Properties of water injection in the diesel cycle

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
What effect does water injection have in the diesel cycle?

In trying to understand this, I have read up on use of water injection from WWII warplane days. I'm trying to understand exactly what happens within either the intake tract or within the cylinder when you have saturated the air with water vapor--as it pertains to either the diesel cycle or the otto cycle.

In trying to find helpful information, I went hunting for a good steam table. I have found a most useful Java applet from the National Institutes of Standards and Technology (NIST); it is linked/listed in my signature.

What I'm specifically after is a model of cylinder pressure-versus crank angle with and without water vapor. I'd like to know if water injection can allow for running aggressive boost and timing and reduce emissions at the same time.

Using the applet in isobaric properties mode, I put in:

pressure 185 bar (guess at peak pressure)
T-low (immediately prior to fuel injection)= 410 C
T-high (peak temperature) = 2177 C

for a stock 1.9L Tdi. It gives:

http://webbook.nist.gov/cgi/fluid.cgi?P=185&TLow=410&THigh=2177&TInc=10&Applet=on&ID=C7732185&Action=Load&Type=IsoBar&TUnit=C&PUnit=bar&DUnit=mol%2Fl&HUnit=kJ%2Fmol&WUnit=m%2Fs&VisUnit=cP&STUnit=dyn%2Fcm&RefState=ASH

And for a very modded Tdi reaching > 221 bar PCP

If it is possible for a modded Tdi to run > 221 bar PCP, then injected water will be in the form of supercritical steam:
http://webbook.nist.gov/cgi/fluid.cgi?P=225&TLow=350&THigh=3400&TInc=20&Applet=on&ID=C7732185&Action=Load&Type=IsoBar&TUnit=C&PUnit=bar&DUnit=mol%2Fl&HUnit=kJ%2Fmol&WUnit=m%2Fs&VisUnit=cP&STUnit=dyn%2Fcm&RefState=ASH

Moreover, the supercritical steam has lower Thermal Conductivity than normal steam beyond 1137 C which should lower heat losses to cylinder walls during that first fraction of the expansion stroke when pressures > 221 bar until temps drop < 1137 again/ and when pressures < 221 bar again:



I'm hypothesizing that if heat losses are lower during that time, it will preserve peak pressures a smidge longer. If coupled with aggressive Start-of-Injection timing, this could yeild some greater BMEP under some loads and give more useful P-V work upon the crank without the same NOx production the equivilant BMEP-sans-water-injection would produce. (see Michael Willmann's comments in reference 2.) Michael's reference 1 below goes into more detail on emissions effects from flame front temperature and rate of heat rise (RHOR)

Some good reading:

reference 1: http://www.ilot.edu.pl/Journal%20of%20KONES%202001/JOK2001%20NO%201-2/R5.pdf

reference 2: http://www.ilot.edu.pl/Journal%20of%20KONES%202002%20No%201_2/01/str275.pdf (relevant point: "there is a high amount of ...water vapor..which cools the reaction zone very well. Apart from this, the combustion process takes more time. As a result the flame temperature is lower."



Q: is PCP lower as well or is it largely insensitive to this?

Finally, the last question of mine is what effect does injecting the water pre or post-turbo have, thermodynamically?

Adding injected water should increase the heat transfer through the intercooler, making it more effective. In a modded Tdi running 2.25 bar absolute (post turbo), water will boil to steam at 125 C. Whether the temperatures post turbo ever get this high, I don't know. I have seen my stock IC temps as high as 65C on a hard run, but I have no idea what pre-IC temps were.
 
Last edited:

Gil

Veteran Member
Joined
May 16, 2004
Location
Wallingford CT
TDI
2002 auto wagon and 2004 stick
Not sure if this is true but I heard that the water injection was used in WWII aircraft for only short periods of time for extra power. Along with the water injection they leaned out the mixture for the power gain.

.
 

dieseldorf

Top Post Dawg
Joined
Oct 11, 2000
Location
MA
TDI
ex- 1996 wagon, ex-2000 Jetta
Nick:

On another clean-diesel front, Chevron is starting to expand its "Proformix" diesel-water emulsion in "selected areas," Ingham pointed out (see Diesel Fuel News 4/29/02, P5). Besides Port of Los Angeles and other customers, Chevron has also converted its own delivery trucks operating from the Montebello terminal to the emulsion fuel. "We're now looking at a marketing opportunity in Seattle, and our Chevron-Canada division is about to commission a plant in Vancouver, B.C.," he said.
 

andypandy

Well-known member
Joined
Aug 10, 2006
Location
Sweden Gothenburg
TDI
Golf MK4 -00 110hp++
Issnt it possible to scale down the technology that the tractorpullers have?

Look at this.
http://www.youtube.com/watch?v=tI2mTRPpJdc

The pullerclass is called "Super Stock"
(and the pro stock where it's only allowed with diesel as fuel)

The block is the tractor originals, they run mainly on Diesel.. and what more.. water/meth?
about 6 liters engines where they could have from 2000 - 3000hp!

