Converter Temperature

[Frank]

From my experience with experimenting with coolant flow through the Model E converter mounted on my 1977 Pontiac, I found that running cooler coolant temperatures had a detrimental effect on my fuel consumption. On a different post about gas valve profiles, franzh confirmed that cooling the propane has the effect of making it denser, requiring a fuel mixture adjustment change. As the cruising fuel mixture is not adjustable on 425 mixers, I wanted to investigate this further.

While searching for more information about converter temperatures, I came across two postings on http://www.eng-tips.com (http://www.eng-tips.com/viewthread.cfm?qid=47012 and http://www.eng-tips.com/viewthread.cfm?qid=53767). One of the interesting points I read in the qid=47012 posting is the following quote:

The primary benefit to lowering the (propane) temperature is not the air charge (although that is a positive benefit) but that you reduce the formation and accumulation of hydrocarbon heavy ends, semi-solids, and paraffin’s. They tend to form at temperatures of 140 deg f. Some severe cases have resulted in vaporizer stoppage in as little as 30 hours of operation.

I tried searching the internet for more information about converter temperatures and about the formation of hydrocarbon heavy ends. I did not find any collaborating information so I tried emailing some petrochemical companies and Impco once again.

I received a reply from ExxonMobil and they write that their fuel department has the following answer:

The ASTM D1835 special duty grade, GPA 2140 HD5 grade and Canadian CGSB grade 1 specification for auto propane all control the evaporative residue with D1837 max temperature at 95% evaporated or by max "butane and heavier" of 2.5 vol% by D2163 GC. The intent of these specs are to ensure that marginally heavier hydrocarbons (butanes, pentanes, hexanes etc.) that are present from fractionation are not present in amounts that would result in excessive collection in vapor withdrawal applications on multiple refills and multiple single plate distillations from vapor withdrawal. Auto propane applications are mostly liquid withdrawal, and heated regulator/vaporizers (commonly called "converters"), so this specification generally does not apply.

Oily residues are controlled by the D2158 evaporated residue and oil stain. This is intended to limit the amount of oily residues that may be left behind at the point of vaporization, including in auto "converters". Sources include traces of compressor oils, pump lubricants, hose plasticizers and other materials that are invariably present at trace levels in all fuels. The specification limits are sufficient that this is generally not a problem in most applications, but could be an issue for equipment that is overly sensitive or severe. For example, most auto propane regulator/convertor usually have enough liquid propane transients to wash any evaporative residues into the engine where they combust along with the fuel (and any traces of engine oil from valve guides etc.). Some regulators have specific provisions to ensure this happens, such as mounting requirements that allow any residues to gravity drain to the converter exit, and no low point in the "dry gas hose" that connects the converter to the carburetor or inlet manifold.

Some newer gas and liquid injection systems have had problems with residues, and recommend the use of solvent/detergent additives to control intake system deposits, similar to gasoline. These are generally solvent oil types, intended to keep a small amount of liquid present to continually flush small amounts of residues into the engine (very similar to solvent oils in current gasoline Deposit Control Additives).

Detergent additives are not added to propane intended for general distribution,, as they are detrimental to most other LPG applications, as they are non volatile and collect at the point of vaporization, for example in a bbq tank or a low temperature vaporizer.

They then followed up with the following message:

Directionally lower temperatures will leave more liquid in the converter and tend to provide more liquid for "flushing" on startup. However there is more that could collect in an improperly installed dry gas hose with a low point, or there may not be sufficient heat in cold weather. Similarly, higher temperatures will vaporize more materials, but the smaller amount that remains will be thicker and more "baked on". The best way to go may be different with different converter designs and even driving cycles (for example the difference between a car with lots of cold starts and a forklift with 24/7 hot operation. May I suggest you consult with the individual converter manufacturer?

I also contacted PetroCanada and they replied with the following answer:

It would take more than the temperature experienced in the engine to convert propane to oil. If it were easy to convert propane to oil we would be doing this at refineries. When propane is certified at the refinery tankage it is tested for purity. Only dedicated lines can be used for propane and trucks that carry propane cannot carry liquid fuels. The only problem I have heard of in relation to oil residues in a propane vehicle relate to oil being deposited on the cylinder walls since is not washed away with the liquid fuel mixture as in a gaseous engine.

