The Waste Oil Furnace

~ Or ~

Yet another Web page With Flames!

Safety First, Safety Last, Safety Always

I'm an idiot, ask anyone who knows me. I did this, if you do it I guess that makes you an idiot too. Do you really want to trust the lives of your loved ones to something done by an idiot?? Or explain to your insurance adjuster why your home is no longer there??? Idiot that I am even I wouldn't do this to heat anything other than a (very) detached garage or workshop

Also, Consider for a minute- have you ever been seized by the urge to plunge your face into a deep-fryer? O.K., howzzabout when its operating? These modifications involve dealing with flammable products heated to high temperatures and pumped all over the place at high pressure. If you do something wrong you're gonna wish you had stuck with the deep fryer 'cause that would only have involved your face. If you are not mechanically inclined with a good understanding of electricity, hydraulics, metal working & most importantly *Pain* you are better off not attempting to duplicate this project.

Legal Disclaimer- You Are Being Warned.

This page does not constitute a set of plans or a "How To" manual. It is more of a recounting of my learning experiences along the way to (Spoiler!) successfully complete my objectives. If you attempt any modifications to a heating appliance based on what you read here I will not be held accountable for your screwups.

So There.

p.s.- Please mention me in your will. :)


When Rachel & Tonka had an accident I kicked dad out of his garage & moved the leftovers in. Dad wasn't too happy about it so we reached a compromise, I took Rachel out of the garage & put her in the hospital.

That seemed to make him feel better :).

So for the last few years we've been spending a lot of time doing repair work on Tonk in my I mean dad's garage.
Now dad is one cold blooded SoB, grew up in the Dirty 30s on a farm, getting up at Oh-Dark-Hundred to milk & feed the cows in a barn where the only heat was from steaming fresh cow poop (yuck). Then walking through 5 miles of three feet deep snow up-hill (both ways!) to get to school where all the budget would stand was one lump of coal for the stove.

Per year.

Then back home working in the woodlot cutting down trees for firewood using a two man cross cut saw. Of course since he was by himself the saw didn't work too well so there wasn't much firewood :(. And by the time he was finished the horse was usually frozen solid so he'd hafta pull the sledge (horse loaded on top) back home through the 3' deep snow by himself.

So when dad built the garage you can probably figure out that central heating was not a major consideration for him. Just a little Itty Bitty wood stove over in the corner, just enough heat to get the temp a few degrees far enough above absolute zero so the floor wasn't slippery with Helium II & that was fine by him.

I, on the other hand, find anything below 80 to be downright chilly.

The first summer of the rebuild wasn't too big a problem since most of the work was happening outside anyhow but when winter rolled around it became apparent that me & the wood stove weren't gonna ever be Buddies. It was a pain to get going. It was a pain to keep going, half the wood you threw in there It didn't like. Once It was going It's appetite was enormous, ignore It for 15 seconds & It went out. When It got tired of playing games & settled down It took an hour to warm anything up & the only things It would warmup were items with a straight line of site of It within ten feet or so.
Or the roof.
Since I wasn't working on the roof that wasn't much help. :(
If you were working (trying to work actually) anywhere's else in the building you were always in danger of ending up like the brass monkey that didn't get brought in :(.

So, after the first winter it became apparent that There was Going to be a Change.

The Plan

Now some people say the greatest invention of all time was the printing press.

Others say it' electricity.

Many will say it's the spoken word (although many others consider that to be more a developmental thing).

Let me state unequivocally right now that the greatest invention of all time is the Thermostat.

These things are pure Magic- If you're cold all you have to do is twist a little knob & you're warm :).

God I Luv'Em.

I toyed with a few different ideas such as a propane heater but somewhere's along the way I got the idea to burn waste engine oil, which was odd because any time I get an idea its usually accompanied by the smell of burning wood. :\ Be that as it may I stated looking into what was needed by researching the subject on the web, after all, we all know if its on the web its got to be right, Right :? I found that there seems to be three approaches to getting heat out of waste oil-

The first is the so-called drip burner as epitomized in its basic design by the Mother Earth News design and other similar designs. Easy to make & operate but it has many of the the same drawbacks as the wood stove,

Then there's the Babbington Burner. These are pretty darn neat but there's more to heating a garage than building a pretty darn neat burner. Somehow you have to convert the hot gases to usable heat & then distribute the heat to where you want it. That would call for the fabrication of some sort of heat exchanger & other bitties.
Sounded like way too much work to me.

Third is to burn the waste oil in a modified oil burner, this is the way commercial units do it. There's also allusions to individuals having accomplished this but details on exactly how either the commercial or home spun conversions work are very thin. :(
But, & heres the important part, Oil Burners are very Thermostat friendly :).

