On Using Ethanol Fuels F uels in Unmodified Vehicles Contents 1. Background 2. Engine life and engine temperatures 3. Ethanol Fuels in Unmodified Unmodified Vehicles Vehicles 4. E20 and and E85 5. E20 and and NREL 6. Comparisons with high air-cooled head temperatures temperatures - Correlating CHT and EGT 7. Additional insights insights from NREL 8. E85 temperatures, fuel economy, and the SAE-MIT-Honda study 9. Satisfactory engine engine behavior behavior with E85 10. Going with 100% E85 in problematic vehicles vehicles 11. Calibration of E85 kits 12. Field testing testing the theories theories 13. Conclusion
©Copyright: John Kolak. Permission is granted to freely copy and distribute this document in its complete and unaltered form only. only. Online postings must link back to the original posting site. Contact:
[email protected] [email protected]
1. Background I subscribe to Abe Shackleton's blog at OpenFuelStandard.org. OpenFuelStandard.org. I have been searching for information on using ethanol fuel in unmodified vehicles. We have had a couple of conversations, and when the topic got technical, he referred referred me to the Alcohol Can Be a Gas forums. While the forum is rather congested, I did notice the Unmodified forum and started looking there for answers. I noticed that Bob Glicksman was the primary driver of inquiry there. It seems that good information is hard to come by, by, but he was asking all the right questions. After some time away from the forum, he returned and posted the results of his studies studies in a document called Ethanol and Internal Combustion Engines, which can be downloaded at this link: http://www.liquidsunenergy.com/learning/ppt/ice.pdf
As of this writing Bob's document is of necessity conservative and reluctant to make recommendations. recommendations. Because of its informational and somewhat encyclopedic nature, it has to limit itself to known facts verified by properly designed scientific studies. At this time, hard information on the effects of ethanol fuel on today's motor vehicles is hard to come by. The US government is working on studies to certify ethanol for use in our nation's vehicles. While the E20 certification is finished, E85 studies are still in progress. I contacted Bob to discuss a few technical issues in his article and shared with him the results of come cursory tests on my own vehicle with my preliminary recommendations. Since he is trying not to include recommendations recommendations in his document, he asked me to publish my results for the public benefit. Section 2.2 has some interesting information on the history of American motor vehicle and motor fuel development development in its discussion of Henry Ford's original philosophy of fuel availability. availability. At the time, gasoline could only be purchased in the city, and alcohol was readily available at farms, so the Model T was designed as America's first flex-fuel vehicle vehicle (FFV). There is a nice video about this made at the Ford Museum by David Blume here: http://www.youtube.com/watch?v=5qDYoEupI28
When prohibition came along, Ford made the decision to drop support for ethanol fuel because he reasoned that he couldn't expect farmers to buy his product if he did not buy their product. This appears to have been a bad decision because the biggest problems we have in engine longevity are related to the higher energy content of gasoline. 2. Engine life and engine temperatures There are some good temperature charts at GAMI aviation that show that gasoline combustion has a high temperature temperature peak in at Air-Fuel A ir-Fuel Ratios (AFR) near the stoichometric mixture (stoich). This is the ideal blend of fuel and air at which they are perfectly balanced so that there is enough fuel for all of the oxygen and enough oxygen to burn all of the fuel with none left over. over. Bob discusess stoichiometry in more detail in Section 3.2 of his document. Figure 2 at GAMI is found here:
http://gami.com/articles/bttfpart1.php (or directly, here: http://gami.com/img/articles/bttf/pt1fig2.jpg ) What we are most interested in on this chart are t he red and blue lines running on a curve at the very top. These represent the Exhaust Gas Temperatures (EGT) of a TCM IO-550 engine running at 70 and 79% loads. The heavy red vertical line is the temperature peak at stoich. The red and green vertical lines to its left and right indicate the boundaries of the peak's hot zone at plus or minus 50 degrees F. Running Running an engine in this hot zone causes serious engine damage when the graph shifts upwards into even higher temperatures temperatures beyond the 79% power graph illustrated here. The GAMI charts measure AFR in terms of fuel flow, flow, whereas automotive AFRs are generally expressed expressed as a ratio or lambda value (Glicksman Section 4.1). Mixtures from 13.1:1 to 15.5:1 are generally considered considered unsafe for engines under load. This corresponds to lambda values of 0.89 to 1.05. The two strategies for the protection and preservation of an engine are to run at the lower temperatures temperatures seen on the chart at either the rich side of the peak on the left side of the chart, (rich of peak cooling), or at the lower temperatures temperatures seen on the lean side of the peak on the right side of the chart (lean of peak cooling). Theoretically Theoretically either is possible, but in actual practice, lean of peak is not done in automotive engineering. If you go back to GAMI Figure 2 and look at the third set of red and blue curves, you will see that the power curve drops off more steeply as you go further lean of peak. What this means in practical terms is that t hat if you don't have tuned induction and closely calibrated fuel injectors, the various cylinders on the engine will not receive exactly the same