That means if scaled down properly we could have 500-800hp in a 2liter machine like in a TDI.

But where are they??? :confused:
350hp i have seen/heard of.. but it seems like there is a lot more room for improvements in this area.
 
Last edited:

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
Dieseldorf,

I agree, I think there's a lot to be learned and gained here.

Can anyone comment on the discussion of water's role in reducing thermal efficiency but increasing useful work in this link?

Peter Fette said:
The thermal efficiency htherm


The thermal efficiency htherm in case of isothermal heat transmission (TE in the H- and TC in the K-cylinders are constant) and at 100% heat exchange in the regenerator is equivalent to the Carnot efficiency:
hthermC = (TE - TC) / TE
This ideal thermal efficiency naturally cannot be achieved within the real engine. The heat transmission may be good, but is just only nearly isothermal. Temperature losses can be calculated from the heat capacities of the irrigation liquid drops (and/or that of the metal chips) and from the thermal capacity of the gas. They have been proved in experiments by the author.
During the heat transmission to the gas, size and mass of the heat exchange surface, gas velocity on this surface, thermal conductivity of the gas as well as the regenerator's quality |9| are of influence for the thermal efficiency. Further the thermal -and the overall- efficiency is reduced with increasing rotation by friction losses in the liquids flow and by acceleration of the liquid masses. These mechanical losses and the temperature losses at heat transmission are kept low at slow rotation. Therefore, if a high overall efficiency is wanted, the engine is to be driven at slow rotations. Like the large diesel engines for ships, high power is attainable for this stirling engine by a large construction and high gas pressure.
If a compound fluid -gas + vapor- is chosen, the increase of pressure |10| by the steam partial pressure of the evaporating working liquid leads to more work output. Evaporating of liquid has the valuable effect, that it performs an optimal heattransmission to the gas with a very small temperature gradient. This is obtained by the very fast movement of the vapor molecules which are mixing together with the gas molecules. But also the disadvantage of compound fluid must be taken into account: It reduces the thermal efficiency within the condensing phase in the compression cylinder. The reason therefore is, that the amount of condensing energy is taken out of the process. At each cycle it must be produced new. This is well to be seen, when comparing the "frames" in Fig. 4a and in Fig. 4c which in both cases represent ideal stirling processes.
  • Fig. 4a -with gas only- leads to the carnot-efficiency. The mass of working fluid is always constant.
  • in -Fig. 4c- the "frame" is the theoretical ideal stirling process for the vapor component as the only working fluid. You see this frame is not even nearly a parallelogram as is the frame in Fig. 4a. At the isochoric cooling phase at V[SIZE=-2]max[/SIZE] the mass of vapor is condensed to liquid until an equilibrium is reached for saturated vapor at temperature TC. Only a small rest of vapor is left. The mass of this small amount of vapor is compressed at temperature TC to V[SIZE=-2]min[/SIZE]. Now new vapor will be produced during the isochoric heating phase and the isothermal expansion phase. The energy needed for the isochoric phase cannot be a regenerator supplied energy. Only the small amount of vapor left during cooling and compressing theoretically can be heated by regenerator energy. The previous condensed liquid must be removed into the expansion cylinders liquid volume; here it is heated and evaporated again. This is the reason for the reduced thermal efficiency of an engine working with compound fluids.
The computer program STMOT2: based on |11| takes into account all these losses.

Let us state: When using a compound fluid -gas + vapor- the work done per cycle is increased but this is done by reduced thermal efficiency.
But inspite of the reduction in thermal efficiency, compound fluid is valuable, whenever the cost for heatgenerating is neglectable, as it is with solar collectors or from waste heat.
press Your browsers Return button to return into previous text
<A href="http://www.stirling-fette.de/english.htm#A0">
Back to contents

The Load Exploitation Factor e i


Apart from the thermal efficiency, the load exploitation factor -sometimes called the "Schmidt Factor" - " n " is an interesting parameter. "n" is the relation of the engine's effective work "W" to the effective work of the ideal Stirling process " W[SIZE=-2]i[/SIZE] " , in the same working space between V[SIZE=-2]max[/SIZE] and V[SIZE=-2]min[/SIZE] and the same temperatures "TE" and "TC" , but with a regenerator volume of VR = 0. "W" is equivalent to the areas enclosed by the graphs in Fig. 2, Fig. 3 or in Fig. 4a.
The ideal Stirling effective work " W[SIZE=-2]i[/SIZE] " , is equivalent to the area in Fig. 2 represented by: (I) - (II) - (III) - (IV) - (I); or to the frame in Fig. 4a, both areas are created by the isotherms TE and TC, as well as the isochores at V[SIZE=-2]max[/SIZE] and V[SIZE=-2]min[/SIZE].