Well, if others can do it I figured I could do it to :). Obviously I needed a regular oil furnace (With Thermostat :)) to start with.
Due to the layout in the garage I decided to look for a ceiling mount unit. Due to $$$ considerations a new one was out of the questions so I started watching the local classifieds & in no time flat, about six months, I found one. Not only that but it came with a bunch of other stuff as well, a couple of spare blowers, a tank, a transfer pump and a *Thermostat*. :)

I actually ran this thing for two years before achieving success with the waste oil, first year I spent buying diesel fuel 2 1/2 gallons at a time to feed it, the second year I got smarter & loaded it with 50 gallon drums of #2 fuel oil. Neither one of these options turned out to be cost effective :(.

So, now I've got the furnace & associated miscellaneous bitties installed & working, how do I achieve my goal of burning nice free waste oil instead of the not nearly so free #2? First thing I did was buy The Book

I found this in a used book store, its "age appropriate" for the furnace which is A Good Thing. This book was invaluable, it details the differences between fuels, workings of pumps, regulators, controls & just so much other stuff I needed to know :). I really wouldn't have gotten anywhere without it. I also found this on the web, some good general info for the hot-air challenged. :) There's some excellent info on oil burner nozzles from an online document put out by a company called Delavan. This document goes into detail on how the nozzle works & how how viscosity & pressure affect the combustion of the oil.

As it turns out, No. 2 furnace oil & waste oil are two very different things, something designed to burn one isn't going to be happy trying to burn the other :(. The main differences that I could find appear to be VISCOSITY & FLASH POINT. Flash point basicly refers to the temperature at which something will ignite, for my purposes it equates to how easy it is to get the oil to burn. Viscosity gets a bit complicated but for this project it equates to how thick, or in technical terms, "Gooey" a liquid is. At room temperature #2 has a viscosity of around 10 Centistokes or 50 Saybolt Universal Seconds. I'm not going to get into how the different units are defined here, look'em up on the net if you really have to know For comparison water has a viscosity of 1 cs or 30 SUS. Waste oil, being mainly comprised of used engine crankcase oil should be in the ball park of 60-70 cs or 300-400 SUS at room temperature, which is apparently a lot thicker than the # 2. If I'm going to have any chance of pulling this off cheaply & easily I'll have to get the waste oil's viscosity to match that of the furnace oil. I may also have to do something about the flash point to get it to burn once its successfully sprayed into the firebox.

Note- I have corresponded with someone who claims success with burning waste oil in an un-modified burner by upping the size of the nozzle & starting it out on kerosene until it warms up. It was actually the first thing I tried & it didn't work at all for me. :( YMMV.

So, I sit & stink, um, I mean think about it for a while & I can only can think of two ways to thin oil:

a) Thin it down with thinner oil or a solvent


b) Heat it up 'till it gets to the right viscosity.

Solution A has the advantage of being able to use the burner "as is" with no add-ons or modifications. The disadvantages are that I'd have to be continually experimenting with a dash of this & a pinch of that every time I topped up the oil tank in order to get the right mixture and that, as the room temperature could be anything from -30 to + 70 as the days wears on, the viscosity would be changing as well.

Solution B has the advantages of having to do research, learn things, design a heater system, modify the burner, design control systems, build, experiment, spend money and waste great amounts of time to make it work. There are no disadvantages I could see so thats the route I went with. :>.

Furnace Basics

Well, before I could start making modifications with any reasonable chance of success I had to understand how an oil fired furnace works, this is what I found out. It's not going to be an exhaustive description of how a furnace works, for that you'll need something such as the previously mentioned Audel's book. There's a lot of little "gotcha" details in there that I'm not going to mention here at all. Be that as it may heres the broad picture- There's the oil burner, the fire box. the heat exchanger, circulating fan and controls.
Well, o.k., if you really insist I'll go into a bit more detail. Sheesh.

The Oil Burner.

It contains the oil pump, the combustion air blower- both run by a common motor, ignition system, burner tube, an air swirly doohickey, the spray nozzle and a flame detector.

The oil pump is actually a bit more than that, it consists of a housing incorporating the actual pump, a pressure regulator, a noise supressor & a filter. The housing has enough volume to hold a few ounces of fuel in it. I found that out the hard way the first time I spun a burner up to see it the motor was good, I really wasn't planning on 3 feet of flame shooting out the end of it :(. The regulator will usually be set for an output pressure of 100 PSI.

