Alcohol / Water Injection Questions!

[Turblown]

I have searched and not found and exact answer...

I'm looking to max out the 14b by using Alcohol / Water Injection. I am assuming that I will need to upgrade fuel injectors first, so the injection kit will fall after fuel injectors in my "upgrade path." Many on this site say that the 14b is less efficient past 17~19psi due to it heatsoaking. I would like to run 22psi+ if possible. Here is my question:

Will a standard Alcohol / Water Injection kit make it possible to run 22psi+ safely?

OR... if I added a Pre-Turbo nozzle in the intake pipe (just an idea) would the 14b be more efficient at higher boost (22psi+) levels?

Thanks for your input!
-Turblown

 

[Turblown]

Are any of you W.I. guys using a MAF+T in a blow-thru setup?

From what I've read about W.I. with VPC or MAF+T, the water has atomized by the time it hits the sensor, so it is just read as cool air.

Also, how do you determine how much flow a nozzle will have or is there a chart out there?

http://www.darklightning.com/gi/wi.htm

 

[4pistons]

I'm considering using two nozzles, one near the TB elbow, and one before the intercooler.

I've been on the DSM Water Injection List for almost two years now and the concensus seems to be that just a properly sized nozzle at the TB is sufficient. Several trials and theories have been used to also inject before the IC but no one has any data that shows the complexity is any more effective than the single nozzle.

Detonation suppression happens in the combustion chamber when the water is vaporized, so wherever it's injected doesn't really matter. Further up the pipeline gives a little more time for the water mist to cool the intake charge but think about it. At WOT, where water is injected, the airstream is moving very fast. Whether the nozzle is at the TB or 4 feet further up the piping is a matter of only a split second.

If you're using a MAFT I'd try to keep the nozzle downstream of it just to be sure it doesn't mess anything up.

Rick - '91 GSX

 

[Jehu]

I'm trying not to hijack this thread, but one last point. You are confusing the presence of air with the presence of pressure. Consider that at WOT on an NA car, the entire intake is under vacumn. The vacumn is what motivates air movement. There is still lots of airflow obviously, but it's all under vacumn. Under light throttle, the turbo does not contribute in moving air. It is instead more of an obstruction, which ofcourse required more negative pressure downstream of the compressor wheel to maintain the airflow required.

As far as having the water evaporate further upstream, even though the time frame is a fraction of a second at those flow velocities, the small sized droplets, especially under the 10micron range, will evaporate very quickly. Is it as beneficial as far as knock reduction goes to have water evaporating only to provide charge air cooling as opposed to having droplets in the combustion chamber? No.

I would consider multiple nozzles, situated close to the throttle body, with different injection turn on pressures to be the ideal set-up. Smaller nozzles have the advantage of better atomization, and setting up a small secondary nozzle to come on at low boost levels prior to the activation of the primary nozzle acts as a buffer to smooth out the transition.

 

[Turblown]

Any comments on what would happen if a nozzle was placed pre-turbo?

I know that it would effect the engine the same way, as far as reducing the knock count is concerned, but would it add to the effectiveness of the turbo at all?... maybe extend the life of the turbo by keeping it from heatsoak?

Stupid idea? ... what do you think?

 

[FireyIce01]

I want to see what you run on the stock injectors. I have a 1995 awd 14b'd and I'm looking into water injection (washer fluid though, not alcohol, not water strictly). I'm leaning heavily towards w/i instead of 550's. 550s will come in due time though. I feel the benefits of the w/i; steam clean engine, run higher boost, lower cooling temps; outweigh the benefits of doing injectors 1st. Since i only plan on running 20psi on the 14b...well, its more of being forced to since the bastard spikes to 20 now with my downpipe on even set at 10psi on a h-mbc....it should be adequate. Let's see how those 450s go!! also, post your afc or logging results if you have them; i'll need a reference for my afc-2. thanks!