or

With M[SIZE=-2]gas[/SIZE] = Mass of gas, R[SIZE=-2]gas[/SIZE] = Gasconstant. V0 = expansion- + compression space, P0 = pressure, both measured in idle position.
The effective work output "Wi" of the ideal Stirling process in accordance with these equations is determined by the temperatures TE and TC, as well as by the volume ratio V[SIZE=-2]max[/SIZE] / V[SIZE=-2]min[/SIZE] that can be achieved, and it is directly proportional to the gas mass in the engine, which is determined by the idle volume V0 and the idle pressure P0. Principally the same applies also to the effective work "W" of the real engine. The volume ratio V[SIZE=-2]max[/SIZE] / V[SIZE=-2]min[/SIZE] is determined by control volume and other deadspaces, regenerator volume "VR" , the phase shift angle d and the geometric conditions of the crank gear movement, which can only aproximate the ideal movement. V[SIZE=-2]max[/SIZE] / V[SIZE=-2]min[/SIZE] and thus "W" and the degree of exploitation "n" are reduced at increasing "VR" and other deadspaces. V[SIZE=-2]max[/SIZE] / V[SIZE=-2]min[/SIZE] is also reduced, if instead of the purely sinusoidally movement of the pistons through the crank slot gears, connecting rod crank gears are used for "KU1" and "KU2". See the summary in table 1
The above mentioned load exploitation factor "n" is only usefull for comparing the work "W" of a real engine to the work "W[SIZE=-2]i[/SIZE]" to its own ideal process within the same space range of V[SIZE=-2]max[/SIZE] and V[SIZE=-2]min[/SIZE].
It would be more helpfull, if we can compare the work "W" of any real engine to the work "W[SIZE=-2]i[/SIZE] " of an ideal stirling process of a common usable ideal standard reference stirling engine. With this new basis of " W[SIZE=-2]i[/SIZE] " -now better named: " W[SIZE=-2]ir[/SIZE] " we should define a new load exploitation factor: " e i " The following proposal may be discussed for the definition of a common usable ideal reference stirling engine, and for the conditions of its process calculation:
  • The ideal work "W[SIZE=-2]ir[/SIZE] " of the reference engine should be calculated with the equation above.
  • TE is the maximum temperature, which can be measured on the surface of the real engines heater -if the engine is heated by a burner or by radiation- or, if heated by a liquid, TE is the liquids temperature at the entry to the engines heatexchanger.
  • TC is the minimum temperature within the engines cooler. That is the temperature of the coolant at its entry into the cooler for the engines working fluid. Or TC is the minimum temperature of the coolers surface if the engine is cooled by radiation.
  • The mass of the working fluid is the mass of the gas which is constant within the real engines spaces including all dead spaces. If compound fluids are used, only the mass of the gas component should be taken, otherwise the above formula cannot be applied.
  • The space ratio V[SIZE=-2]max[/SIZE] / V[SIZE=-2]min[/SIZE] must be a constant standard value. Here we must set statements, otherwise there is no possibility to have a common usable standard for a reference process. There may be other definitions to define a constant space ratio, but
I know the Stirling cycle differs in that it is closed (non-exhausted volumes), but can any of this discussion help us understand what role water injection plays in either the otto or diesel cycles? I've highlighted the parts of the quote which relate to my question. Specifically, is the statement in red correct for diesel and otto cycles as well?
 
Last edited:

GoFaster

Moderator at Large
Joined
Jun 16, 1999
Location
Brampton, Ontario, Canada
TDI
2006 Jetta TDI
It may be true, but there are a lot of other factors to consider that complicate matters tremendously.

As soon as you have a two-phase mixture (gas + liquid), the act of compressing it requires heat transfer between the gas and liquid phases. (The gas is compressing and its temperature goes up, the liquid is absorbing some of that heat, this implies heat transfer.) Heat transfer is an "irreversible" process, i.e. there is a loss of entropy due to heat transfer always being across a temperature gradient. A loss of entropy implies a loss of efficiency in the case of an engine (because you can't get back the loss associated with the heat transfer). If the surface area is enormous (i.e. extremely fine atomization with very even distribution inside the cylinder - Note that high swirl as in a diesel tends to centrifuge-out liquid particles) then perhaps the losses could be negligible.

If you compress what starts out as a mixture of air and water mist, the compression will have a lower ending temperature. I do not know if the pressure will end up lower (due to the lower temperature) or higher (due to the extra space that the water vapor takes up compared to the original liquid) - I suspect it will end up being lower. Thus there is less mechanical work done during the compression stroke - but it means you won't get back as much work during the expansion stroke. The complicating factor is that the reduced temperature and pressure at the end of compression means you have room for burning more fuel before reaching mechanical cylinder pressure limits.

On my previous vehicle, I fiddled with water injection for a while, but my intent was to only inject just enough water to overcome the imperfect intercooling, not enough to significantly affect what happens during the compression stroke. For that, there is a potential power increase due to higher density (lower temperature) of the intake air.
 