I asked the technical department at Impco if the temperature of the water reaching the converter make a significant difference to the operation of a propane system. They replied with the following answer:

Not a significant difference to speak of. During the process of vaporization some heavy ends will collect within the regulator over time. This can be reduced by mounting the regulator with the outlet pointed down and if there are high coolant temperatures a "Thermstat" can be installed in the line coming out of the regulator which keeps the coolant at about 160° F in the regulator chamber

The "Thermstats" are available from Gann Products in CA. Their number is (562) 862-2337. It is an inline thermostat that can be purchased with 5/8 " hose barbs on both ends or 3/8 female NPT on both ends. The regulators produced by IMPCO have a working temperature from -40° to 250° F. If you are working with vehicles that are feedback controlled from the OEM factory on gasoline, you need to use the recommended thermostat because the engine will not go into closed-loop until a certain water temperature is reached.

Therefore, from what I have found about heavy ends, I think that it is more likely that the heavy ends are always present in propane fuel and will always accumulate in the converter if the converter is oriented in such a way that doesn’t allow them to drain into the mixer. Since the converter temperature also has an effect on the fuel density supplied to the mixer and thus fuel consumption, I think that one solution would be to install a control valve in the line supplying water to the converter.

In my opinion, the converter should ideally be supplied with water at a constant temperature. The control valve could be either a globe valve with a hand wheel or a Thermstat. With the hand valve, I would try running the converter with the valve wide open and establish a baseline fuel consumption. After determining the baseline fuel consumption, I would then close the valve by a half turn and measure the new fuel consumption. I would keep repeating this test until I saw that the fuel consumption began to decrease and then I would restore the valve opening to the point where fuel consumption were the lowest.

I have no experience with the Thermstat valve but have asked Gann Products for more information. I would like to know if anyone has any experience with its reliability and its effectiveness.


Frank

[fordCourier]

I had a cooling system routing problem where the coolant wasn't circulated back to the block fast enough. You could feel the coolant in the hoses to the convertor boiling after the engine was shut off. I rerouted the coolant for more flow through the convertor, and the temperature was more stable throughout the system, and starting was easier.

I talked to a local impco distributor/installer who did have a 195 inline thermostat he wanted to unload but I declined. My 140 ci. 4 banger doesn't seem to get hot enough to really push temps to boiling. Besides, in a closed system like an engine cooling system, shouldn't you rely on the thermostat that is already in place?

The Convertor doesn't need very hot water to convert the liquid to a vapor. So, 195 degrees is a bit of overkill. The important thing, however, is the temperature of the gas at the point of fuel-air mixture. As the mixture gets hotter, it loses some capacity to hold air and fuel as these expand.

I found some comments from an Aussie that they had been engineering ways to cool the gas mixture to the engine. They said the sweet spot was around 70 degrees.

Safecontrols.com offers a temperature stabilized converter which looks like its modelled along the same lines of the Model J. Apparently it will regulate the gas temperature to 70F. My installer told me that it works to some degree but is not perfect.

So coolant temperature will affect convertor output to a degree but at the same time the installer told me the best thing for it is to get coolant flowing through it and up to operating temperature as quickly as possible. My 5958 A/F mixture control negates the need for any rich/lean compensation components as it selects the right mixture regardless of other factors, and begins working as soon as the sensor heats up (self-heating).

There are other aspects of the original gasoline-calibrated design which detract from its efficiency runnning propane, such as coolant passages in the intake manifold which further heat the fuel-air mixure before it is combusted. If I ever get around to it I might consider closing these passages off. Ideally if you could reduce the temp of the fuel air mixture to 70F by the time it gets to the combustion chamber, the more power you will have.

[Frank]

Thanks for the advice.

I have rerouted my heater lines so that there is a good source of hot water for the converter. Since the air conditioning system can shut the hot water off to the heater core, I needed to ensure that the converter has a constant source of hot water.

Originally, the car was equipped with a temperature-controlled vacuum switch to shut the vacuum off from the EGR valve when the engine was cold. This provided a spare hot water supply port for the converter but unfortunately GM did not use NPT threads here. I removed the switch and broke off the plastic part of the switch and welded a pipe coupling to the remainder. Into this newly-made adapter, I screwed in a valve so I could control the water flow through the converter, independent of the flow through the heater core. The water returns to the engine by a Wye in the 3/4" hose from the heater to the water pump.

I have this valve fully open right now and plan to gradually close it until I notice there is a drop in my fuel economy. Since the flow through the converter displaces some of the flow through the heater core, I hope that I might even get a bit more winter heat as well.

Anyway, I believe the 195°F thermostat is probably designed for newer cars that run much hotter thermostats than our vehicles. Those newer cars have sensors that require the car to reach its operating temperature (hotter than 195°F) before the computer can operate properly. That converter thermostat would prevent the converter from getting too hot.