The blower will invariably be a squirrelcage type, the inlet to it will have some sort of vane or damper draft control in order to adjust the amount of air introduced to the fire box to achieve the best air/fuel ratio for combustion. The blower is usually mounted directly on the motor's shaft, there will be a flexible coupling connecting it to the pump.

The burner tube conducts the air into the firebox. At the business end there's a doohickey to swirl the air, this works with the nozzle's spray pattern to burn the oil properly. Older burners have a doohickey that looks like a little propeller, newer high efficiency "Flame Retention" heads have ones that look more like a sexy jet turbine. In either case the darn things don't actually turn at all. ;/ Running down the center of the tube is the feed pipe that takes the oil from the pump to the spray nozzle. There's also electrodes running through it from the ignition transformer (10,000 volts +/-) & set to make an arc near the nozzle.

The nozzle, when the viscosity & pressure are correct, will spray the oil in the proper pattern, amount & droplet size to mix with the air in a pattern that maximizes fuel burn. It does this buy forcing the oil thru teensy-tiny angled slots then out thru a teensy tiny hole.

At the back of the burner assembly hidden away from outside light is the flame detector, it's positioned to have a view of the flame when it's going. there's a number of different devices that can be used as a flame detector, the most common seems to be a plain old-fashioned it was good e'nuff fer grandpa & it's by dang good e'nuff fer us cadmium-sulfide photocell. When it's illuminated it has a resistance in the 10-100 ohm range, when there's no fire it's resistance will be 10-100 thousand ohms.

The Burner Controller

The burner controller takes control inputs from the thermostat & flame detector to control the burner motor & ignition. Here's the schematic of the rather ancient Honeywell unit that came with my furnace-

Now, Isn't that Simple??

It's Not???

Ok, heres whats going on-
115 volts come in on terminals 1 & 2, this feeds a 24 volt transformer & the control relays for the burner motor & ignition transformer. The 24 volt transformer supplies power to the timer module, the thermostat & the flame detector. When the thermostat contacts close the power is supplied to the relay windings (Km & Ki) & the timer is activated. After about 2 minutes the timer shuts off power to the ignition to save wear & tear on the electrode. Now, all the burners I've looked at have the motor & ignition wired up together so as long as the motor runs the transformer is making spark. In fact the newer style controller I ended up using doesn't even have a separate output for the ignition.

So There.

There's a thermaly operated timer wired into all this (TT). Essentially the flame detector, through relay Kfd, supplies power to this timer's element. When the flame detector sees flame it's resistance drops, Kfd operates & TT is disabled. If the flame detector doesn't see light after about 30 seconds TT's element heats up enough to trip a bi-metallic strip, opening its contacts & breaking the power to the thermostat. This causes the furnace to shut off.

The timer module has a few other tricks up it's sleeve besides shutting off the ignition. This particular unit, should there be a shut down due to no flame, will attempt a restart after 2 minutes. It also will shut the furnace down for a about five minutes after an hours running time, I'm not really sure why as this doesn't get mentioned in anything I've read on these. :/

The Fire Box or Combustion Chamber.

It provides a stable environment for the flame to burn in. Oil & air come in, The Magic Happens, really hot gasses go out.


The Heat Exchanger.

This is a box thats divided into two airtight sections. The hot gasses from the firebox come into the first section, go through ducting that passes through the second section and then pass out of the exchanger to the flue pipe & chimney. In the second section room air is introduced by a large fan, passes over the conduits that the exhaust gases are passing through picking up heat from them then exits out into the room.

Sorry, no neat pics or diagrams for these, they're steel boxes & thats about it.

Fan Controller

The circulating fan is controlled by a thermostatic switch mounted near the exchanger. (Not to be confused with the room thermostat) When the heat exchanger reaches a set temperature the controller turns on the fan. This thermostat also supplies power the the burner controller, if the exchanger gets too hot the thermostat cuts power to the burner controller shutting down the furnace.

Thats it. Minus a couple of hundred pages of important details.

Design Criteria

O.K.- I've got the general plan.
I've got the background knowledge.
To get the waste oil to properly spray &, hopefully, ignite I need to change it's viscosity to something approaching that of #2 furnace oil.
I intend to do that by heating it up.

So far we're rockin :)

But, we have some other goals in mind as well-

The process has to be as automatic as possible, I don't want it to be a multi-stage procedure to make this thing go. Turning the knob on the thermostat is about as complicated as I want it to get.

The process has to be reliable, I don't want it to work one day & not the next.

It has to be, above all else, safe. I don't want any fires, I don't want anybody to get hurt. The existing safety features built into the stock controls must remain uncompromised.

Anything I build has got to meet these goals or its no good. So, with the above in mind it's time to get design'n.