If you're not going to upgrade injectors right away, think a little more about putting alchy in the injection, as the more boost you run, the more air is being put in your engine... with more air, more fuel is needed, and with alchy injection you'll be adding that fuel, wheras washer fluid won't burn, and even though it lowers the charge temperature, you still risk running too lean. It may not be enough to make a difference, but watch for knocks and temps nonetheless

 

[Shadowfax]

If you're not going to upgrade injectors right away, think a little more about putting alchy in the injection, as the more boost you run, the more air is being put in your engine... with more air, more fuel is needed, and with alchy injection you'll be adding that fuel, wheras washer fluid won't burn, and even though it lowers the charge temperature, you still risk running too lean. It may not be enough to make a difference, but watch for knocks and temps nonetheless

true. The only reason I will run the alchy now though, is that I have it left over from my Stealth. I won't get nearly as far with the alchy and 450's and a stock IC than I would with 550+ and a FMIC.

Alchy/Water injection is good, it keeps things clean, and it does help make up for either a slight lack of fuel, slight in eff IC. Mostly it helps make 93 octane feel like 100

However most of the GN guys will run Methanol, or Denatured to do so. I run denatured.

as for PRE MAF - not a good idea when you get into methanol I believe it likes to eat plastic

 

[Defiant]

Any comments on what would happen if a nozzle was placed pre-turbo?

Blade erosion. Bad. The water feels like a sandstorm to the impeller.

I know that it would effect the engine the same way, as far as reducing the knock count is concerned, but would it add to the effectiveness of the turbo at all?... maybe extend the life of the turbo by keeping it from heatsoak?

The oil and water-cooler prevent heatsoak of the turbo.

Stupid idea? ... what do you think?

No, just not up on all the odd little details. It sounds better, because you'd also be getting the "mixing" effect of the turbine... except it won't. Not in the way we'd hope.

 

[4pistons]

even prior to the throttle body, under part throttle situations you do see vacumn

BEHIND the TB vacuum is the highest at idle (discounting decel) decreasing as the throttle is opened, and lowest at WOT. Since we were talking about WI, and it is typically injected in the intake pipe BEFORE the TB, under part throttle there is no measurable vacuum. Whether you're talking Turbo or N/A, at part throttle the airflow requirement of the engine is nowhere near the flow capactiy of the intake piping and so no vacuum can exist. Air flows from the high pressure area - the atmosphere - to the low pressure area - the intake manifold. But that does not mean there is a vacuum along the route.

Under light throttle, the turbo does not contribute in moving air

It sure does! Part throttle, under load, (not talking about off-idle) when that impeller is spun by exhaust gas energy, it is moving more air than if it wasn't there, unless it is terribly undersized.

Is it as beneficial as far as knock reduction goes to have water evaporating only to provide charge air cooling as opposed to having droplets in the combustion chamber?

The only place to suppress detonation is in the combustion chamber. Any evaporation that takes place in the intake only cools the air. I agree the argument could be made that that reduces knock too but it only reduces the air temp and the threshold for when detonation occurs, not detonation itself.

setting up a small secondary nozzle to come on at low boost levels prior to the activation of the primary nozzle

Like I said before, there is no evidence that a two nozzle system is more effective than a single. That's not my opinion, but a result of seeing data from many people who've tried it.

When you whack open your throttle watch how fast your boost goes from, say 6 psi, where most system start to trigger, (which is before any detonation usually occurs. If you have detonation at 6 psi you have huge issues!) to max boost - it is very short! Less than One second?? That is hardly enough time to "smooth out the transition" and there is really no transition needed. The first few miliseconds of WI there is only a small amount of water being injected and increases to max as the water pressure comes up. There isn't a harsh, abrupt "full spray" of water. And it starts before the detonation threshold, if you're tuned properly. No transition needed.

it does help make up for either a slight lack of fuel

Be Careful!! As a general rule of WI the amount of water/alky to be injected is about 15% - 25% of fuel. If you're running 12:1 air/fuel ratio that works out to less than 3% of your fuel volume. Then assume a 50/50 alky/h2o mix and you're at 1-1/2% of alky! That can't make up for injectors that are too small! Water Injection is not a substitute for larger injectors.

Rick - '91 GSX

 

[Jehu]

BEHIND the TB vacuum is the highest at idle (discounting decel) decreasing as the throttle is opened, and lowest at WOT. Since we were talking about WI, and it is typically injected in the intake pipe BEFORE the TB, under part throttle there is no measurable vacuum.
Rick - '91 GSX

Not true. The point is, the airflow at part throttle flow situations is motivated by manifold vacumn. That means the air at entry point is at atmospheric, i.e. intake air filter, and the pressure drop, i.e. vacumn in this case, increases all the way to the maximum value in the manifold. It's simple fluid dynamics...

