What follows is a very detailed description of the conversion of a 1998 VW Jetta TDI to burn straight vegetable oil. The oil could come from used fryer oil recycled from restaurants, or vegetable oil from various crops.
Diesels designed to burn standard diesel fuel need some modification to burn straight vegetable oil:
The vegetable oil must be heated to
about 160F before it is pumped through the injectors -- this is to lower its
viscosity to about the same as standard diesel fuel.
The engine must be started and run
on regular diesel until the engine internals and the vegetable oil are up to
temperature -- so converted cars have a diesel tank and a vegetable oil tank.
At shutdown, the vegetable oil in the fuel system must be purged and replaced by standard diesel fuel in preparation for the next startup.
This conversion was done using a kit supplied by http://www.FryBrid.com -- their website provides a lot of detail on how their conversion works and vegetable oil conversions in general.
Directory for this very loooooong page:
Installing the WVO tank in the trunk
Installing the heated fuel lines
Installing the under hood components
Installing the In Cabin FryBrid controls
Thanks very much to Jeff for providing this detailed description of the conversion. In the Frybrid forum, he posts as JeffNLisa, and is happy to answer questions or provide information.
Anytime I am filling by pouring from cubies, I use a long funnel that hangs out over the bumper, and that keeps drippies out of the trunk.
This tank is temporary. I always wanted an under-car tank, so as to keep all the trunk space available. I am now shifting my thinking to the idea of having several tanks in the trunk, one in the left side as a vacuum collector, one in the right side as a heated vacuum-boiling de-watering tank, and a main tank that uses ALL of the space behind the seat. I never use the fold-down rear seat anyway, and this arrangement would still leave some space in the middle of the trunk.
I do want to perfect such a system in the garage first, then in a truck where there is more room to maneuver, and only then consider it in a Jetta. Maybe by then, Chris Goodwin will have already perfected it, and I can save a few steps! But for now this is workable. And it is adequate room to fit in luggage, or as I have often done, cubies of WVO . . .
We left the HIH lines long so they could be easily moved,
and we’d have length to get to other location(s) if needed. We
used plastic ferrules for attaching the HIH to the tank, so it
could be removed easily when that time comes. And when putting
in the tow hitch, that time came much sooner than
anticipated!! See THIS thread
a discussion of the plastic ferrules, and see THIS thread
a discussion of the tow hitch.
When we first finished hooking it up, I didn’t run a line from the vent to go outside the car. Found out right away, that some WVO finds its way out the vent routinely!! So we took a canola oil bottle and made an overflow bottle out of it, running the vent tube into it, and then venting that to the outside through the floor of the trunk:
Any overflow is collected in the bottle, and then as fuel is drawn out of the tank creating a slight vacuum, the suction pulls any WVO in the bottle back into the tank. I used a wire tie to hold the bottle secure, so I can move easily if needed.
They turn toward the front,
Route snugly under the stock tank (notice a piece of fuel line split down the length is used to protect the HIH from the edge of metal),
plenty of clearance over the axle,
and follow the same channel as the stock diesel lines:
We opened the holding clamps and stretched them across the
lines and fastened them on either side.
Right at the firewall, they go into the engine compartment also along similar routing to the diesel fuel lines:
We considered that the heat from the HIH might hurt the diesel. But when running on diesel, the fuel is only in those lines for a moment, because it is flowing. And with no insulation or wrap to hold heat or transfer it to the diesel lines (like a HOH) we reasoned that not much heat would get transferred. When running on WVO, the diesel is stationary in that line, but only a few ounces of it would get any heat, and this is used immediately upon purge, replaced with cold diesel. When I tagged the picture I had 5,000 miles, an as of this posting I have over 9,000 miles on WVO, and have not noticed any issues with this. We intend to go back and insulate the HIH lines. When we do that, we will of course leave the diesel lines outside that insulation.
In the engine compartment the routing of the lines is hard
to see, and so the pics under the hood go in the order the
components are placed, rather than the order of flow.
To begin, under the hood we removed the stock air box. It takes a ton of room, and is only somewhat efficient. The TDI Club, I’m told, suggests against the use of K&N filters, but I have used a lot of them, and they are the best.
In the space left from removing the air box, we located the heated WVO filter low in the front right corner, and the HE about where the edge of the air box was:
The K&N we used is model RU-1460. No dealer or supplier we
could find stocks this filter. There is even a K&N warehouse
in Riverside, and they do not stock it there either! This was
only a concern because for special order items, you have to
pay first, and it’s a one-way ticket. If it doesn’t fit the
space, you’re stuck with it. However, we measured, checked and
measured again, and according to K&N’s catalog, the
measurements we took exactly matched this unit. Most parts
stores can order it, and we got in about 2 days. It is a
perfect fit for the space.
