, is going to cause the exhaust to cool off as the liquid water vaporizes. What's more, once water enters the vapor phase, it's a gas. It gets treated as a gas, and it behaves as a gas. It'll reach equilibrium temperature with the exhaust gas it's mixing with. The whole gas mixture will be at a lower pressure than would exist had the exhaust gas not been sprayed with liquid water, because the temperature dropped due to the water taking heat energy away from the exhaust gas in the process of becoming steam.

By the way, what's the pressure of steam at 100 C?

Efficient part-throttle operation is actually around 17:1 with gasoline, which is why Lean Burn is so popular among those who are able to use it. The rest of us are not allowed to use it as a factory option because of the Clean Air Act of 1990. Try again.

Is this a guess, or is there something you can show? And you do realize that exhaust gas has water vapor in it, right?

Assuming that you mean to flood the intake manifold with steam, then yes, pumping losses do approach zero. However, that's because you're diluting the intake charge with an inert gas (water vapor), thereby forcing the throttle to open up to maintain the same oxygen intake, not because of any magical properties of water becoming steam inside a combustion chamber.

This much is also obvious. You keep on with your "thought experiments" and such, never bothering to once run actual experiments to see if your ideas might actually work. Philosophers run thought experiments. Engineers work with reality.

Is anyone really going to reinvent the wheel and put this engine on the street?

your posts are quite informative & you clearly Know your thermo, i get your arguement on latent heats required just to arrive at the point where water turns to steam.
but then again, your only considering the heat energy available/escaping with the exhaust gasses.
let me ask you this, theoretically speaking, an ideal engine will have no/minimum heat transfer to outside the working medium (pison/valves/cylinder ect) correct? that is why ceramic coated pistons/liners/sleeves ect are benificial as i understand it. which means more heat trapped inside to do work.

drmiller I really don't mean to be blatantly offensive, but you touting your "theoretical schooling" is basically just as bad if not worse, especially given how you tend to frequently forget about some important factors at play.

t vago, not all theory heavy people are like him :P

kevman, when the temperature is high in the combustion chamber the pressure is also high, and if the heat escapes then the gas loses energy and pressure, and you get less work out of it. I think the reason we haven't seen the "6 stroke" engine in reality is because most of that heat energy is going out the gases in the tailpipe rather than being retained in the metal, and so you don't have much energy to work with. If you had an engine with no friction and perfect insulation you'd have something like 40-50% (depending on a bunch of parameters of course) of the heat energy of the fuel leaving in the exhaust gas itself. A cooling system can be thought of as sapping heat energy from the combustion itself, but it also cools the exhaust ports so it's not too clear how the heat energy is distributed there.

At any rate something like 30% of the heat energy is leaving the tailpipe at temperatures around 400-700 (or higher, which is kinda bad but it happens) C. The cooling system runs at 90-100C, and the engine internals are kept well below 200C, and the flame front shouldn't be hitting the piston early when temperatures are highest on a good combustion chamber design. The higher the temperature of the heat source the easier it is to get power out, so you can see why people are turning to the exhaust for waste heat recovery.

take a closed volume of 40 cc's at 800 degrees F and 150 psi.

add one gram of liquid water.

water has an expansion ratio of about 1700 to 1.

I believe the pressure will rise.

Significantly.

In order,

In order, pressure of steam at 100c is 1.0142 bar. which is interesting if the rest of the "atmosphere" is not at one bar.

More interesting is what is the pressure of steam at 400c? I can't find the answer.

it is indeed a guess as a place to start based on an old timer I met who did this in the 70's.

Pumping losses are significant. 10 to 15 percent. So if nothing else, we have magically found 10 to 15 percent possible gains.

http://www.greencarcongress.com/imag...soline_ice.png

Just to clarify, by "engine internals are kept well below 200C", you mean the bottom end (and other moving oil-fed bits)...?
Obviously the inner walls of the cylinder are at considerably higher temps and, going back to the 6-cylinder engines for a moment, the ability to make work (and negating the cooling cct) is dependent on the high temperature (and quantity of heat) of the cylinder walls and piston crown.


