Warning: converting a diesel heater to gasoline is NOT safe. First, the heater was not designed for it. Second, gasoline is much more volatile than diesel, so failures can have drastic consequences.
It also voids insurance. This post only describes my journey. It is not
a how-to, safety, or official guide. Do such a conversion at your own risk
Now that the controller and the pump got replaced to help convert the heater from diesel to gasoline, time to see if the combustion can be stabilized, and then attempt to tune it for various altitudes.
The default settings from the original green main board & controller were 4.4Hz (High) and 2Hz (Low) fuel pump pulse rates. The new programmable blue board came with: 5.4Hz @ 4300rpm (High), and 1.4Hz @ 1500rpm (Low).
Neither is suited for gasoline or high altitude use, so some tuning is required. How do I know ? Take a guess... 😅
Yeah, that's what can happen when the Fuel/Air ratio is too rich ! After a flameout, as the fuel pump keeps delivering fuel for over 10 seconds before the controller realizes that something is wrong, a nice BANG! can brighten up your day.
Hot spots in the combustion chamber, exchanger, or glow plug, can ignite the accumulated excess fuel. That's gasoline for you, much more volatile and dangerous than diesel. Stop reading this blog & don't attempt such a conversion if you value your life !
During the tuning, special attention will also be paid to soot buildup which usually quickly shuts down such heaters at high altitude.
Turn On
Running on the High heat setting modified to 4Hz @ 4500rpm with the new 16ml pump resulted in a much steadier combustion sound than the woop woops from the original controller and pump. Great, seems like a step in the right direction ! 💪
However, a couple of times per second, small disturbances in the sound could still be heard, indicative of a randomly unstable combustion. That was remedied by raising the fan speed to 5000rpm, and adding an inlet air filter and a 3' exhaust pipe. We now have a steady and quasi continuous woooossshh sound. ASMR heaven !
I suppose that the higher airflow increases the velocity of the swirl created by the ring of louvers feeding air to the combustion chamber. Which may disperse the gasoline squirts better and result in spreading the flame front in space, and therefore in time. But who knows, maybe it's simply thanks to the gasoline fairy...
Wishing that the controller allowed to explore past 5000rpm, though, as 4900 was barely perfect.
To try getting past 5000rpm, since the fan rpm is regulated via a hall sensor + magnet, the controller was tentatively set to the "2 magnets per rotation" setting instead of 1. That fools it into doubling the fan speed. Unfortunately, the High fan setting only has a 3500-5000 programmable range, thus now making the fan rotate at 7000rpm minimum. No thanks, am no Elon Musk, not fancying a RUD event... 😂
Maybe one day I'll design or find a fully featured & customizable controller to provide more flexibility for tuning parameters, automatic altitude detection & compensation, flame detection and a true room-thermostat function. Possibly inspired from the Afterburner or the Arduino Webastardo, though those also have limitations and might present some safety concerns. So for now let's make do and see how well this current setup can be tuned ?
Combustion Tuning Methodology
Professionally tuning combustion parameters requires expensive gas analyzers to measure air density, Ox, NOx, CO, CO2, HCs, etc. Not wanting to spend 1000s of $$$ for a 1-time tune up, know-it-all Youtube was consulted to see if crafty people had figured out ways to ballpark it.
First is the simple and toolless "Make it leaner until you drop 1 heat bar" method:
It likely works great to guarantee a very lean setup that prevents soot buildup. But it takes a lot of time to set up when moving to a new altitude, and might be too imprecise to ensure reliable and safe operation in varying conditions (inlet air temp, ambient temp) ? And it can notably decrease the heat output.
Instead, I settled on the CO measurement method. Device options here
Low CO emission probably does not automatically guarantee an optimal combustion, but it is a crucial safety requirement. And maybe it will get us close enough ?
Setup
Sweep Results
Sweeping the fuel pump rate resulted in the following CO Emission and Exchanger Temperature points. Test conditions: 5600' (1700m), 42°F (5°C), 5000rpm fan speed, 10 minutes of settling before each measurement.
|
Green: adjusted for a hypothetical 16°C (60°F) ambient temp |
Sweet mother of OpenOffice charts, what an ideal CO bathtub curve ! 😍
4 Hertz seems to be the sweet spot, with leaner combustion to the left of it, richer to the right. Anything between 3.8 and 4Hz would probably be best to limit carbon deposit and slow down soot buildup. We'll want to check that once the tuning is done.
Anecdotally, it was also possible during the sweep to get a rough sense by ear of the quality of the combustion. At very low values like 3.4 the woooossshh became very quiet, raising the risk of a flameout. And starting at 4.3 hoarse-throat like noises showed up.
So using the CO method by optimizing for lowest emissions does seem to correlate well with combustion stability. Sweet !
Also, obviously, the leaner the mix the lower the combustion temp. Which, incidentally, could also be verified on the controller's display where very low pump rates resulted in losing the top heat temp bars.
But happy to report no explosion this time !