Let The Games Begin- The Mk.1 (aka "Calvin")

Here's what we have to work with-

O.K., so I have to heat the oil. This will require, obviously, a heater. I did consider a few different ideas such as piping the oil through the fire box but that would have required starting the furnace on a different fuel, entailing doing things with valves & whatnot whenever starting it. There was also no way to control the resulting oil temperature, it could have come out too cool or dangerously hot, even vapourised. That idea didn't last long. The simplest way seemed to be an electric heater. I also considered heating the combustion air as well, the commercial units do this. I elected to hold off on this until I could see how well the thing worked without it.
So, heat the oil only & do it electrically.

According to my research waste oil should reach the same viscosity as #2 at around the boiling point of water, cooler than that it won't spray properly, warmer than that & we're getting into un-charted territory. In the interests of reliability & safety obviously some sort of thermostatic control of the heater is required. My first thought was to use a heater between the oil pump & the nozzle, located within the burner tube. As I researched a bit more I found that this is the way commercial units seem to do it. I decided I'd go a different route due to the difficulty of designing & building something that could run hot enough to do the job required in the short space available. I was also worried about the effect inserting something into the burner tube might have on the combustion air swirl pattern at the nozzle. So it looks like some kind of heater external to the burner head, the best idea seemed to be a tank to hold the heater & act as a reservoir of heated oil.
Heres the general design I settled on-

The idea is that the heater element warms the oil, that then circulates through the oil pump housing by convection. I could have simply run the output of the pump into a heater & then to the feed pipe but it occurred to me that trying to start at lower temperatures I might run into the situation where the oil would be too thick for the pump to handle. Running the path from the heater to the pump will alleviate this. When the oil reaches the correct temperature a normally open thermostatic switch (90C) (mounted to the tank) that is wired in series with the room thermostat closes, enabling the room thermostat to command the burner controller to start the furnace. Once the pump is running the pump output stream passes through the tank again (via a pipe running through it), this is to make up for any heat lost to the pump body. There's two other thermostatic switches mounted to the tank as well, one rated at 212F/110C to control the heating element & one rated a bit higher than that (120C) to act as a safety in case the heater control switch fails in the "closed" position. Electrically speaking it just don't get much simpler. :)

So thats the design, time to turn it into actual hardware.

The Heater

For the heater element I picked a 50 ohm 50 watt resistor. Now, some genius is about to point out that at 120 volts a 50 ohm resistor dissipates darn near 300 watts & that a 50 watt resister is gonna go *poof*. Well, that 50 watt rating is for operation in still air, sink the thing in oil & the watt rating goes up by about a factor of five.

So There.

I had some 2"x4" steel tubing lying around so thats what I used for the tank. I brazed a number of steel pipe fittings to it for the line ports. The resistor was mounted inside by soldering #14 wire to the lugs, running the wires through the centers of two of the pipe fitting & potting them in high temp epoxy (J-B Weld if you must know). Once it was secure I painted more epoxy over the exposed lugs & wire to insulate them from the oil, not that oil is much of a conductor but I figured "Hey, Why Take Chances?". A mount designed to fit to the end of the oil pump was tacked onto the side. The mount was also intended to provide additional heat transfer from the tank to the pump body. The ends get welded on & she's done.

The tank also got the beauty treatment, some nifty black paint & shiny copper tubing c/w shiny brass fittings just really brought the whole thing together. When something looks this good who cares if it works?

The Hot Box

Now, it was pretty obvious that there was going to be quite a bit of plumbing interconnections going on & I worried about the pressurised sections springing a leak & spraying boiling hot oil all over the place.
And me. :(
I was also concerned about keeping heat loss from the tank & pump to a minimum. The solution to both these concerns seemed to be enclosing the whole thing in an insulated box. Thankfully since I have a sheet metal brake & welder I was able to whip up an appropriate enclosure.

I insulated it with cork faced with aluminum foil & more importantly made it look purdy with gray hammertone paint-

Time to put it all together & see if she'll make some smoke-

Test Procedure-

Now, I didn't just try to fire it up & see what happened. I figured in view of the potential for pain a more structured approach was called for so I came up with this

1- Leak testing

1.a heater tank- bubble test
1.b plumbing - all connections tight
1.c pressure- leaks
1.d pressure- adjust

2- Heater pre test

2.a heater- resistance
2.b heater- ground isolation
2.c thermostat verification

3- Heater test

3.a leak
3.b temp distribution
3.c thermostat operation
3.d venting

4- Burner test

Well Boys & Girls, this is where things started to Not Go According To Plan. :(

The tank flunked the bubble test, i.e. sticking it in a bucket of water & pressuring it resulted in, you guessed it, Bubbles.
Someday I gotta learn to weld better.
After much grinding, welding,grinding,welding, grinding, welding ad nauseum I finally gave up & coated to innards with fuel tank sealer. No more bubbles./ :) As for the plumbing all my flares seemed to work so at least there's something I can do well. :/

For initial tests I set the pump output pressure to 100 PSI.