AFR. This causes the engine to run out of balance, but more importantly creates a situation where some cylinders might run rich enough to burn up. GAMI provides a service to the aviation industry to balance induction and provide tuned injectors, but this is cost prohibitive in the mass-produced automotive market, so automobile engines are always tuned for engine protection at rich of peak. In Sections 3 and 4, Bob discusses the interest of the government to reduce exhaust pollution by ensuring that engines run as close to stoich as possible, and that exceptions were allowed for heavy heavy engine loads and cold start situations. It is precisely because of this need for engine protection that the automotive ECU is programmed to deviate deviate from stoich to allow a richer AFR under load. So, as we can see, Ford might have made a better decision if he had chosen alcohol fuel instead of gasoline. Now that Brazil has been running alcohol fuel for so long, reports are coming back that engines last two to three times longer burning ethanol rather than gasoline. Besides gasoline's higher temperatures, gasoline's combustion explosion is more violent than ethanol's so that bearing and cylinder wear and tear is greater when running gasoline fuel. 3. Ethanol Fuels in Unmodified Unmodified Vehicles Vehicles The purpose of this paper is to provide recommendations recommendations concerning the use of ethanol fuels in unmodified vehicles. The temperature temperature information, and its danger to engines, discussed here is the primary factor that needs to be considered. In the past there have been concerns about damage to rubber and plastic fuel system components being damaged by ethanol. However, However, the US government has mandated that all vehicles sold in the US be able to tolerate ethanol since the introduction of E10 gasohol in 1983. For those living outside the US, you can probably add a couple of years as manufacturers use up their supplies of old stock and adjust
contracts with component suppliers. Abe spoke to one technician who said that ethanol should be safe for all cars since 1990. I would think that is a safe and conservative conservative estimate in comparison to the US government's 1983 requirement. requirement. Ohio Biosystems Biosystems also did a by-the-numbers by-the-numbers check of parts on newer vehicles by comparing manufacturer part numbers of their Flex Fuel Vehicles and equivalent non-FFV models. They found that the part numbers are often the same, and where they vary, it's usually not more than 3 or 4% of the parts. They explain that this is largely a function of the manufacturers' need to preserve economy of scale by not having to manufacture, stock, and distribute parts for parallel model lines. The Ohio Biosys B iosystems tems study can be found here (Frame 2 also has an interesting video on the tear-down of an unmodified Chevy Tahoe which was run over 100,000 miles on E85 with no evidence of ethanol-related damage): http://www.ohiobiosy http://www .ohiobiosystems.org/OBSC-NTEP_file stems.org/OBSC-NTEP_files/frame.htm s/frame.htm So, for those who own unmodified non-FFVs, the primary concern is potential engine damage by high temperatures. temperatures. If ethanol runs cooler than gasoline, why are we concerned about the potential to run hotter? The reason is that ethanol's stoichometric ratio is different from gasoline's. Gasoline's is 14.7 to 1 (14.7:1). Ethanol's is 9.0:1. What this means is that one part of gasoline needs 14.7 parts of air, air, but one part of ethanonl only needs 9.0 parts of air. So if one part of ethanol is run through a car that is programmed to run gasoline, the computer (ECU) will mix it with 14.7 parts of air instead of 9.0 parts of air. air. Or in other words, the computer is giving the ethanol fuel too much air, or relatively relatively speaking, the computer is giving the air too little ethanol. So what that means is that ethanol run through a gasoline Electronic Fuel Injection system (EFI) will run relatively relatively leaner. leaner. This is not enough of an offset to take the AFR lean of peak. All it does is shift it into the hot peak temperatures. The US government studies call this effe effect, ct, "enleanment", which represents represents a *relative* *relative* leaning while the actual AFR is still rich of peak. 4. E20 and and E85 So what needs to be looked at is whether or not burning ethanol fuel in an unmodified vehicle will cause engine-damaging temperatures. temperatures. As it turns out, the answer is different for E85 and E20. E20 can cause a temperature temperature spike, while E85 appears to run sufficiently cooler and leaner to negate this (in unmodified vehicles). So we need to look separately separately at the consequences of using E20 and E85 in unmodified vehicles. The government is still working on certification studies for E85, but E20 has been certified safe for all US vehicles since 2001. Brazil has extended their certification to E25 for all vehicles sold in the country, but there are understandable understandable concerns about accepting a certification by a foreign government without knowing more about their research methodology. 5. E20 and and NREL