As for controlling the gas temperature to 70°F, I think that would be dependant upon the the engine and the mixer. We really have no idea exactly what fuel mixture the engine is getting because Impco regards it as proprietary information. By doing like you did and upgrading to an electronic fuel mixture control system, you could at least be fairly sure that your engine will be getting a stoichiometric (ideal - not rich nor lean) fuel mixture under most conditions. The electronic controller works best with a constant water temperature.

Tom Jennings (1963 Rambler straight propane conversion) told me that he is using an Autotronics (similar to yours) controller with his standard (non-feedback) mixer. Since these controllers can only lean-out a fuel mixture and has no way of making it richer, Tom hasn't noticed any difference in the way it drives and his mileage has improved.

Once I get the starting problems sorted out, I'm going to start looking into getting an electronic fuel mixture controller too.

Frank

[Guest]

Tom Jenning's page has been a source of good information for me as well.

Right now I'm upgrading to a model L from the J to see if that helps me with my current problem: the computer doesn't "React" quick enough and I run rich/lean/rich/lean at idle instead of the computer holding it there.

I would take a look at some European control solutions as well. I read that over there they use a system for fuel control that is based off of manifold vacuum signals and is much simpler and more reliable. I got a good deal on the 5958 but it is a black box and requires a lot of wiring and install work... and it didn't work right from the get-go (two wires out of the black box were switched) and the documentation is somewhat sparse.

So if the regulator upgrade doesn't work I will have to do more detective work on the model L.

I'm also rerouting the coolant lines again to a completely straightforward orientation. I've got several things working against me giving me hot-start problems. The converter is higher than the engine, too many y-pipes, and return line through the intake manifold.

What happens is I can feel the coolant boiling in the lines after shutting down during heat soak, and can't start the engine. I've removed the EGR completely and blocked the hole to the air intake, removing a possible air leak and heat source to the intake side (exhaust pipe).

I plugged the intake coolant ports, and am going to route the flow directly through the core to the converter and back into the block via the temperature sensing location.

[Frank]

From what I can tell about feedback control systems from both Impco and Dualcurve, they are designed to cycle back and forth between rich and lean. I think it has something to do with the response of the O2 sensor to exhaust oxygen content and even gasoline engines hunt around to some extent. Does this cycling cause a drivability problem for you or is this something you noticed with your multimeter?

Also, I believe that both manufacturers use manifold vacuum to assist the feedback control system in controlling the fuel mixture. If the converter is properly sized for your engine, I wouldn't expect a larger converter to work better. Have you asked for tech support from Dualcurve? I found the people there to be quite helpful and they might have an alternative solution for you.

The converters on both of my propane cars were mounted above the engine and, for winter operation, I was running a 195°F thermostat. I never noticed a problem with hot water temperature causing a hot starting problem. Is there a chance that the idle mixture setup on your controller might be causing the hot-start problem?

If you have a V8, blocking off the intake coolant ports isn't a good idea as that is the return route of the engine's cooling water to the radiator. Blocking off the exhaust cross-over passage is a good idea and this will also eliminate the operation of the EGR valve. What engine do you have in your car?

Photos would be a great way of helping us understand how your engine is set up. Also, please feel free to join the forum as a member.

Frank

[FordCourier]

Well I installed the Model L, electric primer, and rerouted the coolant lines to some success. I still had the hunting idle, so I hooked up a digital multimeter to the leads going to the vacuum solenoid on the converter. No matter where I put the idle air mix screw on the CA125 mixer it was above the ideal 4-7V range. I could not get it below 9. The computer was showing that it was controlling the mixture, it was just too high. Frustrated, I took out the AV-1447 "rich feedback" air valve and threw in the old dirty one, dialed in the idle mix screw all the way and out 3 full turns. I started it up and it idled in range!!!

I was able to dial in a perfect range and got excellent results. Now the solenoid constantly clicks where before it would click, stop, and click. Fabulous!

The Model L was kind of a pain to cram in the existing spot and I had to fabricate a bracket but unlike the model J it had room for the primer. The Primer is a little solenoid that screws into the port on top, plumbed to a vacuum pickoff right under the mixer air valve opposite the controller solenoid pickoff. The primer is wired to the ignition start circuit and is quite effective.

The other changes were removing the EGR and switching the coolant outlet from the block to the back of the engine where the temp. sensor was. I put the temp sensor in the intake port to see what it would read, which was around 130-140 using the standard 195 degree thermostat.

I thought long and hard about moving the coolant outlet from the intake manifold to the lower rear of the block but feel pretty good about it for several reasons. Before, the coolant was being pulled across the cylinders, through the intake, and right into the converter, then the heater core, and then all the way around to the front. 1, thats a long way to pull coolant and 2, I was putting the hottest coolant directly into the conveter and intake manifold, likely increasing temps and decreasing volumetric efficiency of the fuel/air mix for no good reason, 3, on an inline 4 cylinder, why not pull from the back of the engine instead of the top-side for better overall circulation.