The heater pre-tests went well at least, got my 50 ohms one terminal to the next & wide open from the terminals to ground. The temperature switches work more or less as advertised, the on/off ranges were a bit wider than spec but I figured I could live with them

O.K., lets heat up some oil!

Now, as you may recall the design idea was that the oil in the tank would heat up, circulate via convection through the pump casing warming it as well. When the oil hit the right temperature(s) the temperature switches would operate, enabling the the room thermostat & controlling the heater. In a fit of genius I had decided to move the switches from the tank itself to the mount holding it to the pump, on the theory thats where they would get the truest picture of the temperature.

The test Did Not go well.

After a few minutes it became apparent that the top of the tank was way hotter than the bottom, where the switches were mounted was barely warm & the pump wasn't warming up at all :(.

Apparently steel does not conduct heat anywhere's nearly as well as I thought it did :(. After a bit more research on the web it turns out steel only conducts heat about 1/4 as well as aluminum or 1/8 as well as copper. Obviously by the time the switches operated my oil was going to be way hotter than I intended which was going to take me into unknown territory, from fuel to feedstock. :( I elected to try moving the switches to the top of the tank (position # 3 if you're keeping track) & not worry about the pump warming up, reasoning that after all the oil would pass back through the heater tank on the way to the nozzle, it would certainly pick up enough heat that way to spray properly.

Have you noticed that I'm starting to drift into rationalizing here??

Well, I fashioned a clip & moved the heater control & overtemp switches to the top of the tank Another test & things seem to be finally working the way I want, the oil temp is hovering around 100C, the thermostat circuit switch clicks on at more-or-less the right temp.

Finally, after a years work & zillions of dollars invested (about a hundred bucks actually) its Time!

And heres what happened!

Now, when I realized that the pump wasn't heating up that I'd be pumping not-quite-hot oil for the first few seconds or so & that this would negatively affect the spray pattern.

Boy, was I right!


About something :(

I let that sucker run for about five minutes & all I had to show for it was a big oily spot five feet in front of the burner. A complete & utter failure. :(

So, What Went Wrong?

In no particular order heres where I screwed up-

I miss-judged (actually it was more of a wild-assed assumption) the thermal conductivity of steel in transferring heat from the tank to the pump body. Or from the top of the tank to the bottom for that matter.

The oil did not circulate through from the heater to the pump the way I thought it would. I'm still not sure what went wrong here, it may have something to do with the copper pipe size I used to connect them together, larger pipe might have worked but I was constrained by the radius of the curves I needed to bend it into & the 1/8" pipe thread holes the pump body uses.

The oil was just not staying warm enough to spray on its way to the nozzle. Either the path to the the nozzle was too long, there was too much heat loss, or (most likely) just not enough volume going out the nozzle, meaning the oil was staying in the pipe too long to stay warm.

The temp switches didn't work out the way I wanted, or even as advertised in the catalog. Their on/off range was, right out of the box, higher than spec'd. The nominal operating temps were off by up to ten degrees C, in fact the "overtemp" switch was opening up before the actual heater switch was. After running through the testing & re-configuration they were even further out of spec, one actually failed completely.

So, What Went Right?

It looked really, really Cool :)


The mk.1, so recently re-christened "Calvin" is a bust. I can't go any further with this design, time to take a break & figure out what to do. I'm still committed (decision, not visit to Funny Farm) to (not for) burning waste oil so a new design is in order, one that addresses to shortcomings of Calvin.

Let The Games Continue- The Mk.2

Chapter 4- A New Beginning...

To re-address the shortcomings of Calvin-

It Didn't Work

Now, what do we do to avoid these shortcoming on the Mk. 2?

Most importantly, I've got to get the oil hot & keep it hot all the way trough the system right to the nozzle.
Calvin's circulation scheme was a bust, the convection didn't work & the steel tank didn't work. After all the trouble I had sealing Calvin's tank I'm in a mood to try something different anyhow.
I need better temperature control.