In 2008, the National Renewal Renewal Energy Laboratory (NREL) produced a report for the US Department of Energy called, "Effects "Effects of Intermediate I ntermediate Ethanol Blends on Legacy Vehicles and Small Non-Road Engines, Report 1", which can be found here: http://oxytane.com/mys http://o xytane.com/mystery/Liquid%20fuels/ tery/Liquid%20fuels/ethanol%20and%20small%20engi ethanol%20and%20small%20engines-Nrel.pdf nes-Nrel.pdf This report covers gasoline and ethanol blends up to 20% (E20). While E85 is beyond the scope of this study, it is listed in Section 4, Next Steps S teps (page 4.1 or PDF page 71). While the study is largely concerned with impact on exhaust emissions, there is considerable data on temperatures and engine damage. Page 3-8 (page 44 in the PDF) has a chart in Figure 3.5 of the effects of E10, E15, and E20 on engine temperature at Wide Open Throttle (WOT - See Glicksman Section 4 for EFI engine modes and WOT). Remember that the ECU keeps the AFR at stoich during closed-loop mode, so we are only interested in what ethanol fuel does to engine temperature in open-loop mode under heavy engine loads. Since cold-start is a noload open-loop mode, it is not of interest to us in considering ethanol's effect on engine safety. The first thing that we learn from Figure 3.5 is that the way your unmodified vehicle reacts to E20 depends on what kind of fuel injection system you have. The study divides this into two types: vehicles with ECUs that add Long Term Term Fuel Trim (LTFT - Glicksman Section 4.1) to the engine-protection enrichment mixture, and those that do not. The blue bars indicate that when the ECU adds LTFT to the enrichment mixture, the engines run cooler with gasohol than they do with gasoline. This is to be expected due to ethanol's lower combustion temperature. temperature. On the other hand, the red bars on the chart show that when the ECU does not add LTFT LTFT to the enrichment cycle, the engine runs hotter with gasohol than with gasoline. This is to be expected because of the *enleanment* *enleanment* effect of low ethanol blends. Remember that the enleanment effect shifts the AFR from richcooling into the stoich region combustion temperature peak. 6. Comparisons with high air-cooled head temperatures temperatures - Correlating CHT and EGT While the chart shows an EGT increase of up to 32 degrees C (58 degrees F), the conclusion of the NREL study is that none of the engines tested showed any sign of heat stress upon inspection. By comparison, a study of aircooled Volkswagen Volkswagen engines is useful since they run hotter than water-cooled water-cooled engines and are thus more susceptible to heat damage. Jake Raby has written an article on this at 914 World: http://www.914world.com/specs/JakeRabyHeadTemps.php He indicates that a healthy VW engine should run with cylinder head temperatures temperatures (CHT) in the range of 300 to 340F for a well-tuned performance engine, or 350 to 375F for a stock engine at cruise speed, which can approach 400F under heavy load on hill climbs. Notice that he says that if you are seeing 375F 37 5F constantly, constantly, even at cruise, you are already starting to see mild heat stress damage or wear and tear "due to the extreme heat cycling the engine and valve train parts are seeing." Reading further, further, he says at 400-420F "the rules change, 400 is the magic number that stretches valve valve train parts and really where cracking issues begin," and at 420F+, "Well your engine has been damaged."
So this is a range where a normal engine running acceptably at 350F is only 50 degrees F away from engine damage. But it's important to distinguish between 50 degrees on a CHT gauge and 50 degrees on an EGT gauge. If you go back to GAMI Figure 2, you will see that the 50 degree increase in EGT from rich cooling to peak is reflected by only a 5 degree increase in CHT. This is because of the buffering effect of the cylinder head and its ability to wick heat away from the combustion chamber. chamber. So a 50 degree increase in the head temperature of an air-cooled VW engine is certainly a reflection of a much larger increase in EGT. EGT. So we can probably conclude that it is safe to follow the NREL conclusion that E20 will cause no discernible heat related damage regardless of the type of FI ECU your car employs. 7. Additional insights insights from NREL Before leaving leaving the NREL study, I would like to direct attention to some of the other charts. First, Figure 3.6 show the effect of ethanol blends on temperatures temperatures of "non-road engines". This is what they call "open-loop engines", or what we would call carbureted engines. With no computer to enrich them, they predictably run hotter due to the enleanment effect effect of E20 under all engine operation conditions, but again, no engine damage is observed in the study. Before running ethanol blends in an older carbureted engine, it should be considered whether the carburetor contains rubbers that can be affected by ethanol. But again, if the rubbers were made after 1983, and the carburetor has been running standard pump gas (E10), it is probably okay. okay. Older cars run in foreign countries that have never been exposed to gasohol and never had the carburetor rebuilt with a newer rubber set may be vulnerable. This is somewhat unlikely as a 20 year-old car that has never had the carburetor rebuilt likely has detiorating rubber in it anyway that would show problems even when running gasoline. Note that the chart reflects an average EGT increase for E20 in carbureted engines of 35C, or 63F, 63F, which, as expected, is nearly the same as the 58F increase seen for the open-loop WOT in non-LTFT adjusted engines. Figure 3.15 is interesting because it tests another type of engine, portable generators. It shows a bit more of a temperature temperature increase for E20, about 55C, or 99F. It's good to have more tests and numbers to look at for possible worst-case scenarios since the reaction of a particular car to E20 is very much a matter that varies according to individual make and model. Again, no heat-stress engine damage is detected on these generators generators either. Figure 3.16 shows the variation in individual cylinders on a 2-cylinder engine as discussed by GAMI. Figure 3.17 is valuable because it shows the temperature change under varying load conditions. In the GAMI chart, we saw the effe effect ct of AFR on combustion temperature, temperature, but here we get to see the effect of load. Predictably, Predictably, higher loads mean higher temperatures because higher throttle positions increase the amount of fuel provided to the engine, and consequently the amount of heat energy in the larger fuel charge is greater. greater. The higher temperature seen at no load in cylinder number 2 should be seen as an inadequacy of the engine's cooling system, system, and not as a factor of higher heat released at idle. Figures 3.18 and 19 show the highest temperature temperature trend slopes with average average increases of about 65C, or 117F, 117F, for Briggs and Stratton power washers; and 75C, or 135F, 135F, for Stihl line trimmers. So this concludes the discussion of E20 blends. In conclusion, US government tests indicate E20 to be safe for all