There were two drawbacks to removing the stock air box and using a K&N as we did in this application. One is that it eliminates the cold-air intake. (Placing the filter in the location of the cold-air intake led to another problem we had to solve, discussed in a moment.) And second, the un-insulated HE is here, and air coming to the K&N must pass by it if not over it.
But I have noticed no degradation in performance, and the greater surface area of intake with the K&N may also reduce or eliminate any associated problems. I plan to insulate all the Frybrid lines anyway, to keep heat in for WVO purposes. By the summer, if performance degrades, we have contemplated an air dam and tunnel that will more directly feed cold air to the K&N.
In locating the valves, we were concerned with appearance, but only to the degree that we could hide them under the cover. Aside from that, functionality was the primary criteria in locating them. We fabricated brackets to hold them in the desired locations. For the supply valve, as close to the HE as possible was important, so we would have a minimum of unheated line after the HE and before the valve. Then we wanted that whole assembly to be as close to the IP as possible. We got about 4 inches and less than 2 inches for the final result!!
Next came the problem of mounting the K&N. The stock air intake feeds through a plastic elbow, around the back of the valve cover to the turbo. Fortunately it is bellowed for about 10 inches, and we were able to cut it at one end about 3 bellows from the MAF, and cut off all but about 3 bellows on the other end, just before the bend. We then fit several bellows from the one end into the other, smeared some sealant in and wallowed it around to make a good seal, and achieved a perfect shortened intake!
For maximum efficiency and speed of the purge cycle, it is important to have the shortest possible return line from the IP to the return valve. We achieved just over 3 inches here:
When setting up the purge cycle we put the sight glass in
the joint at the end of this line, and then from the sight
glass into the valve. We timed the cycle for fuel-change to be
12 seconds for the fuel to “completely” wash thru to the
replacement fuel at this point. I have contemplated putting a
delay relay into the switchover process, so as to give the
return valve about an 8-10 second delay when switching over to
WVO, just to save the few ounces of diesel that normally feed
to the WVO tank at switchover.
The next most important thing in setting a good purge is to minimize unheated line after the valve and before the entrance to heated line. We could have located the terminus of the HIH for the return line right next to the valve, but it was much cleaner to locate it where we did:
But to be sure we didn’t leave any VO in unheated line; we originally timed the purge cycle to run diesel all the way to the entrance of the HIH. An additional 12 inches. By placing the sight glass there in the line, we timed that to be a total of 20 seconds. So an additional 8 seconds purge time to have the HIH terminus where we wanted it, for that additional 12 inches of unheated fuel return line.
However, Chris Goodwin later told me that it was safe to
purge only to pass diesel thru the switch valve, and leave WVO
in that short of a line. He said that by the time the car has
warmed up enough for switchover next time, sufficient heat
would travel thru the ounce or two of WVO in just 12 inches of
line to make it flow, and flow is the only concern. So we
reset the purge cycle to 12-14 seconds, and it has worked
To get the heat to the HE, at first we teed into the heater hoses that feed the cabin heater. When buying a kit off of eBay, I got one with parts and instructions for a different car, that IS supposed to be plumbed that way!! But we found when running it that virtually no heat goes thru. When I posted a question to ask, Chris Goodwin said that either an aux water pump or re-plumbing in series was the correct way to do a TDI VW. We had already tried it the other way and found that out for ourselves by the time I saw his answer on this forum . . .
So on my car, the coolant travels into the firewall to the heater core thru the hose on the driver’s side, and then out the firewall back to the engine thru the hose on the passenger side. (Apologies to Aussie Bloke, Tony from West Oz, WVO Downunder, et al. for the wrong orientation for them!) We spliced into that return line, and fed that coolant to the WVO system. In TDI VWs (mine at least) coolant flows thru the heater core at all times, and heat is diverted into the cabin or not, simply by controlling whether the fan blows air thru it or not.
Later we also found that we could still decide on-the-fly
whether to take heat in the cabin first or send heat to the
WVO system first. If you select heater on in the cabin, the
air will transfer all the heat it can from the coolant to the
air flow, just as if the WVO system wasn’t there. If you
select heater off, virtually all the heat travels right thru
the heater core and on to the WVO system. This is remarkably
efficient. I really can select it on the fly and take all the
heat, or give it all to the WVO system, or any combination I
feel like depending on how cold it is!!