As for the compressor wheel turning fast enough at part throttle, I still don't think there's enough exhaust gases flowing to get the turbine/compressor to spin fast enough to move more air than the engine itself moves. If you hit WOT from a dead stop, boost onset is pretty sudden once a threshold is reached. It's not a steady increase, which suggests to me that under low exhaust flow conditions, the compressor does not aid flow to any degree...

I don't know about your system, but with a solenoid, check valve right before the nozzle and system pressure kept at ~90psi, the onset of water injection is pretty instantaneous once the solenoid is triggered. The potential of bogging increases as the water injection amount goes up. I think it makes sense to go to multiple nozzle set-up once you get into larger water requirements, eg 400cc/min plus. I mean, the largest nozzle aquamist sells only goes up to 310cc/min at the "race pump" pressures. You'd HAVE to use multiple nozzles anyway .

Don't think of water injection as being able to make up for a slight lack of fuel, but think of it as allowing your to lean things out a little...

Oh, and water before the intake is a BAD idea. Light gauge Al at 100,000+ rpm and water droplets don't mix...well.

 

[4pistons]

vacumn in this case, increases all the way to the maximum value in the manifold

No, it doesn't because the TB restrict the flow into the manifold.

The only way to create a vacuum in a fixed space at the same temperature is by removing air faster than it can be replaced. Are you trying to say that at PART throttle the air filter can't flow enough to replace the air the engine takes in? If that were true how could it POSSIBLY flow enough air to make 200 or 300 or 400 horsepower at WOT?

If a fixed space flowed exactly as much air in as was sucked out, air would flow through it but there would NOT be a vacuum in the space. Air can flow from a high pressure area to a low pressure area without following a "path" of vacuum to move it along. If you had ultra-precise measuring equipment you'd only be able to measure some amount of vacuum in the area where the air entered into the vacuum source, i.e., at the Throttle Body. For our practical purposes it is imperceptable and immeasurable anywhere else.

Here's an extreme example but may help to illustrate. Put the most powerful Shop Vac you can find outside and put the hose through a window in your room. Now go open the door. You can feel (exagerated) the air moving through the door and across the room but there is no measurable "negative pressure" in the room. The air is moving into the room and replacing molecules as fast as they're being taken.

Put a vacuum gauge in your intake pipe and see if you can note any amount of vacuum. You won't because your intake pipe can flow way more air than part throttle operation can take out.

If you hit WOT from a dead stop, boost onset is pretty sudden once a threshold is reached. It's not a steady increase

No, it's progressive, almost exponential. Look at a compressor map http://www.stealth316.com/2-3s-compflowmaps.htm#int and note the wheel speed compared to airflow. It rises pretty rapidly. Compare the Turbo pounds of air flow to the same N/A engine's pounds of airflow to find when it actually becomes effective. What does this relate to in engine RPMs? Don't know, that would change from setup to setup.

I think it makes sense to go to multiple nozzle set-up once you get into larger water requirements

You're arguing against established experiences, which is fine, that's how we improve things and find new ways to go faster, but single nozzle systems have proven most effective so far.

Rick - '91 GSX

 

[Jehu]

Think of the vacumn as pressure drop in reverse. Any flow in any constrained area will cause a pressure drop, or in this case vacumn. This is how it works...

Let's say at 300cfm of airflow, the pressure drop through the following flow components are as follows. Airfilter - 0.5psi, intake tube 0.2psi, turbo (0 if it spins fast enough not to impede flow) IC piping 0.5psi, IC - 0.7psi, UIC pipe - 0.5psi, partially opened TB 3psi.

This means that the air MOVEMENT is being motivated by 5.4 psi of vacumn. Imagine that air is water, and the pressure is the level of the water. Water will only flow from a higher elevation to a lower elevation. Sure, the TB is a big restriction. But if the TB elbow is at atmospheric pressure, then nothing will draw air in from the airfilter. Air does not flow from 1 atm - 1 atm.

In your example, the pressure change in the room is not measureable because of the immense change in your volume. The flow in the engine is analogous to you placing a restriction in the vacumn hose. As long as air still moves, then the restriction is still under vacumn.

Ofcourse once the turbocharger start producing boost, things change...

As I said, it's basic fluid dynamics. If you don't believe me, why don't you do a little research?