types of engines, with open-loop high load modes of operation increasing exhaust gas temperatures on average from about 58 to 135 degrees Fahrenheit Fahrenheit depending on the engine, and with some types of ECUs actually lowering the temperature slightly. 8. E85 temperatures, fuel economy, and the SAE-MIT-Honda study This section moves into discussing the suitability of E85 for unmodified vehicles and its differing issues. The first issue is that since E85 has so much ethanol, ethanol's lower combustion temperature becomes more of a factor than the enleanment effect observed in E20, so we don't generally see unsafe EGTs for E85. The next study to look at is published by the Society Soc iety of Aumotive Engineers (SAE), and written by MIT in collaboration with Honda: Effects Effects of Ethanol Content on Gasohol PFI Engine Wide-Open-Throttle Wide-Open-Throttle Operation http://www.tricktuners http://www .tricktuners.com/forums/a .com/forums/attachment.php? ttachment.php?attachmenti attachmentid=86&d=1288800657 d=86&d=1288800657 This study is a more complete examination and covers all ethanol blends from E0 to E100. In the popular discourse, there is a lot of conflicting information about exactly how much cooler ethanol burns than gasoline. On page 344 of Alcohol Can Be A Gas (See Glicksman, Section 6), David Blume lists ethanol as burning 400F cooler than gasoline, or 350F cooler in his YouTube YouTube video lecture of the same title. David does say that, "Under ideal research engine conditions, where everything is tuned perfectly, perfectly, the difference difference between alcohol and gasoline exhaust exhaust can be as little as 130F, 130F, but those conditions are not found in most cars on the road. The difference difference is far more pronounced on older vehicles without sophisticated sophisticated electronic control units." This agrees with the test literature here. Figure 9 of the SAE/MIT/Honda SAE/MIT/Honda study charts EGT against ethanol blends from E0 to E100. With a slight variation in individual empirical data points, they plot a straight slope of -2.0 degrees C for each 10% increment of increasing ethanol content for a total delta of 20C, or 36F. They say this slope tracks the expected difference predicted by the Adiabatic Flame Temperatures emperatures of gasoline (2289K/3660F) and ethanol (2234K/3561F) (2234K/3 561F) (See also Figure 4). As you can see, the Adiabatic Flame Temperatures of the two are quite close, which explains small difference in combustion temperatures. temperatures. One might expect the difference to be tracked by the Heats of Combustion instead. The Handbook of Chemistry and Physics lists them in kg-calories as follows. Ethanol Ethanol is 327.6, which is a small fraction of that found in some of common components of gasoline such as Benzene at 782.3, n-Octane at 1302.7, and n-Hexane at 989.8. So with such a vast difference difference in Heats of Combustion, this is possibly one of the sources of the erroneous concept that ethanol fuel economy must necessarily be much worse than gasoline, with a 30% reduction being the common figure cited. However, a study posted at OpenFuelStandard.org shows that in some cases fuel economy can even be increased with ethanol fuel in unmodified vehicles. The Toyota/Ford study notes that, "Three of the four cars got better gas mileage on twenty or thirty t hirty percent ethanol than on pure gasoline." The post with a chart and link to the study can be found here:
http://www.openfuels http://www .openfuelstandard.org/2011/06/ethanol-can-i tandard.org/2011/06/ethanol-can-increase-your-mp ncrease-your-mpg.html g.html There are various reasons for this, and the studies indicate that all of this is not understood at the present time, but the SAE/MIT/Honda study has a wealth of data and charts that provide some indications. One indicator is that Figures 13 and 15 chart the various energy losses in an engine, and ethanol is shown to have about 6% less energy loss than gasoline. Another indicator is that Figure 12 indicates that Net Indicated Mean Effective Pressure Pressure (NIMEP) is increased when timing changes from knock-limited timing to Maximum Brake Torque (MBT) timing. MBT timing is the ideal timing setting that an engine would use for maximum output and efficiency. efficiency. However, However, since gasoline more readily causes engine knock, timing must be retarded to prevent it. Ethanol is more knock resistant and thus has a higher octane rating than gasoline. Thus the timing can be released to MBT timing instead of running on gasoline's knock-limited timing. This give the engine more power, power, or the same power at less fuel consumption for better fuel economy. A third indicator is that Figure 8 shows a fairly steady NIMEP across all blends of fuel from E0 to E100. This means the combustion pressure remains nearly constant. Remember that the engine is not driven by the heat energy released, but rather the rapid expansion of combustion chamber gases create the explosive explosive pressure wave that drives the engine. While heat is needed to drive the expansion of the gases under the gas laws, it is an excessively excessively produced waste product that must be disposed of, and is one of the primary enemies of engine longevity. longevity. So this means that ethanol's lower heat content is not the driving factor in predicting engine performance, but rather its expansive pressure potential. Ethanol has more "push" per unit of volume even though it has less energy content. The chemical reaction of oxygen with ethanol produces more combustion products (mass - Glicksman Sections 2.3, 3.2 and 6), than the reaction with gasoline. Here is the chemical formula for the combustion of ethanol: C2H5OH (liquid) + 3 O 2 (gas) → 2 CO2 (gas) + 3 H2O (gas)