The coolant travels over to the HE and enters there. The temp sensor is supposed to go in the line right before the HE, but we wanted to be safer at the beginning, so we located it after the HE and before the filter. So coolant flow is HE, temp sensor, then it goes into the filter. From there it goes into the HIH back to the tank, through the in-tank HE, then thru the other HIH back to the terminus where the fuel return line is. It then is routed back to the engine head.
The WVO comes from the tank, thru the HIH, thru the filter, thru the HE, the supply valve, the IP, the return valve, the other HIH, and back to the tank. The lines are routed tightly and closely, but fit in very efficiently.
Under the front of the right fender is the access to drain or change the filter element:
This is real nice, because the drips that leak when
changing it all land on the ground, not in the car. Also, when
changing the filter, I simply fill the new one to the top with
diesel or fresh filtered WVO, and at the next switchover,
there is never any hiccup from any air in the fuel line.
I was concerned about having the filter this exposed. I have contemplated putting either insulation around it, or even an electric or coolant heated wrap. However, as it was engineered, it stays blistering hot, even with cold wind that might hit it. I have had a few problems which may be related to cold in this filter, but have not been able to conclusively determine that to be the cause. Still, when I go to insulate all the lines, this will probably get an insulated cover for it.
As I said already, it sits in the space where the cold-air intake was. To protect it from direct air blowing across it we welded in an air dam to block the original path of the cold-air intake.
It is also visible in this shot, which shows a better view of some of the lower hose routing.
And from this angle you can get a good perspective of the filter location.
And another perspective:
This orientation of the filter and HE led to a problem of
how to flow the fuel from the filter to the HE. The path it
must take exits the fuel outlet of the filter, travels several
inches, then must make a 180* turn and into the HE fuel inlet.
Heater hoses are made that have 180* bends in them, but we
could not find any fuel line that does. But heater hose would
not hold up to fuel. We considered that it MIGHT hold up to
WVO, but there would certainly sometimes be diesel or
BioDiesel in the line. I asked on the Frybrid forum, and
several including Chris Gooodwin responded that even WVO would
break down heater hose over time. Chris suggested bending a
piece of aluminum line and using that. And that’s exactly what
Here you get the best view I can show of how that line goes:
With the K&N removed, you can see the routing of the diesel lines:
There were several electrical and structural arrangements that were designed for the location of the MAF. When we relocated it, at first we thought that several of those electrical and at least one mechanical assembly would have to be relocated. But in the end, they reached from the existing locations, and nothing had to be moved:
All in all, I think it’s a clean looking installation!
The location of the valves and components was designed not only for efficiency of operation, but also for cleanliness of appearance, and to appear either hidden or as close as possible to stock with the cover fitting back on:
And again, this is the “final product” view under the hood!!
In and through all this, I am not suggesting that these are the “best” locations or the only logical locations for these components to be located. I am simply showing where we decided to locate them, and why; and what we did to solve any associated problems.
And then into the cabin for the Frybrid controls.
I did not want any extra gauges in the cabin. I wanted everything to remain as stock looking as possible. So to begin with, we did not mount the provided VDO fuel guage. Instead we wired it in to the stock fuel guage, which I will show in a moment.
We considered a custom panel for the LED lights and button and switch. But we found an unused blank in the dash board that we could very tightly fit the lights and button, and although this was a difficult fit on the back, it looks very clean on the front:
I did not want the switch in view in the cabin. In my “previous life” with Mustangs, it is common for the cooling fans not work well in fox-body Mustangs, particularly the turbo charged ones. And so in nearly all of my Mustangs we had wired a fan over-ride switch, and located it just under the dash to the far right side. I was very familiar with how easy this location is to reach, while at the same time being hidden from sight. This is where we located the master power switch for the Frybrid system:
And the Frybrid controller is located just up under the dash, hanging right under the steering column. That area does not lend itself well to photographs at all, so here’s a “ghost” drawing of its location:
(Picture of Fuel Sender hatch in trunk -- to be added later here)
We then spliced the live wire that comes from the stock sender and attached it to the default closed position on a SPDT relay. We brought the live wire from the Frybrid tank sender, and attached it to the normally open position on that relay.
(Picture f wires coming from stock sender, and splice of sender wire from WVO tank -- to be added later)
We wired the trigger power to the relay to come from the power to the supply valve. That way, whenever the supply valve is not powered (fueling the car with diesel) the relay trigger is not powered, and defaults to its #1 position, which closes the link between the gauge and the stock tank sender unit. Whenever the supply valve is powered (fueling the car with WVO) the relay trigger is powered, and switches to the #2 position, closing the link between the gauge and the WVO tank sender unit.