You think boost is progressive? How much boost do you get from 0-2500rpm? How much from 2500 to 5000rpm? Also notice what wheel speeds the maps start at. 90,000. It takes quite a bit of exhaust gases to get that going...

Well, as I said, seeing how the largest nozzle aquamist has is a 1.0mm nozzle, flowing 310cc/min at the race pump pressures, you want to suggest to me how to use a single nozzle system to flow 420cc/min using a readily available pump and using aquamist nozzles?

As I said, I'd still rather have two nozzles due to the better atomization they deliver. That might be counterproductive to knock reduction as the smaller droplets will evaporate prior to entering the combustion chamber, but with a single injection point, I'm more worried about having smaller droplets that will be entrained in the airstream and evenly distributed to the cylinders.

 

[kicker91laser]

I wonder what I could run wiyh my evo 16g, 660s, and water injetion, on a stock smic? Anybody run a similar setup? I have really been considering putting w/i on my 89 colt gt that does not have an intercooler

 

[nellystyle]

ok ok ok, i'm on like information overload here....so with my current setup (see profile, basically 95 w/ 14b) I should go 550s instead of w/i with the stock i/c; correct? then get the w/i. (these are steps simply because of funding, not because i want to do them seperately, i'd throw them both in together if i had the money...but then again, i'd also have a front mount if i had the money ) so anywho...i think i'm swaying back towards 550s... so i at least have the fuel going in there, THEN, build a kit for w/i to help with the intake temps and i/c.

 

[kicker91laser]

ok ok ok, i'm on like information overload here....so with my current setup (see profile, basically 95 w/ 14b) I should go 550s instead of w/i with the stock i/c; correct? then get the w/i. (these are steps simply because of funding, not because i want to do them seperately, i'd throw them both in together if i had the money...but then again, i'd also have a front mount if i had the money ) so anywho...i think i'm swaying back towards 550s... so i at least have the fuel going in there, THEN, build a kit for w/i to help with the intake temps and i/c.

Thats what I would do, that way you know that you will not run lean, then put the w/i when you have the money, and turn up the boost a bit. This is similar to what I plan to do, also.

 

[4pistons]

In your example, the pressure change in the room is not measureable because of the immense change in your volume.

EXACTLY! Thank you for seeing my point. The small opening of the TB is the hose and the intake pipe is the room. Under part throttle the volume of the intake piping compared to the small TB opening is huge. The engine can't draw enough volume through the small opening to produce a vacuum in the intake pipe.

The flow in the engine is analogous to you placing a restriction in the vacumn hose.

You are assuming the intake piping is a restriction at part throttle. The only restriction (at PT) is the throttle body. The 2G TB has an area of approximately 3.3 square inches. If we even allowed up to half as 'part' throttle, which would be well into boost, that's roughly equivalent to the same flow as a 1.02" ID pipe. The intake piping would have to be smaller in diameter to even BEGIN causing a restriction. It is very simple to discover without the math and flow theories - put a vacuum gauge anywhere in your intake piping and you won't find any measurable vacuum.

You think boost is progressive?

I don't THINK it's progressive, every compressor map SHOWS it is. More precisely, the mass of air it moves is progressive. Actual boost and volume depends on other factors like inlet and outlet temperature, outet pipe volume, etc.

How much boost do you get from 0-2500rpm?

It doesn't matter for the sake of our discussion. All it has to be is positive.

Also notice what wheel speeds the maps start at. 90,000.

There are two maps on that page. I was referring to the first map, which starts at 46,050. The 90,000 map is not complete and has the low end cut off. That's not where volume starts.

you want to suggest to me how to use a single nozzle system to flow 420cc/min using a readily available pump and using aquamist nozzles?

I commented about using two nozzles spraying in different areas or in a progressive fashion. The complexity and expense of two Hobbs switches, two relays, the wiring and additional points of failure have not proven to work better than a single system. If you need the extra volume and just add another nozzle then that's not what I was talking about. You changed the basis of the argument.