Waldek Walrus

Veteran Member
Joined
May 15, 2006
Location
Central Pennsylvania
TDI
2006 Beetle TDI
nicklockard said:
Dieseldorf,

I agree, I think there's a lot to be learned and gained here.

Can anyone comment on the discussion of water's role in reducing thermal efficiency but increasing useful work in this link?

I know the Stirling cycle differs in that it is closed (non-exhausted volumes), but can any of this discussion help us understand what role water injection plays in either the otto or diesel cycles? I've highlighted the parts of the quote which relate to my question. Specifically, is the statement in red correct for diesel and otto cycles as well?
Nick,

In an Otto cycle engine the water injection keeps the combustion chamber temperatures low enough to prevent pre-ignition. This was used in WWII aircraft engines to allow more supercharging to get more power without destroying the engines through pre-ignition. This is still used with heavily supercharged or turbocharged gassers to get the last bit of power out for racing.

In the Diesel engine you won't get pre-ignition, so this use is not needed. Adding a bit of water can limit peak combustion temperartures and prevent melting things like valves and injector tips when running at ultra-high power for a short burst (i.e., the tractor pull engines noted earlier).

In both of these there may be a tiny bit of extra work done by the expansion of the water vapor during the power stroke. This is TINY. The gain from the water is its cooling effect to keep the working surfaces from melting. The extra power comes from being able to burn more fuel in the same size engine.

The Stirling cycle is different. The gain in power is from the extra expansion of the water vapor during the power stroke. It costs a LOT of energy to get that water vaporized under pressure. Most of that energy is lost to condensation at the end of the power stroke. This is basically a steam engine. An eye-opening exercise on this would be to search for the term 'boiler horsepower' - this is the energy to turn ten pounds of water into steam at atmospheric pressure and that is the amount of steam early steam engines needed to make one horsepower. Hence the idea this is only good if you can get 'free' energy.

Does this help at all?
 

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI
Nick, your first problem is assuming that combustion in a high-speed Diesel engine occurs at constant pressure. Tsk tsk tsk. :)
 

EddyKilowatt

Veteran Member
Joined
Mar 1, 2006
Location
Carmel Valley CA
TDI
2003 Golf GL 5M
TDIMeister said:
Nick, your first problem is assuming that combustion in a high-speed Diesel engine occurs at constant pressure. Tsk tsk tsk. :)
Aw shucks... I think my first-year thermo textbook tried to warn me about that, but all I can remember is Otto = constant-volume combustion, Diesel = constant-pressure combustion.

Guess it is time to pull out The Tautology of Dave (courtesy of an intinerant engineer I know): "In theory, theory and practice are the same. In practice, they are not." :eek:

Eddy
 

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
Dave (and Dave,)

Yes, I know the modern diesel cycle is probably a hybrid cycle: 60% CP and 40% CV. But with the late late laaaaaaaaaaaate fuel injection hoohaa being pursued in order to kill the mighty mighty NOx monster these days, the cycle is heading back toward's a CP heat addition process, is it not?

What if we were to directly inject water mist/fog inside the cylinder, with a separate high pressure injector? Then, like that hot rod fellow who has proposed eliminating the cooling system and injecting water, we could dramatically lower the size of the radiator and cooling system components: most heat in the diesel cycle happens at high loading: precisely when you would want to inject water for NOx control and for keeping PCP's under control. By dialing down the A/F ratio, smoke could be avoided. The water injection could be mapped to occur at optimal times and delivery volumes/durations to minimize NOx--it looks like, according to Michael Willmann's papers, that staying below 1800 C PCT's entirely avoids NOx, and staying below 2400 C is a practical way to minimize NOx. Instead of retarding timing and wasting fuel to kill this big bad monster, why not just prevent it from forming in the first place? Throwing fuel down the cylinder after the exhaust valve is open is plain dumb, as is lowering the compression ratio. I say raise CR, advance timing, and kill NOx with peak temperature control via water injection. We're slowly seeing all the advantages diesel has being sapped away with the current path.

I think that Crower guy is onto something, but you don't need to add a fifth cycle. However, maybe it could be combined with an Atkinson cycle usefully.
 
Last edited:

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
GoFaster said:
It may be true, but there are a lot of other factors to consider that complicate matters tremendously.

As soon as you have a two-phase mixture (gas + liquid), the act of compressing it requires heat transfer between the gas and liquid phases. (The gas is compressing and its temperature goes up, the liquid is absorbing some of that heat, this implies heat transfer.) Heat transfer is an "irreversible" process, i.e. there is a loss of entropy due to heat transfer always being across a temperature gradient. A loss of entropy implies a loss of efficiency in the case of an engine (because you can't get back the loss associated with the heat transfer). If the surface area is enormous (i.e. extremely fine atomization with very even distribution inside the cylinder - Note that high swirl as in a diesel tends to centrifuge-out liquid particles) then perhaps the losses could be negligible.
Hi Brian,