(Picture/schematic of relay. This is a SPDT relay, with default to postion 1 -- to be added later)
We got some resistors and an electric multi-meter. After we
tried several resistances, we determined that even NO
resistance still would not allow the gauge to read all the way
to “full” with the float ball all the way up in the tank. This
was easy enough to check with the sender still in the tank; we
could use a piece of wire to reach into the tank and gently
lift the float ball to the ceiling. That only took the gauge
up to about 2/3 full reading. But I drove it this way for
several tanks, just to make sure that it was moving up and
down, and that it was switching to read the 2 tanks.
So we had to remove the sender unit from the tank, and adjust it. We determined that with the wire of the float ball all the way up to the max position on the sender, that the gauge would indeed read full. But then the ball would be WAY higher than the top of the tank! We found the spot where the “fuel” warning light would come on, and that would have the ball floating at over half full in the tank.
We emptied the tank and put in exactly 2 gallons. We used a
stick to dip into the sender hole, to mark the level of fuel
for 2 gallons, and then set that alongside the sender unit. By
trial and arror, we adjusted the height of the sender up and
down on its rail, and a bend in the wire that holds the ball;
and located it so that when the ball was where a full tank
would be the wire had the sender all the way up, and when the
ball was at the level of 2 gallons, the wire had the sender at
the point where the “fuel” light came on.
This alignment works perfectly, and gives 10-11 gallons of usage before the “fuel” light comes on, and gives a reserve of 2 gallons of WVO when the light does come on.
For testing, it was necessary to turn on the ignition to the run position, but not start the car, and short pins 3 and 4, so the green light would come on, and adjustments to the tank sender would register immediately on the gauge. (When the car is moving, changes do not register instantaneously, as described below. But when the car is still, they do.) When the ignition was on, the “door open” buzzer was irritating, so I facilitated a temporary disabling device, with the help of a golf club and a cubie of WVO . . . .
In the cabin, the fuel gauge works perfectly with the exception of a significant delay in changing readings when switching. In this photo, the car has almost warmed to switchover temp, but you can tell by the temp gauge that it is not quite there. And you can see the RED light on the Frybrid controller to indicate diesel. The gauge shows the level of diesel I have in the stock tank:
I resumed driving and once it warmed enough and switched over, the gauge then moved to show the amount of WVO in the Frybrid tank:
You can't read the clock on the dash board very well, but
the elapsed time is about 10 minutes.
The only problem with this operation when driving is that the gauge doesn’t move to the new level instantaneously, unless you are sitting still. The ECU takes readings from the sender every few seconds. When the car is still, it displays those readings directly to the gauge. This way, when you stop to fill up your empty tank, when you turn on the car, it senses the new level and displays the new level immediately. But when the car is moving, the float ball bounces around in the tank. Even with momentary readings, this would cause some erratic behavior on the gauge! So while moving, the ECU takes these readings from the sender every few seconds, and holds them for a period of time, and averages them for the reading it sends to the gauge. This smoothes out and eliminates the erratic behavior of the needle when driving, and the level of the tank drops relatively slowly, so that in the stock design each reading gets slightly lower as the fuel is used, and the overall average does go down so the needle on the gauge does go down as the fuel is used.
This functionality though serves to foil the switchover. When driving, if the ECU has been detecting readings near half let’s say, and then at switchover the readings are now at full, let’s say, it will add a full reading into the average, which raises the average a little, but not a lot. As more full readings get included into the average, the average gets to be larger and larger, and the needle gradually moves to the new level.
In practice, this takes about 10 minutes. I have gotten quite used to it, but does take about 10 minutes for the needle to reach the correct level after switchover when driving. After the switchover photo, I purged and the needle went back to the diesel level:
Then I switched off the Frybrid controller and switched it on again so it would switchover to WVO again, and the needle went back to the level of the WVO tank:
You can email Jeff at: email@example.com
One further comment from Jeff:
It should be noted that the Frybrid kit comes with rubber fuel lines designed for fuel that will withstand hot WVO. However, those lines are NOT designed to withstand BioDiesel, and it is important to know that if you install ANY WVO conversion kit, if you also plan to use BioD you need to install viton fuel lines instead of the ones I used, which are supplied normally by Frybrid. When they have stock, I believe Frybrid is willing to supply you with BioD-safe lines as an option, if you request and pay the difference.