Rick - '91 GSX

 

[Jehu]

EXACTLY! Thank you for seeing my point. The small opening of the TB is the hose and the intake pipe is the room. Under part throttle the volume of the intake piping compared to the small TB opening is huge. The engine can't draw enough volume through the small opening to produce a vacuum in the intake pipe. You are assuming the intake piping is a restriction at part throttle. The only restriction (at PT) is the throttle body. The 2G TB has an area of approximately 3.3 square inches. If we even allowed up to half as 'part' throttle, which would be well into boost, that's roughly equivalent to the same flow as a 1.02" ID pipe. The intake piping would have to be smaller in diameter to even BEGIN causing a restriction. It is very simple to discover without the math and flow theories - put a vacuum gauge anywhere in your intake piping and you won't find any measurable vacuum.

Look, as I said before, PLEASE go do some research, and you'll soon find that any piping, any boundary that contains a moving fluid is a flow restriction and hence generate a pressure drop. Besides, a shop vac in a room is not even in the same order of magnitude of volume ratios compared to the part throttle scenario. The pressure drop along a length of pipe has nothing to do with how much of a restriction exists before it. As long as there is flow in a pipe, there is pressure loss. Now, I'm not arguing that the partially open throttle plate is the major component of flow loss in that system at part throttle, but the fact remains that the negative pressure prior to the TB will not be insignificant, prior to the turbo spooling up.

Seriously, go look up piping flow losses on the internet. It's really pointless arguing about this when you are just pulling your theories out of your hat...

As far as introducing an additional point of failure by having two solenoids of injecting water, I think it's more of an additional safety. The chances of having both injection systems fail is a lot smaller than having 1 system fail, and in the case of a dual system, at least you still have 1/2 the water going. Better than not having any water at all, don't you think?

 

[suicidal2af]

Without getting into any fancy fluid dynamics stuff, since I'm no expert, I think some people may have the wrong definition of vacuum in their head.

Air flows from high pressure to low pressure areas. This is standard knowledge. Vacuum is simply a term for low pressure. Actually, the exact definition as used in relation to pressure is "Containing air or other gas at reduced pressure". How does this relate to the argument at hand?

An engine moving at 7000rpm is going to be flowing much more air than an engine at idle. Can we all agree on this? Speaking in a purely NA sense, in order to flow more air you need a higher differential between the high and low pressure areas. The air outside the vehicle isn't changing (much) in pressure, in fact it's staying pretty much dead-on at 1atm. Simple logic is going to dictate that at WOT and higher RPM's, we are creating a larger vacuum.

The bottom line is, even if the air is being replaced, you are still in a condition of vacuum -- otherwise there would be no flow.

 

[98GSX_MTX]

I just installed the www.snowperformance.net stage two kit. The reason I went with stage 2 is because I do alot of partial thottle boosting. I rarely have to floor my car to make it move so the adjustable boost onset and such is very beneficial to me.

For what its worth. I was able to pull 5% more fuel out and add 2 degrees of timing at WOT and I have zero knock. So it definately works like a champ. I am running winshield washer fluid injected just before the TB.

I could probably add more timing but I am scared that it will do "bad things" to my high mileage 7 bolt.

 

[4pistons]

All right, this is my last post on this subject. This discussion should be on a new thread. I apologize to the moderators and those on this thread looking for Water Injection info. I tried to end this before but was dragged back in.

I haven't pulled any theories out of my hat and I don't just make things up. I used real world experiences and practical applications to demonstrate my points. I do real world research.

Here's some real world research to demonstrate what kind of "vacuum" really exists in the world of air movement. I have a 0-30 inch, 270 degree sweep, handheld vacuum gauge with a long piece of tubing on it. It's a large Marsh Instrument gauge, not a POS. In my two story house the furnace moves an incredible amount of air. Much more air than a 2.0 liter engine moves at part throttle. (Or full throttle for that matter.) I put the tubing down the cold air return where I could feel large quantities of air moving as it passed into the ducts. Not a wiggle on the gauge, nothing. I then probed the center of the squirrel cage in the furnace blower itself. Nothing. NO MEASURABLE VACUUM.

Next, At my shop I have a 16 gallon, 6hp, Shop Vac with a 6' pickup hose. I put the Vacuum probe in the end of the hose. Nothing. I then ran the tubing all the way down the hose INTO the tank of the Shop Vac. Nothing. NO MEASURABLE VACUUM, in a vacuum cleaner no less! I then covered the end of the hose with my hand. Voila! 3 inches of vacuum! It really does exist.