After giving it some thought here's what I came up with-

I'll remove the end cap off of the pump & mount the tank directly to the pump. The tank will have a hole in it allowing oil direct access into the casing. It will be fashioned from an aluminum utility box which will transfer the heat around far better than the steel did.
I'll make sure the hot oil is circulated right to the nozzle, in order to do this I'll need a feed pipe that is both a feed line to the nozzle & a return line from the nozzle back to the heater. Pressure will be set by an external regulator installed in the return line, this will require disabling the pump's internal regulator. To insure that the hot oil makes it to the nozzle ASAP, thus producing a proper spray pattern & ignition before the flame detector/burner controller times out, I'll put in a timed bypass circuit.

With the addition of bypass obviously the heater controller is going to be a bit more complicated than Calvin's. I've also taken the pledge and sworn off temperature switches for controlling anything. A little digging around revealed the existence of programmable temperature controllers known as PIDs utilizing thermocouples and having multiple independent outputs.

Here's what the Mk 2 Unassembled looks like-

Once again I'm going with 50 watt power resistor heating elements, but two this time because to new tank has twice the volume of the original, I don't want to wait all day for the oil to heat up to the working temperature. The project box for the tank comes with a neoprene seal in the lid, advertised as "oil tight". The black thing is the PID, the silver wire is the thermocouple that came with it. The beige thing with the black knob is the timer for the bypass. there's also a stand alone temperature switch (using a different technology from Calvin's units) in the white envelope to act as an overtenp shutdown incase the PID fails Big Time. Some relays to handle the high current & a box to stick it all in & thats mostly it.

The Heater

This went just *So Much* easier than the Mk.1 :)
To mount the resistors I drilled the holes & installed compression fit nutserts. Once again I inserted the wires through & potted with high temp epoxy, then again covered all exposed wiring & lugs with more epoxy. Once everything was cured some spade lugs get soldered on & we're done with that.
You can also see in the pics the fittings for the lines, I turned down standard pipe fittings, brazed them to fender washers & bolted/epoxied them to the box.
The tank gets bolted to the pump housing but first I removed the filter screen that normally resides in the pump housing, placing it in the tank will allow for easier cleaning if required in the future.
The cover (note the gasket) gets screwed on & we're done. :)

The Feed Pipe

The original was simply a piece of 1/8" pipe- the line from the oil pump hooked to one end, the nozzle to the other. Since I now need to provide a return path for heated oil this isn't going to work. I considered a dual pipe but this solution seemed easier to implement, it's a 9" length of 1/4" pipe with a piece of 1/8" tubing down the middle. At the back end the copper tube is inserted into a street elbow, this gets screwed into the 1/4" tee. Oil goes into the elbow, down the copper tube to the end, whatever doesn't go out the nozzle comes back up the 1/4" & out the side port on the tee fitting.
Obviously the new pipe is a bit clunkier than the old thus requiring some mods to the burner case for it to fit.

To connect the feed pipe to the oil pump & regulator assembly I needed some flexible hoses, both due to the limited space & so I could move the feed pipe to adjust the nozzle position within the burner tube. I forked out some $$$ to get stainless steel braided lines made up. These are definite overkill for the 100-150 PSI the system will see but I really like the idea of the reinforcement in case the heat or compounds in the fuel ever causes the hose rubber to deteriorate.
Most importantly- They look Way Cool. :)

The Regulator

Here's the regulator assembly which includes the solenoid activated bypass valve. Oil returning from the feed pipe comes into the tee, one side leads to the regulator, the other to the bypass valve. When the valve is open the oil has a free path back to the heater tank. When it closes the regulator maintains the pressure in the feed tube that the nozzle requires to form the proper spray pattern. The regulator is an adjustable unit so I can "fine tune" the system. As this mounts in the hot box I was also a little concerned about the knob melting, as it turned out that was the least of my problems but we'll get to that later on.:/
I did make sure the valve was spec'd to handle the heat.

The Controller

Here's the diagram of how it goes together & interfaces with the burner control. The sequence of operation goes like this-
When power from the fan controller is applied the PID initializes. The PID has two outputs with independent set points.
The first operates relay K1 applying power to the heater until such time as the thermocouple reports the oil temp has reached the required temperature. Power to K1 is routed through a "Low Fuel" switch which resides on a day tank. In the event the fuel supply stops, when the tank is near empty the switch opens in order to prevent the power resistors in the heater from operating when not immersed in oil. In the event of a failure of the PID or thermocouple there is a separate "overtemp" shutoff circuit consisting of a temperature switch (mounted to the heater tank) which operates relay K2, interupting the PID's control of K1. Associated idiot lights signal when the heater is in operation or that the low fuel or overtemp conditions are present.

When the thermocouple reports the second setpoint (set to a temperature *below* the 1st setpoint) the PID's secondary output contacts close, enabling the room thermostat circuit to call for heat.

