Sunday, 15 November 2009

Friday, 29 June 2007

Coolant heating diagram for SVO/WVO/UCO/UVO fuel systems

I thought I would make a coolant heating diagram that shows how to tap into the coolant system of a Mercedes 300D. The same principle will apply to most vehicles. Just remember that the veg oil heating circuit and the flat plate heat exchanger (FPHE) must be in parallel with the main coolant circuit, and not in series. If you connect it in series then you need to have your cabin heater on at all times to keep the coolant flowing through your FPHE:)

Vegetable oil flows through the FHPE on its way to the filter, and takes up heat. Typically with these FPHEs, the oil is heated to within a degree or two of the coolant's temperature. They are extremely efficient and far better than the copper coil variety which can also cause copper ions to enter the veg oil and contribute to oxidation and polymerisation problems.

If you need to heat your tank as well, you could add a 2nd FPHE in series with the 1st FPHE and use that to reheat the oil in the veg oil return line before sending it back to the tank. Nice and simple, and it doesn't heat the oil too much; just warm enough to keep it nice and liquid to aid the flow and reduce problems associated with the oil's high melting point components clogging the lines and fittings.

While on the subject of heating the veg oil tank, I've seen some designs that extend the main coolant circuit all the way back to the veg oil tank where it heats the oil via a coil, before sending it back to the engine. This is not a particularly good idea for the following reasons:

  • increased risk of rupturing the coolant and damaging your engine.
  • if the heater coil is metal, particularly copper or aluminium [aluminum to our American friends ;) ], it will contribute to oxidation/polymerisation problems in the tank, because these metals are very good pro-oxidants.
  • It could heat the oil too much, and again contribute to oxidation/polymerisation problems.
For a larger pdf version, please click this link

Wednesday, 27 June 2007

Wiring the Glow Plug Heater

Wiring the glow plug heater was pretty simple. A mate of mine who is an engineer helped me with this as I am not a very good electrician!

From Jaycar Electronics I purchased the following:
  • 100 degC thermo-switch (normally closed)
  • Relay (234 12vdc)
  • Wiring loom for relay (with 15amp fuse)

100 Deg C Thermal Circuit Breaker Switch:
AUTOMATIC RESETTING BI-METALLIC DISC, SNAP ACTING. This phenolic-housed unit attaches to any metal surface. When that surface meets or exceeds certain temperature, the SPST bi-metallic element goes open circuit, i.e. unit is normally closed,opening on temperature rise. The unit resets automatically when temperature falls below specified limit.

Relay 12vdc:

Relay loom:

It was important to choose a "key on" power source as I didn't want the heater on with the ignition off!

Here is the wiring diagram:

Click this link to download a larger pdf version: Glow Plug Heater wiring

Tuesday, 26 June 2007

Changes to the fuel diagram

Click the images to enlarge them :)

I've finally drawn a decent fuel circuit diagram for the 300D.

For a larger pdf file version, click this link 

Changes made to the original setup are:

1. Added a glow plug heater just before the injection pump (IP). This modification allows me to change over to veg after only a couple of minutes from start-up. Ideally, we need to try to achieve a temperature of 90degC at the IP. At this temperature most veg oils will have a viscosity that approximates dino-diesel, for which all modern IPs are designed.

My home made glow plug heater is made from commonly available plumbing parts and uses a single 14V pencil-type glow plug. It's installed in between the lift pump and the IP and is controlled by a 100degC thermo-switch, so it is constantly cycling on and off. A fused relay ensures that if the thermo-switch fails the GP will not "run-away"and burn unchecked. In actual fact, I thought it was a 24V volt GP as that is what I asked for, but have since noticed that 14V is etched onto it????!!! Previously I was blowing the 12V variety quite regularly (11V etched on that GP). Anyway, this single 24V/14V (whatever) plug has worked flawlessly for the past 10 months and does an excellent job at heating the oil very quickly.

I think the secret is to use a nice lump of brass for these GP heaters. They heat the oil that passes though instantly, but as soon as the fitting gets to your selected temperature, the housing does most of the heating and the GP doesn't have to work as hard, so cycles off and on less often. Also, it retains a lot of it's heat if you have stopped for 1/2 hour or so. It will do an even better job if you take the time to insulate it.

You will see that I've used a "3-port"plumbing fitting. Suitable reducers were found to match the threads of the GP and the hose fittings. The GP can be installed in any of the "ports". I chose this configuration to suit my layout and to facilitate needing to change the GP easily.

I added some thermal grease to facilitate better heat conductivity from the brass body to the thermo-switch.

BTW, I am getting this reliability from a $10 GP off eBay!


  • Strip the paint off the GP before installing it or you run the risk of the paint entering your fuel line.
  • Ensure that you prime the fuel lines properly. Air in the fuel line might cause premature failure of the GP or "charring"of the fuel.
  • Install an in-line filter between the GP heater and the IP as insurance just in case the GP fails and fragments, or to catch any charred bits. Better the bits in the filter than in the IP :)
  • You could also install another heat exchanger in place of the GP heater. This will work very well, but it will not have the advantage of the GP heater that will heat the oil to allows quicker changeovers.
  • As further insurance, you do have a serviceable fire extinguisher in your car don't you? They come standard in all Mercedes models and are essential equipment in any vehicle.