http://en.wikipedia.org/wiki/Ethanol The composition of gasoline includes a variety of compounds, but Wikipedia uses octane as a typical example. It is also convenient for easy comparison because it contains exactly four times as many carbon atoms as ethanol: C8H18 + 12.5 O2 → 8 CO2 + 9 H2O
http://en.wikipedia.org/wiki/Gasoline So if we multiply the ethanol equation by 4, we see here that octane combustion uses about 4 times as much oxygen as expected (12.5 molecules vs 12) and produces exactly 4 times as much carbon dioxide (both yielding 8). But where the difference really comes in is in the water production with ethanol producing 12 molecules against octane's 9, or in other words, ethanol produces 33% more water mass than octane and its similar components in gasoline. Note that combustion produces water vapor rather than liquid water. Water is normally a liquid at room temperature. At normal atmospheric pressure, carbon dioxide has no liquid state, but
changes directly from a solid phase to a gaseous phase above −78.51 °C (−109.32 °F) (See Wikipedia W ikipedia – Carbon Dioxide). I would expect this to give water more potential pressure energy than the carbon dioxide under the gas laws to contribute to the driving of the engine. Adding a water mist into the induction stream is known to provide additional power and economy for internal combustion engines because of the superior expansion effects effe cts compared to plain air induction. I would also expect the increased water presence in ethanol to provide additional cooling of the combustion process, but this is not necessarily supported by the emprical data. Aviation safety requires that aircraft be equipped with CHT/EGT CHT/EGT gauges, so a lot of our information in terms of data and experience comes to us from aviation sources. An aviation tuner at E85vehicles.com E85vehicles.com reports that "Airplanes burning AGE85 (Aviation Gas E85 – 88% ethanol) usually have 200 degree lower EGTs compared to 100LL avgas. To get better cruise economy on ethanol fuel the mixture is leaned till EGTs get back up to 100LL temps." Link: http://e85v http://e85vehicles.com/e85/i ehicles.com/e85/index.php?topi ndex.php?topic=4380.15 c=4380.15 This brings up the important point that a vehicle tuned for E85 will not run significantly cooler than expected gasoline temperatures. temperatures. The next user on the same thread ran experiments on a Honda with a customizable ECU. He programmed it with both E85 and gasoline fuel maps and found the EGTs to be nearly identical with just a slightly noticeable temperature temperature reduction for ethanol. He has posted the test result charts from his Innovate TC4 EGT datalogger datalogger at the same link above. They are well worth looking at. Again, this is consistent consistent with the temperature data in the SAE/MIT/Honda SAE/MIT/Honda chart. Now let's look at the SAE/MIT/Honda Figure 9 chart. Note that it is performed at WOT and a lambda value of 1.0, or in other words, at the very top of the EGT peak. This means that we are seeing the absolute maximum temperatures temperatures possible under the given load on this test engine. Remember that the GAMI chart teaches us t hat EGT varies with AFR, NREL teaches us that EGT varies with load, and this chart shows us that EGT varies with ethonal content. WOT WOT is the highest possible fuel charge for the given engine load, and lambda 1.0 is the highest possible mixture temperature, so this graph should be indicating the maximum possible temperatures for all blends of ethanol from E0 to E100 for the given load, and therefore, therefore, only increasing the load will yield higher temperatures. temperatures. The chart shows a maximum temperature for pure gasoline at 675C (1247F), and 640C (1184F)for pure ethanol, with E85 slightly higher at about 645C (1193F), or in other words, a difference of only 54F. So this chart substantiates substantiates David Blumes statement that research engines tuned for ethanol only show a slightly lower temperature temperature for ethanol fuel. Aviation data doesn't substantiate substantiate his report of 350F lower ethanol combustion temperatures, temperatures, but he does say that it very much depends on the vehicle, and that the cooler running is found more commonly in older non-ECU vehicles. The ethanol properties chart on page of Alcohol can be a Gas states that ethanol can be run satisfactorily satisfactorily on a wider range of lambda values. While ethanol's stoichiometric ratio is 9.0:1, he lists ethanol as having tolerable AFRs from 5.3:1 to 23.3:1, 2 3.3:1, so he may have gotten the 350F figure from lean-jetted carburetors carburetors getting a leancooling effect on carbureted engines. engines. So it would appear that our EFI-equipped cars are getting the 200F cooling effect reported reported from aviation rather than the 350F reduction reported in Alcohol Can Be a Gas. This 200F reduction is also suggested by my personal test reported in Section 12 below. below.