This issue about water injection being akin to a steam engine does seem to come up quite often, but as has been said it is completely unfounded. However, the arguments for using W.I. that do make sense are all founded in cooling the charge (whether for power[turbo'd] or efficiency[pumping-loss]), yes?
So why not just use up the water as an external, evaporative cooling agent prior to the charge entering the cylinder? Would the effects be the same or better (providing the charge is not so cool as to deter atomisation)?

I'm neglecting the available heat energy available due to friction losses, and available due to heat transfer of exhaust heat energy into the engine block. Yes, that's true. Then again, my example has been very much simplified in other respects. You'll find that all of the simplifying assumptions I make are placing the example into the absolute best-case scenario, and that real-world considerations only worsen the case for steam generation.

To a layman's perspective, for instance, nothing is added to the discussion by dragging in steam tables or air tables to more accurately model exhaust gas having water sprayed into it. Treating exhaust gas as an ideal gas is the best-case scenario, and more realistic modelling by using tables would only show worse results.

Nor is the discussion helped by dragging heat transfer equations into the surrounding metal of the combustion chamber. For purposes of the example, no heat transfer is assumed because, again, it's the best-case scenario. That it greatly simplifies the example is a beneficial plus. Besides, most engine heating is due to exhaust heat radiating into the engine block. Think about it - there's an entire stroke devoted just to manipulating exhaust gas. During all the time it takes for that stroke to complete, exhaust gases are touching the sides of the cylinder, the combustion chamber roof, the valves, the piston, and the walls of the exhaust port(s). If we posit that we cool off exhaust gas by spraying water into it, then heat transfer into the engine block will be much less than before, so there is much less heat energy available anyway. This observation is borne out by the fact that the two 6-stroke engines mentioned elsewhere in this thread did not need water cooling at all.

The same goes for the actual spraying of water. In my example, I'm assuming the water is atomized into the exhaust gas such that it's perfectly uniformly mixed with the exhaust gas, so as to uniformly raise its temperature as it sucks heat energy out of the exhaust gas. Again, best-case scenario. In the real world, it's kind of tricky to get truly perfectly uniform atomization of water into a combustion chamber filled with compressed exhaust gas at TDC, and water droplets will fall out out onto the metal surfaces as the exhaust is cooled below its ability to vaporize that water.

Heat energy due to friction can be ignored since there's actually very little of it available to do anything, compared the the heat energy directly found in the exhaust gas itself. Warm-up time, remember?

These are the major assumptions I made with my example. Remember, they push the example into an impossible-to-attain ideal condition, for which worse results would have been shown had real-world considerations been taken into account, and for which the example itself would have been needlessly complicated.

That is exactly why ceramic coating is popular. However, it's also pricey to have it done correctly to engine internals, because you have to ensure that the coating is correctly bonded to the surfaces where exhaust gases would touch. Otherwise, the coating would eventually flake off and cause no amount of mischief to the piston rings. You also can only coat the surfaces that themselves do not experience frictional rubbing. Therefore, the cylinder wall (which is most of the surface area that exhaust heat leaks into) is not coated.

I know. Not all engineering types come across as dicks, either. However, as jerky as my posts come across, they're gentle love pats compared to how the real world would act. I'm not interested in gentle persuasion or cajoling, because the real world doesn't gently persuade or cajole people who engage in wishful thinking.

There are, to my knowledge, only three proven effective methods of reclaiming exhaust heat energy to do something useful. The first method involves shaping exhaust piping in such a way that the exhaust gases themselves generate a partial vacuum to help evacuate the cylinders so as to provide less dilution to charge air inside the engine. Devices that do this are popularly known as headers.

The second method involves passing hot exhaust gas under pressure through a nozzle to convert thermal energy into mechanical energy, and then recover the mechanical energy by use of a turbine wheel. Turbochargers are the best known device that uses this principle.