2. The other modification of the original installation was to send the IP overflow and IP returned fuel back to the veg tank. Previously this was looped back into the fuel circuit just before the heat exchanger. This task is managed by the 6-port Pollak fuel selector valve which has performed flawlessly. The returned fuel is still quite warm and sending it back to the veg tank warms the fuel and has cured the problem mentioned in my previous post where the high melting point (HMP) fats were clogging the fuel line fittings on the marine tanks. The relatively small size of the tanks (25litres) means that they warm up quite readily. I carry two of these tanks at all times, with the second one ready to be connected when the other runs dry.

I also added a small Facet lift pump and an in-line filter before the flat plate heat exchanger. See diagram.

In my next post, I'll upload a diagram that shows how to wire the glow plug heater.

Issues that needed to be resolved:

Twelve months on and the 300D is running fantastically well on unmodified vegetable oil (UVO). Actually, it is used cottonseed oil from a local restaurant, which is merely filtered to 5 micron before adding it to the marine fuel tank in the boot. To those of us who run our cars on vegetable oil, these vegetable oils are commonly known as

SVO – straight vegetable oil
UCO - used cooking oil
WVO – waste vegetable oil
UVO – used vegetable oil or unmodified vegetable oil

Note that vegetable oil is NOT biodiesel. Often times the media and others do confuse the two when discussing fuels. Biodiesel is made from either or both of vegetable and/or animal fats/oils through a process of transesterification, which removes the glycerine or “soap” component. Biodiesel is in fact a by-product of soap making!

But I digress. My conversion hasn’t been all plain sailing as I have had the following problems:-

1. High Melting Point fats. I found that the HMP components of the cottonseed oil were clogging the hose fittings on the tank in cold weather. It took me a while to find the cause of this as at first I blamed the CAV filter cartridges. These blockages caused fuel starvation symptoms, like a huge loss of power, so that I was forced to switch back to dino-diesel which is my start up fuel. Remember, I have 2-tank system, with my start-up fuel being held in the 300D’s stock tank and my veg oil in the 25 litre plastic marine fuel tanks.

These blockages were occurring despite the fact that I “cold-filter” all my oil, which gets rid of most of the HMP fats. Temperatures here on the east coast of Australia are such that this problem is non-existent in summer time.

2. Insufficient fuel temperature before the injection pump (IP). This was a relatively minor issue, as the car ran just fine anyway. Even though the 26-plate heat exchanger is so efficient at transferring heat from the coolant to the veg oil, I found that a lot of heat was lost through the hoses, fuel filter (CAV), and the lift pump. This heat loss was some 40% of the heat supplied by the flat plate heat exchanger (FPHE). Insulating these components did help but something needed to be done to boost the temperature immediately before the IP.

The reason I wanted to correct this, was twofold. Firstly, I needed to lower the viscosity to that approximating dino-diesel to lessen the workload of the IP. Even though the Bosch injection pump on these Mercedes 300ds is by reputation such a rugged unit, there is no sense in punishing it unneccesarily. Secondly I needed to shorten my changeover times. This is the length of time required after start-up before the switch to veg oil can be made, and is determined by how quickly I can get the oil heated to a high enough temperature. The benefits are clear, i.e. to lower the consumption of dino-diesel as much as possible!

I will "tell and show" how I resolved these issues in my next post!

Thursday, 6 July 2006

Fuel system diagram

A few people have asked me to publish a diagram of the fuel system used. Well, here it is. My apologies for the rather rudimentary sketch. I'll try to improve it one of these days.

There are a number of possibilities available for a twin tank conversion such as this, and there is no such thing a the "perfect system". There is always a compromise of some sort. I’ve found that my system works very well. Initially, I was concerned that this particular system would be prone to air intolerance, but as it turned out, it is very good in this respect. In fact there are NO air leaks at all, and even after the car has been left standing for a few days, the system remains perfectly primed.

Initially, this was not the case, and I was going crazy trying to eliminate the air when bleeding the lines. I refitted all the fuel lines and retightened the hose clips, so I was satisfied that the air leak was not from poor attention to the plumbing work. The source of the air leak turned out to be the design of the 3-way return line banjo fitting on top of the stock Mercedes fuel filter. The fitting has a hole through the middle of the screw so that air was entering from the canister. It was easily fixed using some metal epoxy to block the hole. Once that was done, priming using the stock priming pump on the side of the IP was a cinch!

Note that the spill line for the diesel circuit goes back to the diesel tank, whereas the the spill line for the UCO/WVO circuit re-enters the UVO/WVO line just before the heat exchanger. This helps to conserve a small amount of heat.

A separate filter for each fuel system ensures that changeover times are minimised. A one-filter system would require that the entire contents of the filter would need to be emptied before effective changeover occurred.

No electric heaters are employed at this stage. I wanted to avoid placing additional load on the alternator, as a decent electric heater would draw something like 200-300watt.

You will also notice that I have not used a separate fuel pump in the WVO system. I’ve found that the stock lift pump is quite capable of the task.