So far, far, we don't see a scientific explanation for why real-world engines are showing a 200F temperature decrease when burning ethanol instead of gasoline, and this is the subject of ongoing certification studies currently being performed for E85. In time, we should see another NREL study that will advance our understanding understanding of E85 fuel just j ust as their 2008 study did for our understanding of E20 fuel. For the time being, all we have are the reports of gauge-monitored experience experience in the aviation industry and anectotal experience from individual advanced automotive tuners. As FFVs increase their ethanol tuning efficiences, we may see the 200F EGT drop go away as engines are brought up to running the Figure 9 temperatures through more efficient tuning. 9. Satisfactory engine engine behavior behavior with E85 So having observed that E85 temperatures are safe for an unmodified vehicle, the decisive issue will be whether or not a particular unmodified vehicle can run on E85. There are two issues at stake that can make E85 unsatisfactory unsatisfactory for a given car. car. Both are a re the consequences of ethanol's leaner burning in gasoline calibrated systems. systems. The first is the ECU not being happy with the LTFT changes and setting the CEL light. The second issue is experiencing poor driveability caused by the lean AFR. We have all experienced this kind of coughing and sputtering when we are driving a car that is starting to run out of gas. In an interview with Wisconsin Eye, David Blume states that any vehicle can run at least a 50% mixture of E85, which is effe effectively ctively about E50 when mixed with gasoline, considering that pump gas is effectively E10. The interview is found here: http://www.youtube.com/watch?v=pzgk9Dq http://www .youtube.com/watch?v=pzgk9DqnIrA nIrA
He explains in more detail how he learned this in Chapter 13 of Alcohol Can Be a Gas. He describes a group of ethanol fuel enthusiasts in South Dakota that developed a method to test a particular vehicle for E85 tolerance. Basically the procedure is to go to the filling station with a near-empty gas tank and add a small amount of E85. If your car is happy, happy, you're fine. If not, the recommendation is to start experiementing experiementing with various mixtures to see what mixture makes your car happy. For example, example, first try running a 50/50 mix of E85 and gasoline. If that works, try 60/40 and so on until you learn exactly how much E85 you can tolerate. 10. Going with 100% E85 in problematic vehicles vehicles Of course, if you are one of the ones who can only run a 50/50 mix, while that might help the fuel budget, it may not be satisfactory at today's fuel prices. So for those who are bold and brave, you can start looking for what to do to get your car to run on 100% E85. In Section 5.3, Glicksman explores the various options for converting a non-FFV into an FFV. The options break down as follows: 1. Buy an OEM ECU from your dealer. dealer. This is perhaps the best option because your dealer will also know vehicle specifics such as whether or not a fuel sensor has to be added. The downsides are that ECUs are expensive, expensive, which will cut into your fuel savings, but more importantly, importantly, FFV ECUs are not available available for all makes and models. 2. Buy an aftermarket programmable programmable ECU. This is a good option, and may even even turn out better than an OEM ECU. David Blume says that the manufacturers manufacturers wanted to do as little work as possible to certify their vehicles as
flex-fuel capable, and that as such, the ECUs are not really optimized. Having a programmable ECU optimized for E85 may give superior results. The downsides are the high programming skills required, or if not doing it yourself, yourself, the high cost of paying to have it done. This solution is a winner for those who already know how to do it. 3. Install one of the E85 conversion conversion kits on the market. These generally generally run in the $300 price range. While there used to be a variety of technologies to choose from, now they all generally tend to be of the same type. Shopping around might find additional features features offered such as cold-start enrichment. David Blume also says that auto manufacturers did not want to pay $50 per vehicle to add cold start kits, so for this reason E85 was certified as an alcohol fuel. The 15% gasoline content content is supposed to give enough volatility to eliminate the need for a cold-start kit. I have never tried to start a car with E85 on a cold morning, so I have no direct experience with this concept. Note that any type of kit will still add to the cold start injector pulse and volume. Note also that this is an E85 kit, not a FFV conversion. If you fill up on gasoline, your open-loop modes may run too rich, and possibly damage your catalytic converter. converter. For these situations, David Blume suggests installing a switch to shut off the kit while running gasoline. Many people have reported reported a bad experience with the kits. Complaints run from poor performance to bad fuel economy. economy. In Section 5.1, Glicksman briefly touches on the important point that the way any particular car will react to E85 (or a conversion kit) is very much dependent upon the particular make and model of the car, car, and depends on a large number of variables in the design engineering. This, for example, example, is seen in the NREL test that found that E20 effects depend on how the ECU handles LTFT. Other factors that can be an impact with E85 include whether or not a system designed for gasoline can deliver sufficient volumes of ethanol to keep lambda values at 1.0 in closed loop mode, and whether the ECU can tolerate the LTFT LTFT values necessary to run ethanol at lambda 1.0 without setting the Check Engine Light (CEL Glicksman Section 5.1). 11. Calibration of E85 kits The biggest problem with the E85 kits is that they are being sold as a s "Plug and Play" kits. While it may be true that a child can install one in a few minutes with no mechanical experience, the poor performance problems are coming from not checking the results of the installation. installation. By the way the tests described in this section are not a bad idea for E85 users who do not need an E85 kit as well, just to know what is happening in your engine. If your ECU can calculate the correct E85 volumes, and get the EFI system system to apply them, to keep your lambda at 1.0, everything is happy. If it is not able to do so, or if you can't run E85 because the CEL light comes on, then an E85 kit can help you. All the kit does is add a little more time to the fuel injector electronic pulse so that the injectors deliver a little bit more fuel than what the gasoline program calls for. These kits include a potentiometer to fine tune them, but mostly these are ignored because whatever you set it at, in most cases the ECU is able adjust the AFR according to the feedback from the oxygen sensor, sensor, so E85 kits are shipped with the potentiometer set at an intermediate intermediate position and most people just leave them there. Also, if your ECU is not adding LTFT to open-loop modes, the E85 kit potentiometer potentiometer setting can help you there as well.