The third method is really only useful on engines that require a throttle valve to begin with. Exhaust heat energy is used in this case to heat up the intake air such that the throttle plate is forced more open than what would otherwise be needed, in order to lower pumping losses associated with that throttle valve.

Thermal piles are not considered because it's currently not practical to use them in real-world applications, given the state of the art.

Several companies make water injection kits that do exactly this. The idea is to push operating conditions away from where engine-destroying pre-ignition and detonation occur.

I know you believe. You believe that water will just magically flash into steam. Your belief is akin to a religious belief.

Again, you neglect the latent heat of vaporization.

Using your example of 1034 kPa and 700 K (or 427 C, if you will), and assuming that the liquid water was near 100 C, we find that:

The mass of your mystery gas in that closed volume is unknown. However, for a simplified exhaust gas containing 80% diatomic nitrogen, 11 percent carbon dioxide, and 9 percent steam, it's 0.2 grams.

It holds about 150 J of thermal energy. Of that, 91 J is available that can be used to vaporize water.

That gram of water needs 2260 J to completely turn into steam.

Roughly 4% of that gram flashed into steam, bringing the temperature of your closed system to that of the water, i.e. 100 C.

The pressure dropped, too. Ooooh, lookie - 101 kPa.

And you still have 0.96 grams of liquid water to deal with.

So, now you have 39 ccs of mystery gas, 1 cc of water, and it's all at 100 C and 101 kPa. Oh, excuse me - 212 F and 14.7 psia.

Given your example above, it's about 1/10th of the pressure you mentioned.

That's because it's undefined. Water has a critical point of 373 C and 22064 kPa.

Why do we not know the name of this old-timer? If he could get this thingy to work, then his name ought to be at least as well known as Otto or Diesel or Brayton or Rankine or Carnot or Atkins or Wankel.

So how would you go about preventing ignition quenching?

A variation on #2 that is also proven is turbo-compounding.

#3 is home-built used often on this site.


There are many other methods that have also been proven effective, however they have not been proven cost effective:

TEC Modules to recapture heat energy. IIRC one of the big 3 had created a cat covered with them that generated around 1KW average. Of course these things are ridiculously inefficient.

BMW used a steam turbine in the exhaust: BMW Turbosteamer gets hot and goes

<insert heat recovery device here> Sterling engine, etc.

I figured I would chime in on this. The boiling point of water is not always 100C. It can be less or more depending on the pressure its under. under ten atmospheres of pressure its boiling point goes up to about 180C. Also at that pressure the steam expansion ratio is less than 175 times its volume as water.

In my research into water injection on a gasoline engine it is most useful as an internal coolant. The next most useful thing it does is act as a knock suppressant allowing more boost or higher compression ratios. Lastly it can be used as a intake charge coolant to cram about 0.5 to 1.5% more air/fuel mixture into the cylinder. None of these things is extremely useful for hypermileing .

On a diesel it is a little different story when used as a charge coolant. Its documented to be useful increasing fuel economy, power output, and reducing NOx emissions. Unfortunately these effects aren't really prevalent unless the diesel engine is operating constantly over 80% of its full output. Even then its only good for an improvement of 0.5 to 2% in fuel consumption and power output.

For purposes of illustration, it's enough to show that the water will cool off the exhaust gas due to its sucking out the exhaust gas's heat energy to satisfy its latent heat of vaporization requirement. I'm not about to go into overly complicated details for somebody who does not know what water's critical point is (can't find the pressure of steam at 400 C, indeed).

Agreed. I have done some study into water injection, too. At one point, I was going to do a binary staged water injection system that would deliver approximately the same amount of water per cylinder, regardless of engine speed, for purposes of turbocharging. I even worked out the math using Excel and found out that the water charge would vaporize completely while the cylinder was on its compression stroke, while providing the cooling mentioned above.

BTW, this is why extra gasoline is used at WOT for WOT enrichment, as well as why tuners enrich the charge mixture for forced induction applications. Gasoline also provides cooling by evaporation on the compression stroke, but nowhere near as efficiently as water does. IIRC, gasoline has a latent heat of vaporization of about 2 J/g-K.