Next, My 1997 Dodge Dakota 5.2 liter V8 truck. I attached the Vacuum Gauge to an intake manifold port as a baseline - 17.5 inches, just like it should be. I moved it to a port on the intake tube that connects the throttle body to the air filter box. Engine idling. Nothing. Rev to 3000 - Nothing. Rev to 4500 - Nothing. Crack it open quickly to WOT - Nothing. NO MEASURABLE VACUUM. How could that be? All kinds of air moving but no vacuum.

To have a measurable vacuum in a gas it has to exist in a steady state and space where the molecules are pulled away from each other, i.e., the space is made less dense, and that state must be sustained. In all the situations I've illustrated in this thread the air is rushing in as fast as the molecules are being pulled away and cannot maintain vacuum but instead moves rapidly. That's called "flow." Higher pressure air rushes in to fill the void where the vacuum is. That's how a cylinder gets filled as the piston moves down. In the case of the furnace blower and Shop Vac the vacuum exists at the tips of the impeller blades but is not measurable at that point by any of our methods or tools.

Any frictional losses due to the pipes (in the examples above) are negligible. Even the accordion shaped intake pipe on my truck, which induces all kinds of turbulence and friction, caused NO MEASURABLE VACUUM because the pipe can still outflow part throttle engine requirements. You can argue pipe head flow losses, coefficients of friction, and Reynolds Numbers all you want, but we're talking about automotive applications here, not plumbing Power Plants and Skyscrapers. It is very easy to step into the real world and actually measure it, like I did, and prove it to yourselves.

Rick - '91 GSX

 

[4pistons]

Back to Water Injection...

I was able to pull 5% more fuel out and add 2 degrees of timing at WOT and I have zero knock. So it definately works like a champ.

That is exactly what water injection is supposed to do!

Rick - '91 GSX

 

[Jehu]

Your pressure gauge isn't sensitive enough. Whenever I do Pitot measurements in the inlet side of fan( and I used to do A LOT), the static pressure is always negative. I typically use an inclined tube manometer that measures in factions of inches of water. Ask anybody who measures airflow for a living. Even if it's an open pipe, the inlet of the fan will have a negative static pressure, if there are no fans upstream of the measurement point.

I do agree that this probably isn't exactly relevant to the thread, but there are some good discussions in here that will help people understand the dynamics of fluid flow.

Higher pressure air rushes in to fill the void where the vacuum is. That's how a cylinder gets filled as the piston moves down. In the case of the furnace blower and Shop Vac the vacuum exists at the tips of the impeller blades but is not measurable at that point by any of our methods or tools.
Rick - '91 GSX

So maybe I can illustrate it using your example. Yes, ambient air rushes in to fill the void. The void pulls on the air molecules, an increases the void in between. The molecules that gets pulled on in turn pull on the molecules down the line. To get the ambient air to move with a vacumn, you have to pull on it. The moving air thus has molecules with wider spacing in between, and that my friend, is vacumn.

I don't know how to tell you this, but you're arguing against established science here. I'm done with this too, I think I've done as much convincing as I'd like to do.

 

[97TSIAWD]

To the subject of this orginal post. I use the smc alky kit and I run higher boost pressure, without fear of detonation. I was seeing good results with 93 octane and the alky but ultimately found myself using leaded gas also, the leaded gas gives it the extra power to hit higher mph then with just 93/alky. I kept the alky on of course and see great numbers in timing and I am running very rich for my boost pressure. I could probably take out much more fuel to increase the amount of power, but that is for another time. I have used the kit for over two race seasons and have seen my car go faster every time I go to the track without modifications to the current setup, merely changing boost and fuel mixtures. It is a great addition to the car, and I would recomend it to anyone who is looking for a save way to increase boost and power without detonation.

 

[98GSX_MTX]

IRT the above....

Are you running windshield washer fluid or something else? When you say Alcohol where are you getting it from?

 

[97TSIAWD]

I use denatured alcohol from Home Depot, at 9 dollars a gallon. The chemical properties I don't know off hand but it is basically methanol with a 2-6% dillution with rubbing alcohol. I also use the table spoon of the lubricant, can't remember the name that is recomended by steve.

 

[98GSX_MTX]

Holy crap is that expensive.....I cannot see your mixture giving you that much of an increase over blue water! I live by the 80/20 rule. Get 80% of what you want for 20% of the cost. The other 20% will cost you the additional 80%.