If we presume that all of the heat tranfer to the droplets from the hot compressed N2 and O2 occurs prior to fuel injection (let's define this as step A,) then entropy will be conserved, since there is no chemical reaction, only a state change, correct? We will have used up some enthalpic energy of hot air, which is a poor carrier of enthalpy anyway, IIRC. However, we may recover that enthalpic energy once the fuel is injected, if we can get the steam to flash into the supercritical steam region: it will do as much useful work as it took to reach step A. Now, since the water molecules are crowing the cylinder just prior to fuel injection, the kinetics of combustion will be slowed because there will be fewer collisions between oxidizers and hydrocarbons, but that might actually be a good thing! It might allow for a more complete droplet vaporization prior to combustion. This seems like yet another path toward HCCI, but would require higher compression ratio's and more advanced timing in order to get PCP's sustained over the supercritical steam threshold of 221 bar.

In practical effect, it should mean that PCP is about the same, but extended over a longer time domain (and doing more P-V work upon the crank for a given amount of fuel,) but peak temps are held lower, which is obviously good for NOx reduction.

This seems like it could be another path toward HCCI, but would require higher compression ratio's and more advanced timing in order to get PCP's sustained over the supercritical steam threshold of 221 bar. Higher CR's and more advanced timing have obvious fuel efficiency benefits too.

There is also some work which suggests OH radicals chew up soot, IIRC. Supercritical steam should be chock full of radicals at such high pressures and temps, but I'm only guessing.

GoFaster said:
If you compress what starts out as a mixture of air and water mist, the compression will have a lower ending temperature. I do not know if the pressure will end up lower (due to the lower temperature) or higher (due to the extra space that the water vapor takes up compared to the original liquid) - I suspect it will end up being lower. Thus there is less mechanical work done during the compression stroke - but it means you won't get back as much work during the expansion stroke. The complicating factor is that the reduced temperature and pressure at the end of compression means you have room for burning more fuel before reaching mechanical cylinder pressure limits.

On my previous vehicle, I fiddled with water injection for a while, but my intent was to only inject just enough water to overcome the imperfect intercooling, not enough to significantly affect what happens during the compression stroke. For that, there is a potential power increase due to higher density (lower temperature) of the intake air.


Yes, I agree. Essentially what you did was repeat this: " But also the disadvantage of compound fluid must be taken into account: It reduces the thermal efficiency within the condensing phase in the compression cylinder" in other wording.

Yes, I can see how water injection can reduce thermal efficiency, but still, it can allow for more useful PV work--IOW: it is a direct corrolary to a turbocharger! A turbo decreases VE by backpressure on the exhaust, yet it allows the engine to act like a larger engine--on demand. In effect it is a mock variable displacement engine. It seems that water injection does the exact same thing, but instead of reducing VE or adding backpressure, it reduces TE (for short duration under higher loads, just as a turbo does) while producing more work, on demand.

Now, I am still not sure of this. It all depends on if that statement in red in the quoted section of post #5 is true or not. Can you or Dave or anyone expound on that?

Also, that author mentions a 'load exploitation factor (Schmidtt factor.)" Does this apply equally to all heat engines, and what do you make of it?
 
Last edited:

jnecr

Veteran Member
Joined
Oct 9, 2004
Location
Raleigh, NC
TDI
2014 BMW 328d
nicklockard said:
There is also some work which suggests OH radicals chew up soot, IIRC. Supercritical steam should be chock full of radicals at such high pressures and temps, but I'm only guessing.
Nick!

Theoretically the water would begin to split in to OH and H at such high temperatures. But I think you'll find that at such a short amount of time it just won't happen. Plus, if you think the equation out, what would you get. Well you'd get HOH going to HO and H and then the H would quickly find another H and then just as quickly combust (provided that the diesel hasn't used up on the O2), making what? HOH. Yeah, that ain't gonna happen.


Anyhow, back to the rest of it. I really think you're barking up the wrong tree here. Water injection is meant for one thing:
Waldek Walrus said:
In an Otto cycle engine the water injection keeps the combustion chamber temperatures low enough to prevent pre-ignition. This was used in WWII aircraft engines to allow more supercharging to get more power without destroying the engines through pre-ignition. This is still used with heavily supercharged or turbocharged gassers to get the last bit of power out for racing.
All water will accomplish is the lowering of chamber temperatures. That's what it's meant for! I've read the thread and I'm still not sure whether you are trying to lower temps or not? If you are, then bingo! you found it! :D
 

GoFaster

Moderator at Large
Joined
Jun 16, 1999
Location
Brampton, Ontario, Canada
TDI
2006 Jetta TDI
Lowering the temperature nowadays also can serve the purpose of reducing NOx emissions.

Re entropy/enthalpy: Entropy involves more than just combustion. As soon as you have a degradation of a high-grade energy into a lower-grade energy there is a loss of entropy. If you compress a two-phase mixture of air and water droplets, the heat transfer from the air (which is going up in temperature as it compresses) into the droplets (which stay at a temperature that corresponds to the boiling temperature as the pressure changes - slower rate of increase with compression) there is a degradation from a higher-grade energy (higher temperature) to a lower-grade energy (lower temperature).