To properly calibrate your E85 kit, you need an AFR meter that displays lambda as well as AFR, such as the LM1. The first check should be to see if the system is able to maintain a lambda value at nearly 1.0 during all closed-loop operating conditions. Next should be a check of open-loop lambda values. While good cold-start values are nice, we are more interested in the values at heavy engine loads such as WOT. It takes a fairly steep hill or hard acceleration to kick the ECU into open-loop mode, so you need an assistant to ride with you to observe the meter to see when the system goes open loop. This will be apparent because the lambda values you are seeing on the meter will suddenly change from the vicinity of 1.0 to something around 0.85. Ideally these numbers should be higher than 0.81 and lower than 0.88. If so, the ethanol is being run in the same lambda values as open-loop gasoline, but the precise value may not be important if engine performance is satisfactory and the exhaust gas temperatures temperatures are safe for the engine. Otherwise, adjust the potentiometer until the readings are to your liking. This brings up the next point. The final step to certify that your engine is not suffering from enleanment enleanment overheating due to some bad combination of ECU values giving a fuel mixture causing an E85 temperature peak. This should not be happening because E20 seems to have caused the maximum EGT peak of 50F like the one that is seen on the t he GAMI chart, and from there, the SAE/MIT/Honda study shows that E20 to E85 should give back 13C, or 23F. But if you want to know what is happening with your car, you can get the EGT readings. Unfortunately, this reading is the most difficult to get because, for best accuracy, accuracy, the probe should be installed on the exhaust manifold as close to the cylinder head's exhaust port as possible. Tuners Tuners are not that interested in this because they can install it in a more convenient convenient location and just watch for temperature differences differences over the norm. Plus probes placed in the hottest location tend not to last as long. But if you want to know absolute combustion combustion temperatures temperatures as close as possible, you should install the sensor in the closest position. Most people will not take the testing to this level because it requires drilling and welding a sensor bung (mount) onto the manifold or pipe. If you are lucky, your car may have another sensor located in such a location which can be temporarily removed to use intall your test sensor. sensor. Of course, if said sensor is part of the engine management, its removal may change the validity of your test. 12. Field testing testing the theories theories Not being willing to make this modification, I did the next best thing by testing with a cylinder head temperature temperature (CHT) gauge. CHT readings are not as meaningful as EGT readings. Again, the GAMI chart indicates that a 50F increase in EGT is only registered as 5F on the CHT. CHT. But, again, in the spirit of the tuners, it is useful to check for differences differences and trends. I ran the test on a 1998 Hyundai Accent. Hyundai varied their FI system suppliers, suppliers, so my results will not necessarily apply to all owners of the same make and model. This car has Bosch FI. E85 kit sellers I have seen seem to think the car should have Nippondenso. Nippondenso.