I plumbed it using the shortest possible path as well meaning the heater core and converter are not in paralell but in series, which is something I can live with since the blower motor in this particular truck is of absolutely no use as a cooling fan. The converter/heater is in parallel with the block though via a 1/4" hose from the block to the waterpump, using combination fittings (1/4 and 5/8 outlets) just in case I shut the heater valve (you never know).

From what I can tell about feedback control systems from both Impco and Dualcurve, they are designed to cycle back and forth between rich and lean. I think it has something to do with the response of the O2 sensor to exhaust oxygen content and even gasoline engines hunt around to some extent. Does this cycling cause a drivability problem for you or is this something you noticed with your multimeter?

Yes, it causes a drivability problem to some degree at very low speeds in gear because my truck has tall gears and it tends to make the truck surge. Since ditching the AV-1447 and putting the old one in (no indentifying marks on that one) its much smoother but still hunts a little (Tom Jennings site says to drill out the calibrated leak orifice to get the cycling under more control, but I'm fully within range right now so thats on the back burner.)

I talked to a semi-local Alt fuels installer who sold me the SafeControls 5958 and the AV-1447. We tried drilling out the orifice to gain control over that setup but it didn't work. From what I understand there are several air valves, and given all the possible cimbinations of engines and components, one will work best. I seem to have found mine via serendipity. The only identifying mark is on the tip of the "cone" on the underside that shows the number "7".

Also, I believe that both manufacturers use manifold vacuum to assist the feedback control system in controlling the fuel mixture. If the converter is properly sized for your engine, I wouldn't expect a larger converter to work better. Have you asked for tech support from Dualcurve? I found the people there to be quite helpful and they might have an alternative solution for you.

Maybe they do, but this particular unit (5958) is strictly electronic. The 5959 features a throttle position sensor for even greater accuracy, I suppose. My setup is a 2.3L ford Pinto/Mustang type motor of 1979 vintage, with nothing fancy like TPS.

My alt fuels guy tipped me off that the converter might be undersized from his experience seeing Ford 2.3L factory with a model L. I obtained a 2.3 OEM manual for a Ford Granada on eBay that was issued for propane in 1982, which showed clearly the Model L and the electric primer. I got both of those from the "Propane Guy"'s website from Canada. This 2.3 is only rated to around 90hp and the model J topped out at 100hp max. Either way, the L has the nifty primer port so I'm happy.

My other complaint was the engine running a bit out of steam at freeway speeds, and it may have improved somewhat but I'm guessing it has more to do with running propane on a gasoline distributor curve. The stock vacuum advance maybe allowing too much advance at high speeds, hurting my power. My alt fuels guy suggested replacing the vacuum advance with mechanical advance for this reason.

One thing I didn't test that was hypothesized was the action of the calibrated leak to the relative size of the converter diaphragm. Would the L be more effective than the J because of greater surface area/force to overcome, given the same relative size of the leak hole vs. intake pulse?

The converters on both of my propane cars were mounted above the engine and, for winter operation, I was running a 195°F thermostat. I never noticed a problem with hot water temperature causing a hot starting problem. Is there a chance that the idle mixture setup on your controller might be causing the hot-start problem?

Well, for one thing I still don't know if its explicitly the heat causing the problem, just that the heat and the starting problem are coincidental. I just ordered a 160 degree thermostat after doing some thinking and checking out Franz Hoffman's site who recommended running the engine between 160-180.

What I conluded is this- theres no reason not to run it as cool as possible. The only reason to run it around 195 (meaning around 212 fully open) is to compensate for the properties of gasoline. Propane's combustion temperatures are higher than gasoline so I figure it needs all the help it can get, especially without EGR for additional cooling in the combustion chamber. A colder engine develops more power and wears out oil slower. And a colder engine should to some degree shorten the post-shutoff heat soak time period.

As this project progresses, it seems to me that the small incremental changes all add up. The other small change was plug gap- I had them gapped to .038 and cut that down to .028 which increased running voltage and drivability. I also played around with ignition timing- set at TDC it had wonderfully smooth power up until about 30. The engine would idle even with the timing backed off 50 degrees! I found a median setting which works fairly well as I said but it is a compromise. Too close to TDC or stock at idle and top end power suffers, too far away and low end suffers. I really believe a lot of power loss is simply due to the advance mechanism and am looking into switching to a different mechanism.

I'll take some pictures hopefully tomorrow and upload them so you can see the set up.