Re injecting water at the end of compression stroke: In my mind, this is thermodynamically the wrong way to do it. Doing it this way maximizes the loss in entropy - it means the full work of compression is done to compress the air, and you throw a good portion of it away by soaking up that heat by squirting in water. If you are going to mix water and air with the objective of reducing peak cycle temperature, you might as well use the opportunity to reduce the work of compression - by creating a mixture of air and water mist at the end of the intake stroke.

Keep in mind that if the objective is to still operate on some semblance of a compression-ignition cycle, there is a limit to the extent to which the peak temperature can be reduced before you have a misfire.
 

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
jnecr said:
Nick!

Theoretically the water would begin to split in to OH and H at such high temperatures. But I think you'll find that at such a short amount of time it just won't happen. Plus, if you think the equation out, what would you get. Well you'd get HOH going to HO and H and then the H would quickly find another H and then just as quickly combust (provided that the diesel hasn't used up on the O2), making what? HOH. Yeah, that ain't gonna happen.
J,

The possible reactions are:

HOH <--> H(rad) + OH(rad)

H(rad) finds OH(rad) and the reaction is reversed
H(rad) finds H(rad) and forms H2 and combusts with O2 making water
OH(rad) finds other radicals, like carbene centered radicals or partial decomposition radicals of fuel, or other reaction intermediates and forms a new radical.

What the radicals (if they can form in this time scale) react with is purely statistically driven. It has been estimated by many others that the average combustion event in an IC engine has thousands of reaction intermediates, many of them radicals.

Net reaction of combustion is simply: fuel + O2 --(irreversible)---> H20(gas) + heat energy.

The amount of heat is the total enthalpic change, which is path independent. It doesn't matter if you have a million intermediate reactions driven by radical chain reactions or if you only have the one primary reaction, the end result will be the same.

Now, you may be right about the time scale thing. I'll have to look into that. However, it is well known that combustion produces many thousands of radicals from fuel and O2. Radicals, in general, participate in reactions with much higher rate constants due to the harpoon mechanism. At these temperatures, I presume that collisional reactions dominate.

Ultimately, the reaction kinetics determines what reactions can happen. If water formation is much faster kinetically than other reactions, then yeah, that's all you'll get, but if not, then the radical reactions of OH(rad) will proceed.

Jnecr said:
Anyhow, back to the rest of it. I really think you're barking up the wrong tree here. Water injection is meant for one thing:


All water will accomplish is the lowering of chamber temperatures. That's what it's meant for! I've read the thread and I'm still not sure whether you are trying to lower temps or not? If you are, then bingo! you found it! :D
I'm wondering if there is a better way to control NOx than wasting fuel and losing power. And, if soot can be reduced at the same time, that is a win win win situation.
 
Last edited:

GoFaster

Moderator at Large
Joined
Jun 16, 1999
Location
Brampton, Ontario, Canada
TDI
2006 Jetta TDI
I've heard (through Tdimeister - ask him!) that water injection can indeed be used to reduce NOx - and the good thing is that a water injection system can be used all the way up to full engine load, which can't be done with EGR without significant de-rating of the engine's power output.
 

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
Excellent.

So water injection in the diesel cycle may not do much for increasing power except this: it may allow you to advance timing back toward optimal (10 deg BTDC, or perhaps even earlier since the water molecules will crowd out the O2 molecules and slow reaction kinetics some) and increase compression ratio to get back to the diesel ideal efficiency.

It might actually complement EGR well: high EGR flow rates at low load in complement to variable and ECU-mapped water injection at all loads, so you don't have to retard timing, except at the extrema.
 

jnecr

Veteran Member
Joined
Oct 9, 2004
Location
Raleigh, NC
TDI
2014 BMW 328d
The other thing that gassers do is methanol injection. You discussed that at all? Again, it will aid in cooling, not sure what else it would do though.
 

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
J,

The methanol is an added fuel. The cooling effect allows one to cram more fuel into the cylinder yet get lowered PCT's.
 

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
I chatted by PM with Dave (TdiMeister) and he had some good inputs.

If you read up on supercritical steam powerplants, you'll see that what I'm proposing is neither new nor novel. I've been doing a lot of reading. Here are some interesting sources:

http://www.lulu.com/items/volume_13/...eview-thpb.doc

http://peswiki.com/index.php/Heat_engine read last paragraphs.

http://www.sgrajeev.com/geometry-of-thermodynamics/

http://www.uet.edu.pk/ht2/steam.htm

Water injection holds promise to do a lot more than simply reduce EGT's, although it might be worth it for that reason alone.
 
Last edited:

MethylEster

Veteran Member
Joined
Sep 16, 2004
Location
Lakeville, MN
TDI
Jetta, 1999.5. Red, 5 Spd
Nick,

Where do you cross into the territory of degrading the lubricating aspects of the liquid diesel fuel within the cylinder walls themselves ? And possibly start scoring the cylinder walls. In other words, what's the upper limit to the amount of water you can safely add ?