So, with all limbs still intact, this setup was ready for a sweep at higher altitude. Onward to the Rockies: 8800' (2700m), 5000rpm fan speed (tried lower again, but it was still best), 19°F (-7°C):
The optimal pump rate is now 3.6 Hertz. That's a 10% decrease in fuel rate for a 57% increase in altitude.
The lowest CO emission value jumped from 90 to 130ppm. At sea level it'd probably be around 20ppm, in line with David McLukie's CO video above.
Household CO alarms go off at 20-25ppm so, obviously, it is a good thing that the exhaust pipe will be outside the van. Courtesy of the heater being of the exchanger type with complete separation between the burner's air circuit (outside the van) and the breathable heated air circuit (inside).
We'll just have to make sure that when parked there is plenty of clearance around the Ford Transit for the exhaust gases to dissipate. Don't turn on your diesel heater if your van is in a shed or surrounded by snowbanks.
Now, in case anybody missed it, the fantastic news from these results is that even at 8800' the combustion was stable. So there we have it: confirmation that, if soot buildup can be kept in check, this gasoline conversion will work at our high altitude target of 8000' 🥂 Phew, what a relief !
High Heat Setting
Having acquired 2 altitude points, a pump pulse rate line can now be drawn. It will be used to determine the appropriate High Heat value when moving the van to a new altitude. The resulting slope is 0.000125Hz/ft (0.00041Hz/m).
Since the controller's minimum increment is 0.1Hz, the steps represent the values to program. And substracting an additional 0.1Hz or so can further help keep soot under control, if need be.
Just for giggles, similarly charting the exchanger's temperature results in a 0.0025°C/ft (0.0082°C/m) line:
Low Heat Setting
Since these parking heaters choke up at high altitude due to soot buildup, running on the Low Heat setting should be avoided as a lower temperature in the combustion chamber, and a slower airflow, promote soot generation and do not flush it out as efficiently.
So the strategy for the Low heat setting is to tune the fuel rate at the max fan speed (2000rpm?) allowed in that mode. This should result in a higher heat output, i.e. lower risk of soot buildup, than if using the much lower default fan speed. 1.9Hz seemed to work well at 5600'. Which can then be derated by 0.00125Hz/ft at different altitudes.
The Low pump rate could have been meticulously swept like the High pump rate. But a rough estimate is probably good enough since the operating procedure below will hopefully avoid running at the Low setting.
Soot Buildup Check
Now, let's see how much damage was inflicted after all these modifications and tuning runs ? Since the initial turn-on the heater experienced a dozen ~1 hour long runs, while changing controller, pump, settings, altitude, etc. Gasoline only, never saw a drop of diesel except during QA at the factory.
So it was abused quite a bit, with a handful of flameout events, an explosion, and about half the time spent in unstable or too rich or too lean combustion conditions.
This said, before shutdown, most of the runs were ended with 10 minutes on the High setting, with a fuel pump rate set for a lean but stable combustion. That should theoretically have helped clean up some of the carbon deposit.
So, with all that in mind, let's see how many battle scars this fire breathing dragon now sports...
First, the exhaust pipe has a very thin light brown coating (pic doesn't do it justice). No soot. That's a very good start 😀
No sign either of combustion gases backflowing or backfiring into the fresh air pressure chamber. Absolutely pristine !
The glow plug shows some thin carbon deposit that flakes away. The atomizing mesh screen is a mixed bag, with clear signs of carbon deposit at the bottom (left), but even there the mesh is still fully open when checked against a light.
And the combustion chamber holding the glow plug and the fuel screen is very clean:
The exit of the combustion chamber, and the inside cylinder, show no trace of soot or carbon deposit. Now THAT is an excellent sign !
And finally, the most susceptible location for soot buildup, the exchanger, shows no sign of soot on the exhaust side (left). However, a bit did settle on the fins on the side opposite the exhaust (right side of the pic), where the flow from the exhaust gases is the slowest. But it is very light. I won't even bother cleaning it before reassembly 😅
Conclusion: the results blew away my expectations ! Super clean despite all the abuse. So we'll just replace the mesh screen and the glow plug out of an abundance of caution, reassemble the whole contraption, and leave it alone from now on.
This fire breathing apparatus from hell should now be able to shoot flames for a long time without getting a sore throat. Sweet !
After all these weeks obsessing over countless reports on the web of soot buildup fast killing these heaters at high altitude, our modifications & tuning seem very promising. So it is clearly time to put this worry to rest and go overthink other areas of this van build 😂
Temperature Control Constraints
For now, the heater will be used in the manual on/off mode.
It does have a thermostat mode, but that constantly runs the fan at low speed, even in standby when not burning fuel. Because the thermal sensor is on the main board in the heater, not on the remote controller. So the ambient air must be constantly circulated through the heater to measure the room temp. Idiotic design, a waste of electricity, and unnecessarily noisy at night.
Another issue is that after running on the High setting until the desired ambient temperature is reached, these heaters switch to the Low setting for a while before shutting down. That's a no-no at high altitude, as it increases the risk of building up soot.