When the burner controller receives a call for heat from the thermostat it applies power to the "M" output lead activating the burner motor & ignition transformer. This also triggers the timer which activates the bypass valve allowing heated oil to be freely pumped through the feed pipe. At the end of the timer period the valve closes, effectively inserting the regulator into the oil's circuit, at which time the nozzle starts spraying &, hopefully, ignition occurs.

The timer also disables the flame detector during this period by applying a short across the leads to the photocell. This is the *only* liberty I took with the furnace's stock safety systems and it's not something I did lightly, I spent a lot of time contemplating what unintended effect(s) this could and as far as I can determine this modification has no likely dangerous consequences. The only possible dangerous failure mode would be the timer contacts failing in the "closed" position, disabling the flame detector during normal operation. If this should happen there would be the possibility of the firebox becoming saturated with un-burnt oil. My take on the situation is that the timer is a CSA/UL approved unit designed for use in commercial/production environments, the likelyhood of it failing in a way to compromise the flame detector is no more likely than the flame detector circuit components themselves causing a similar problem.

So There.

So much for the theory, lets put'er together.

Construction is plain ol'fashioned point to point wiring. For wiring going into the hot box I used "Range wire" which has insulation rated to 200C/450F.
Once everything was together we apply some power & it lights up, which Is A Good Thing. What is actually showing here is that the first setpoint (heater control) is set for 111C, the thermocouple is reporting a temperature of 16C. As 16 is lower than 111 the power for the heater is enabled(green idiot light).

The thermocouple as supplied with the PID is not a liquid-tight unit. This necessitated fabricating a sealed casing that could be screwed into a spare port on the pump body, the thermocouple then screws into back of the casing. It's made out of a 1/8" pipe plug drilled out, the outer end is tapped to accept the thermocouple, the other end has a short length of copper pipe solder in with the end sealed up. I elected to mount the thermocouple in the pump body for two reasons- first there was a spare port available saving me the trouble of adding another port to the tank and second I wanted to be sure the oil in the pump is at the right temperature when a heating call is issued.

Test Procedure

Where with the Mk.1 I just sat the thing on an old oil drum with the Mk. 2 I got a little fancier & built a test stand/

As with the Mk.1 I followed a test procedure-

1- Leak testing

1.a plumbing - all connections tight
1.b pressure- leaks
1.c pressure- adjust

2- Heater pre-test

2.a heater- resistance
2.b heater- ground isolation

3 Controller

3.a PID operation- thermocouple calibration
3.b PID operation- set point operation
3.c Over temp & low fuel shutdown verification
3.d Timer operation- bypass activation & flame detector disable

4- Heater test

5. Burn test

First the leak test. Different from the Mk.1 on the Mk. 2 the heater tank isn't pressurized & as the box came with a neoprene gasket on the cover, ah hell, just look at the damned picture :(.

Those ain't tears of joy dripping off the bottom of the tank there folks. Off comes the cover to reveal the ^$%* gasket in pieces. I really have no idea what happened to it but it went back on with a goodly quantity of high temp silicone around the edge. Tried it again & no more drips. :)

O.K. with that behind us we can move on.

A little juice applied to the motor & we can set the pressure. First I plugged the pressure guage into the pump output & cranked the internal regulator in the pump casing all the way up to take it out of the picture so to speak. It topped out about 150 PSI, I figure this will make a nice pressure relief feature if the external reg ever seizes up. In order to set the external regulator off comes the guage & the output is routed back to the regulator assembly input to bypass the feed pipe & nozzle- I don't want oil spraying out there quite yet.

Needless to say there's a major screwup in the works :(.

I fired up the motor again & tried to set the regulator. For something with a knob with that many turns it seemed to have a remarkably narrow choice of settings. :( One half of the range produced a reading of 150 PSI. The other half produced a reading of 0 PSI. Right in the middle was a spot where it would just go nuts.

As it turns out the regulator I bought operated ass-backwards from what I needed, it was designed to maintain a set pressure on the *downstream* side of the oil circuit, NOT the upstream side.


I ended up rebuilding it by combining some original parts with a new valve seat & valve stem that faced the right direction.

When I took it apart I found the spring was broken anyhow. :/ I made the new valve seat out of a bolt head & used a rubber faucet washer for the disc.

Lets try it again-

Ahh, Thats better :).

After that little setback it got rather anticlimactic. The resistors read exactly what they were supposed to & there were no shorts to ground.
Next step, lets try the controller.

The PID worked as advertised heating the oil to the set temperature, the timer activated the bypass valve & flame detector cheat for the set amount of time.