Ethanol is coming to my area. My town now has E20, but the nearest E85 station is 200 miles away. I needed to go there the previous weekend, weekend, so I ran E20 the week prior to see how the car reacted to it. Then on arrival I filled up only a half tank of E85 since David Blume said that some vehicles can only run 50% E85. I actually had more E20 in the tank on arrival than expected, so I estimate I ran an E85 dilution equivalent equivalent to E70. I installed a Dakota Digital CHT gauge to monitor the engine to see how the readings would compare with Jake Raby's recommendations for VW head temperatures. Running on E20, I found head temps to be running about 225 to 250F, 250F, with lows in the 190 range at idle or deceleration. deceleration. Flooring it going up a mountain climb, I could not exceed 269. I'm not sure what the temps are supposed to be, but these numbers look very safe and normal to me. I checked some CHT gauges on ebay for water-cooled engines, engines, and they have a display range of 140-340 degrees, so I feel the 269 max is well in the safe zone, and seems to collaborate that Brazil has certified that all vehicles in the country are capable of running E25. I was, however, however, surprised on the return trip. I was not able to get the CHT to even reach 250. I pushed hard for a while up a hill and could only reach 249. The trip up is mostly uphill, and the trip down is mostly downhill, so the uphill test on returning wasn't as long as the outward-bound trip, but as you know when travelling travelling in mountains, there are ups and downs going both ways. Plus you don't generally climb hills with the gas pedal floored, so I felt that I pushed the car harder than I would normally anyway. anyway. So it would have to be a good serious and longer climb if I wanted to test for getting hotter. I got home with an empty tank and wanted to run AFR tests before before getting more gas since it will be a long time before I can get E85 again. So I installed the AFR meter and did some hard driving on the way to the gas station. I have to say that the experience was invaluable. From my studies, I would have thought that closed-loop mode was only running in light to medium throttle ranges, and the open-loop mode would kick in sooner for engine protection in the medium to heavy throttle positions and load conditions - maybe at 65% throttle or so. But I was surprised at how well modern FI cars do at keeping the AFR at stoich (and emissions down) under MOST driving conditions. I did some accelerations that were not WOT, WOT, but were pretty brisk stepping out with my foot deep into the gas pedal without triggering open-loop mode. In fact, I could only get open-loop by pushing the car pretty hard with maybe 80% throttle. But those were the readings I was most interested interested in. As long as the ECU was keeping the AFR at stoich, everything is great and there is nothing to learn. But I'm interested in the unsafe zone temperatures temperatures (Section 1). I found that using E85 in open-loop mode under high load, throttle, and acceleration, I was at Lambda (λ) values in the 0.91 to 0.95 range, which, translates to gasoline AFRs of of 13.4 to 13.9:1, which are right in the heart of the 13.1:1 to 15.5:1 unsafe unsafe zone. zone. Conversely Conversely,, after refilling with E20, I was at Lambda (λ) values in the 0.79 to 0.85 range, which, if it were gasoline, puts me more comfortably back in the expected gasoline AFRs of 11.6 to 12.5:1, or safely out of the unsafe zone.
So this leads me to the conclusion that E20 is as safe as gasoline per NREL and Brazil's certifications. E85 is an interesting interesting situation because the AFR is in what would be an unsafe zone for gasoline, gasoline, but the temperatures are not there to substantiate the mixture being too hot for the engine. I n fact, in light of the temperature graphs graphs cited above, I would have expected E85 temps near to gasoline temps, but my readings clearly show E85 running CHT temps a good 20 degrees less at WOT under heavy mountain-climbing mountain-climbing load. As previously mentioned, 20 degrees on the CHT should correspond to considerably greater differences differences on the EGT. EGT. If 5 degrees on the CHT is approximately 50 degrees on the EGT per the GAMI chart, we might reasonably assume that 20 degrees on the CHT is about 200 degrees on the EGT, EGT, which confirms the aviation post at E85vehicles.com E85vehicles.com stating that aircraft using AG E85 run about 200 degrees cooler than when using AG LL100 gasoline. It has been indicated that pending certifications, reactions of any one particular car are unique to the particular engine and EFI system. So right now, I have to say that my test results are preliminary, anecdotal, and valid for this car only. However, my low open-loop “unsafe zone” temperatures seem to suggest that anyone can run E85 since the old unsafe zone no longer appears to cause unsafe temperatures. Regarding Regarding the car's performance. At no time on the trip did I experience any of the coughing, hesitation, lack of power, etc that you expect of a lean-running car. But since I did not experience any of this, and my lambda values were within acceptable parameters, parameters, I was very impressed with the ability of the Bosch FI system to handle E85. The Hyundai owners manual does say that the car can run gasohol, but it is likely that back in 1998 they were thinking of E10, not E85. When I went out for the AFR test, I did experience one issue I hadn't had on the trip. At WOT I got what can best be described as an electronic clipping where the engine cut in and out at about half second or smaller intervals. I suspect this is the only limitation and failure of the Bosch system to handle E85. I don't know if this is a non-E85 related problem that just happened to develop at that time, but I suspect that it was just a matter of being the first time that I ever pressed the gas pedal to the floor hard enough to compress the carpet and floor mat. Possibly this triggered a WOT switch that I had not been able to trigger trig ger before. In my mind, I envisioned WOT as heavy throttle near the end of the pedal range, such as 80%+, but if WOT is only triggered by a switch at 100% Wide Open, I wasn't aware of that and can't positively say that this is what is happening. 13. Conclusion So in conclusion, the recommendation is that if an a n unmodified car will run safely and satisfactorily on E85 like the Hyundai or the Ohio Biosystems Biosystems Chevy Tahoe example, example, then don't use a kit. A kit is needed if the CEL on the ECU shuts down the engine. It is also needed where a cars won't run well on E85 and the owner is not satisfied to run diluted E85. So for these cases, the thing to do is NOT to just plug and play like the ads say. Rather use one with a potentiometer potentiometer and use an AFR meter and temperature gauge to dial it in to the best setting. The ads say the ECU will take care of it, but we know that in some cases the t he ECU is not helping in open-loop mode under heavy loads when the engine most needs it. So I would adjust that pot for Lambda around 0.85 under load and safe
temperatures. temperatures. Not only will these tests give you the most satisfactory satisfactory driving experience, but it will also give you the peace of mind that you are not running any hot temperatures.