Unfortunately, these heaters can not be controlled with a regular thermostat. Their Start / Stop button just triggers a pulse signal. So a thermostat would either need to send a compatible short / long duration signal to start / stop it, or yank the power off the heater.
Yanking the power off would be outright dangerous 😱 It prevents the heater from going through its critical ~1 minute long power-down sequence that turns on the glow plug again to remove potential fuel excess from a flameout event, and that cools down the exchanger to avoid cooking the main board.
Ahem... in other words: our control process just has to keep a few balls in the air. No biggy, right ?
Operating Procedure to limit Soot Buildup
Coaxing the heater to run a long time on the High setting is done by simply setting an overly high ambient temp target on the controller, like 90°F (33°C). That ensures a long runtime with a high enough combustion temp and fast airflow to eliminate startup soot.
This will get the brunt of the heating work done when turning on the heat in the morning or after a day out frolicking in the snow.
Once the large thermal mass of the walls and cabinets has soaked up the heat, and a high wall surface temp has been reached, the supplemental ~900W electric heater can be turned on to maintain a steady ambient air temp. The electric heater is automatically controlled by the room thermostat.
This said, this approach is clearly not optimal for comfort as it may lead to yoyoing temps.
It also can't be used at night with people trying to sleep, and won't allow maintaining a steady temp in the van when not present during the day. However, once more familiar with using the gasoline heater at high altitude, and if soot is successfully kept at bay, using the controller's thermostat mode, followed by a daily 10' High setting purge run prior to shut off, might sometimes be appropriate.
Oh well, maybe I'll automate this process one day to simplify it down to a user friendly room-thermostat-but-failure-safe like behavior. A few people have prototyped electronic doodads in that vein. But, honestly, why is my TODO list never shrinking ? 😢
Limitations of this Tuning Method
Obviously the tuning method used is not optimal. It neither ensures the highest heat output possible, nor the cleanest combustion.
CO was plotted against altitude when combustion only cares about barometric pressure. I should have tracked the BP at the time of the tests
Measuring just the CO does not tell us much about other combustion results like unburned HCs, NOx emitted or the Ox/CO2 ratio
A simple line was drawn between 2 points to derive the fuel pump rate at any altitude. Additional measurements at intermediate and lower altitudes would be helpful (but both oceans are 1100 miles away 😂)
The outside temp likely plays some role in the performance of the combustion. Turning on the heater on a -15°F (-26°C) day probably impacts things differently than on a 50°F (10°C) day
(*) The heater automatically decreases the pump rate further if the exchanger's max temp limit is reached
(**) No cold startup tuning
In other words: any of my settings above might just be horse manure ! A cat chasing after a laser pointer could possibly stumble on as good a set of values... But at least my numbers are supported by a science-y looking write-up and authoritative charts !
Anyhoo, feasibility has been proven, settings have been science-ingly optimal-ized, and all limbs are still intact.
So, I am declaring this gasoline & high altitude conversion attempt of a cheap 'Chinese diesel heater' a RESOUNDING SUCCESS ! 🥂🥂🥂
Time to procure a 5" hole saw and drill a big hole in the floor of the Transit to install the heater... Will probably have cold sweat for days before I can muster the courage to do that !!! 😂
[Update Aug'23] Would I go through this whole conversion ordeal again ? It was a super fun learning experience but, in the meantime, a new option showed up: Velit released a $650 4KW gasoline heater with true thermostat control (unit completely turns off once temp reached) and barometer for operation at up to 4800m (allegedly). Same casing / cast / fan as the cheap diesel heaters, but proprietary combustion chamber & electronics. In-depth review here: Mispronounced Adventures
So yeah, even though it's oversized for a Ford Transit van, I would have gone that route instead, had it been available back in Fall 2022 !
Finally, now that we have a working gasoline heater, time to install it in the camper van.
Posts in the Diesel-to-Gasoline heater series:
- Installing the Gasoline Heater
- Installing a Fuel Pickup Line in the Ford Transit's Tank
- Gasoline & High Altitude Heater Conversion - Success !!!
- Programmable Controller & Pump for Diesel Heater
- 'Chinese Diesel Heater' - is it Good out of the box ?
- High Altitude Gasoline Heater for Camper Vans ?
(*) That can happen in mild weather, or with a rich mix. When the exchanger reaches Tmax, the controller switches to a very lean mix and CO shoots up.
It actually happened a few times while testing rich mixtures: the CO would
shoot up 2 to 4X for several seconds before coming back down, and
repeat 10 to 20" later. I dunno of a way to tune for that,
short of drastically reducing the High setting's pump & fan rates to lower the heat output
(**) Startup was a stuttering business, even sometimes explosive after a flameout 😨, when using far out-of-tune settings. Things improved markedly
with the tuned settings. But those were optimized for steady state burner
conditions, after settling for several minutes. They aren't likely best for
the rapidly evolving combustion chamber conditions of a cold startup. For
instance, gasoline engines require a choke during startup. No such mode
here. So we might find out later in the field that further tweaking is needed
in some cases
Comments
Post a Comment