Also any thoughts on what additional acidic species would be generated and further degrade the lubricating oils ?
 

GoFaster

Moderator at Large
Joined
Jun 16, 1999
Location
Brampton, Ontario, Canada
TDI
2006 Jetta TDI
That's likely to be different on a case-by-case basis. Dave / Nick - thanks for the tech papers, lots of interesting stuff in there. Nick - Those two links are great!

One thing that seems apparent from the tech papers is that if you use direct water injection (late in compression stroke, either water/fuel emulsion or separate high-pressure injection) in an engine that starts off with a good combustion chamber and injection system design, it degrades the thermal efficiency, probably for the reasons described in an earlier post. None of those tech papers described the situation more common in the aftermarket ... water injection into the intake air.

The first link that Nick posted above is for such a system (on a gasoline engine with multiport fuel injection), basically using it the way I used it when I played with this - as a substitute for an intercooler. That link may be a sales brochure ... but it's a good one with lots of information. For that application, very low flow rates are required, and the intent is to have very little atomized water left by the time the compression stroke starts. In that case, cylinder wall effect should be minimal. Also, in those applications, water injection only happens at full load, and with a warmed up engine, the cylinder wall temperature is likely to be above 100 C anyhow.

If you go *slightly* beyond the amount of water required to get to the wet-bulb temperature, the effect will be to absorb some of the heat during the compression stroke ... in effect, this reduces the "low" temperature of the heat engine, and that is thermodynamically a good thing to be doing as long as you don't overdo it to the point of causing combustion problems.

The RSR paper that Nick posted suggests the water-to-fuel ratio is between 12.5% and 25% (i.e. water delivery rate is one-eighth to one-quarter of the fuel delivery rate). That sounds about right. That document gives a whole bunch of rules for implementing water injection systems - and having been through this exercise, they all make sense.

The second link that Nick posted is for large marine engine emission reduction ... and they have water-injection systems both for intake air and for direct water injection.
 

TDIMeister

Phd of TDIClub Enthusiast, Moderator at Large
Joined
May 1, 1999
Location
Canada
TDI
TDI
Interesting that marine engines and Wärtsilä were mentioned. I attended a seminar last week presented by the general manager at Wärtsilä on the future development of large-bore engines. Interesting. :) Those two-stroke suckers are pushing 55% BTE running on bunker oil.
 

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
Hi fellas,

I'm really considering getting a membership/subscription to www.diesel.net.

Can anyone tell me if they are as good a technical resource as they appear? They consistently come up on google and scholar.google searches on diesel technical topics. There are a few papers I'd love to get my eyeballs on :)
 

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
MethylEster said:
Nick,

Where do you cross into the territory of degrading the lubricating aspects of the liquid diesel fuel within the cylinder walls themselves ? And possibly start scoring the cylinder walls.
Did you mean to say lubricating oil? Well to be truthful, part of the goal of this thread is to gather enough data on pressure-vs-crank angle and temperature-vs-crank angle to help me build a model in Excel.

I *think* that water injection under precisely timed, quantity-controlled delivery can be done in such a way that you never have wet steam in the cylinders. But, I need to build a reasonable model first. At this point I make a guess that we could have the droplets flashing to dry steam moments before fuel injection (never reaching the walls), heating to supercritical steam moments after, then staying far above the condensation point throughout the rest of the downstroke.

However, it would have to be mapped to engine running temperatures and air temperatures as well. There'd be no point in doing it in a cold engine. Wet steam could degrade the oils by hydrocracking. I don't know what effect dry steam would have. Suffice it to say that in an OEM application to control emissions, the oil would have to be compatibility engineered.


MethylEster said:
In other words, what's the upper limit to the amount of water you can safely add ?
That's what I need some data for: to do heat calculations. As long as you don't try to extract too much heat too quickly and kept the injected water amount below that threshold, I'm confident you would never wet the cylinder walls or any components inside the combustion path/exhaust path.
 

Scott_DeWitt

Vendor
Joined
Apr 7, 2004
Location
Texas USA
TDI
2000 Audi A4 1.9TDI quattro
When introduced into the hot intake stream of a diesel water flashes to steam which requires alot of heat. So your intake temps are decreased, to a point. Methanol is typically added because water/steam takes up a certain amount of space and contributed nothing towards energy release during combustion.

However too much water can cause it to pool in the intake tract, and lower the temperature of the air in the combustion chamber to a point where diesel ignition won't occur.
 

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
GoFaster said:
That's likely to be different on a case-by-case basis. Dave / Nick - thanks for the tech papers, lots of interesting stuff in there. Nick - Those two links are great!
Thanks! Yeah, I'm pleased that there appears to be a fair amount of good research in this area already--it already has some momentum.