One disadvantage of my system design is that the vehicle must be started and shut down on diesel fuel. There is always the possibility that the driver could forget to switch over before stopping for extended periods, thus leaving UCO/WVO in the IP and injectors. While I've found that this is not really a problem and re-starting from cold is still possible on WVO, albeit a little more cranking is involved, this is not really ideal because it can cause coking of the combustion chambers.

Overall, I’m quite happy with it. I’ve found that the 300D seems to reach operating temperature quite quickly, and that I am able to switch over to WVO fuel after about 3 minutes. Change overs are not noticeable at all…

Tuesday, 4 July 2006

It all comes together... Conversion complete!

The final stage of the veg-oil conversion of the 300D has been completed! This stage was almost all plumbing work. It now runs on

  • Diesel,
  • any new vegetable oil,
  • any used vegetable oil,
  • biodiesel or
  • any combinations of the above!

The veg fuels are filtered to at least 5 micron before filling the tanks and then again by a pre-filter and an on-board water-removing CAV cartridge type filter.

In operation, I start the car on diesel. It takes 3 minutes for the 3litre Mercedes OM617 diesel motor to heat up the coolant suffiently to enable burning of the cooking oil to commence. I flick the on dash fuel selector switch and about 15 seconds later I’m driving along smelling like a mobile Maccas! When I stop and turn off the engine, if I’m not intending to restart within a couple of hours I’ll shut down on diesel to make starting easier. However, it will start OK on the cooking oil, but to avoid any damage to the injectors, a proper shut-down is preferable. I’m thinking of the possibility of using a turbo-timer to automate the shut-down.

Here are the final shots of the installation:

This one shows the brass tee into the heater circuit to take off coolant for the heat exchanger. I’ve discovered that quality brass fittings are expensive! This tee routs the coolant to the 26 plate heat exchanger, from where it returns it to the cooling system via another tee before the thermostat on the other side of the engine. As you can see, I can still access the oil filter quite easily, so servicing is not a problem.

This shot shows the brazed plate heat exchanger (imported from Taiwan via the US) with all plumbing connected. The coolant enters from the top and exits at the bottom of the exchanger. Fuel (waste veg oil or “wvo”) enters at the bottom and exits at the top. This is called a contra-flow setup, and makes the most efficient use of the heat from the coolant. I found that it was not necessary to move the windscreen washer tank or the self-levelling suspension system’s hydraulic oil tank, as all the hoses fitted comfortably around and under the tanks.

From the heat exchanger, the heated WVO then goes through the Delphi CAV filter and from there to the Pollak valve (US made) for distribution to the lift pump on the IP. The Pollak valve is controlled from a switch on the dash.

This shot shows the connection from the lift pump to the injection pump, or IP. The line is transparent to allow me to see if there are any air leaks in the system. Air leaks will immediately be revealed by bubbles appearing in the line. Normally this supply line goes to the stock Mercedes filter before being routed to the IP, but the fuel is now filtered before it enters the lift pump, so this is no longer necessary. The glass pre-filter on the diesel supply line is just to the left of the transparent line. This transparent pre-filter is important as it also helps to detect air in the system. Additionally, it is important as it monitors any algae (“diesel bug”) build-up in the tanks which is a natural occurrence in oil fuels, but of course must be eliminated.

It was convenient to utilise the stock three-way return-line banjo fitting on the stock filter housing to route the overflow lines back to the Pollak valve. However, to eliminate air and facilitate priming this particular set-up required that the venturi hole in the three-way fitting had to be blocked off. I did this using metal epoxy. The rear of the CAV truck-style filter for the WVO fuel circuit is in the foreground. I have been asked why I have the two filters, and not just the one? After all the Delphi does have 4 fuel ports. The simple reason is that fuel changeovers are more rapid this way, as the fuel in the filter does not have to be used up before full changeover occurs.

The WVO tank is in the boot. I have two of them for easy changeovers and handling. Honda bayonet fittings ensure easy, no-mess changing of the tanks. They are restrained by “occie” straps and a bar that slips into a recess under the tank. In due course, I may decide to use these outboard tanks for the diesel fuel and put the WVO in the main tank. At the moment, it has all to do with the ease of fuel handling, and this system allows me to refuel the tanks out of the car. It also allows me to use a brew-heater under the tank overnight in mid-winter to prevent the possibility that cold temperatures could gel the WVO.

Sunday, 4 June 2006

Pollak valve wired up and control switch installed...

I didn't get around to the plumbing this w/e but I've installed the wiring for the Pollak valve and the On/On control switch is mounted on the dash. To install the switch, I drilled through one of the switch panel blanks.

I've taken the power for the Pollak from the rear window demister as it operates only on auxiliary.

I encountered a problem with the valve operation. The problem was that I only had +12V to one terminal on the Pollak valve. Flicking the switch resulted in no voltage to the other terminal, so the solenoid did not operate back and forth as it should.

Thanks to Tony in West Oz, the problem was identified. I had omitted to bridge the diagonal terminals on the 2-Position, Double Pole Double Throw, On/On switch :( What a klutz I am!