GoFaster said:
One thing that seems apparent from the tech papers is that if you use direct water injection (late in compression stroke, either water/fuel emulsion or separate high-pressure injection) in an engine that starts off with a good combustion chamber and injection system design, it degrades the thermal efficiency, probably for the reasons described in an earlier post. None of those tech papers described the situation more common in the aftermarket ... water injection into the intake air.
Yes, I believe that in a direct-injection situation, TE is sacrificed but total PV work is increased while NOx and soot reduction is maintained for peak loads--highly analogous to a a turbo :)

But in the aftermarket case as we commonly find, I'm still struggling to understand its full implications. Obviously, it helps lower the knock limit in SI engines, but in general, in a heat engine, what effect on TE or VE does it have? My suspicion is that it would lower TE a fair amount, increase VE a tad, but provide no PV work increase than in the direct-injected case.

GoFaster said:
The first link that Nick posted above is for such a system (on a gasoline engine with multiport fuel injection), basically using it the way I used it when I played with this - as a substitute for an intercooler. That link may be a sales brochure ... but it's a good one with lots of information. For that application, very low flow rates are required, and the intent is to have very little atomized water left by the time the compression stroke starts. In that case, cylinder wall effect should be minimal. Also, in those applications, water injection only happens at full load, and with a warmed up engine, the cylinder wall temperature is likely to be above 100 C anyhow.

If you go *slightly* beyond the amount of water required to get to the wet-bulb temperature, the effect will be to absorb some of the heat during the compression stroke ... in effect, this reduces the "low" temperature of the heat engine, and that is thermodynamically a good thing to be doing as long as you don't overdo it to the point of causing combustion problems.
Well put.

GoFaster said:
The RSR paper that Nick posted suggests the water-to-fuel ratio is between 12.5% and 25% (i.e. water delivery rate is one-eighth to one-quarter of the fuel delivery rate). That sounds about right. That document gives a whole bunch of rules for implementing water injection systems - and having been through this exercise, they all make sense.

The second link that Nick posted is for large marine engine emission reduction ... and they have water-injection systems both for intake air and for direct water injection.
The interesting things about that Wartsila application for me are:

a. Ships are under increasing regulations on port emissions. It's especially interesting that they gravitated toward this idea...IOW: it must have quite a bit of merit, or they wouldn't be wasting any time nor resources with it.

b. It got me thinking over the weekend about the possible effects of injecting water post-combustion:eek:

Okay, going to fundamentals, a heat engine such as our CI engines throws a lot of heat away, in the form of warmed up coolant, basically just to protect the metallurgy from melting down. If you could pre-inject just a smidge of water prior to fuel injection for the previously stated goals, then post inject a butt-load of water (at high throttle applications only!), you could extract a lot more useful heat and turn it into steam pressure, which would drive the turbine exducer harder for a given temperature. Obviously, you'd want to always stay below the wet bulb temperature and not actually wet the cylinder, exhaust tract (think fully warmed up exhaust header at 700C or hotter) or turbo. However, I think that the rapid pressure drop across the exducer would cause the water to cool and condense post turbo.

This would allow you to kill several birds with one stone: keep NOx under control in the high-load regime (pair it with EGR for low load NOx control,) keep PCT's under control, control high load heat loadings on engine, extract more useful PV work (at the expense of temporarily lost TE.)

Now, this should allow the engineer to size the coolant system size and capacity to only handle engine loads up to say, 75% of full throttle applications. You could potentially offset some or most of the weight penalty of carrying several gallons of deionized water.

A final consideration is that the dI water could be preheated and pre-pressurized by (strictly the heat from) post-turbo exhaust. If BMW is trying to develop a micro-turbine to recapture waste heat, wouldn't this be more direct and effective?
 

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
Scott_DeWitt said:
When introduced into the hot intake stream of a diesel water flashes to steam which requires alot of heat. So your intake temps are decreased, to a point. Methanol is typically added because water/steam takes up a certain amount of space and contributed nothing towards energy release during combustion.
If you use the NIST applet or any good steam tables, you find that water injection in the intake tract can *at best* hold temperatures below the boiling point of 125 celcius (for 2.25 bar, absolute,) which is really not that great. Water/methanol can depress that to around about 85 C (guess.)

The methanol should definitely burn, right? It should give back some energy of combustion I think.

Scott Dewitt said:
However too much water can cause it to pool in the intake tract, and lower the temperature of the air in the combustion chamber to a point where diesel ignition won't occur.[/
Yes, and it could also be dangerous way to hydrolock the engine if a poor control method is used.
 

nicklockard

Torque Dorque
Joined
Aug 15, 2004
Location
Arizona
TDI
2010 Touareg Tdi w/factory Tow PCKG
TDIMeister said:
Interesting that marine engines and Wärtsilä were mentioned. I attended a seminar last week presented by the general manager at Wärtsilä on the future development of large-bore engines. Interesting. :) Those two-stroke suckers are pushing 55% BTE running on bunker oil.
:eek: :) Ooooh, tell us more!
 
Top