So much for the bench tests, its time to go for the gusto.

We're moving outside People, please stay together in a group & don't touch anything.

Here's the setup. To test the Mk.1 I sat the burner on an oil tank in the garage & aimed the business end out the back door. This time around I decided to get a little bit more elaborate.

And Safe.

On the off chance that I actually might succeed I decided to maximize my chances by fashioning a fire box out of some surplus air duct & added a short chimney. This would give me a shooting chance at establishing a proper draft for the test burns. For the first test I just wanted to verify that everything was still tight & more or less capable of proper operation so I disabled the heater & ran regular #2 furnace oil

And we actually got a flame!

Now if we can get it to work with waste oil-

Well, not waste oil. I didn't want to guck up the guts in case It had to come apart again for some reason so I some used fresh 10W-30.

Remember, at this stage I really had no idea if this was going to work. The viscosity was only one part of the equation, there was also the issue of a far higher flash point to worry about. There was always the possibility that I may have missed something important, I was digging up info on this one piece at a time from different sources, there was no guarantee I hadn't missed something vital along the way. But, I had to give it a try sooner or later so lets give it a shot- The heater gets re-connected, the oil goes in, the power gets connected. We watch the display on the PID as the oil heats & at the right spot on comes the power to the burner. The bypass idiot light comes on for 30 seconds and


Before I even had a chance to look WE HAD FLAME!
It worked PERFECTLY!!


The flame isn't quite as "crisp" and the pattern not quite as well defined as obtained from burning #2 oil but what the hell, its making heat :).
Now since we have a nice test bed going here & its the middle of summertime anyhow lets play a bit. :) Because I went with the PID on this version instead of fixed temperature switches its a simple matter to adjust the oil temperature. I had initially set the heater temp to 100C, with the actual burner activation temp @ 90C. It worked but playing around I found it worked a lot better at 130C, with a flame virtually indistinguishable from burning #2. However, since some of the info I turned up on the net indicated that at this temperature the oil could start changing chemically I elected to drop back to 120C with the burner activating at 110C. Actually as the thermocouple is in the pump body when the pre-heat cycle starts the pump sucks in hotter oil from the tank, about the time the pre heat cuts off & we start spraying the indicated temp has risen to about 120. It's not quite as good as the 130 setting but its *that* close. I also tried tweaking the pressure & found that anywhere's from around 80 to 120 PSI worked well. It would actually run as low as 50 & over 120 the pattern became unstable. I settled on 100 which is the book value anyhow.
I gotta say this is a big improvement on Calvin, I ran it through cycles for about five hours total & absolutely no hicups.


Time to put this puppy to work. Well, no its not, its still August fer cripe's sake.

Pre/Post Inspection

Now before I was ready, or willing to mount this thing to the furnace I tore it down to see how everything looked inside. The only thing that caught my attention was the feed pipe- the copper tube showed a lot of disscolouration. Even after some cleaning it looked like hell & I discovered the end was actually wearing away. In the picture its showing .023" wall thickness, it was a lot closer to .035" when it went in there. I'm still not sure what caused this, I was burning fresh oil so there shouldn't have been anything corrosive in it. One possibility for the wear is mechanical, the oil exiting from the tube & then reversing direction as it (mostly) flowed back up the outer return pipe may have set up a mechanical vibration. This isn't that big a setback because before I manhandle this puppy up the ladder to the furnace there's some changes that have to be made anyhow, & one of them is the feed pipe. The burner assembly I've been working with has a shorter burner tube than the unit the furnace came with. I'll have to swap the longer tube to the Mk. 2 and fabricate another, longer feed pipe to go with it.

On this one I'll use a steel tube instead of the copper & I'll pinch the end shut with two drilled holes for the oil to exit from.

The burner's all back together & ready to go up.
Notice I've traded in the museum piece Honeywell burner controlled for a museum piece White burner controller. This thing is so ancient it has a real honest-to-gosh transister in it! Don't see many of them anymore. :/
Also note I've added a plenum around the burner's air input & damper. I haven't forgotten that the commercial units pre-heat the air supply, at this stage I'm not sure I'm going to have to do this but if I do I figured I'd be somewhat prepared. If required the air heater will slip onto this thing.

Fuel Delivery

Now, as I've opted for the ceiling mount furnace I've had to get a bit elaborate with the fuel delivery system. My burner came with what's known as a single stage pump. This basicly means it needs to gravity fed. There are dual stage pumps available that will suck fuel 8-10 feet uphill but I don't have any of those. What I do have is the transfer pump that came with the furnace, I use this to top up a "day" tank that is mounted a little higher up than the burner pump. Why a day tank? It's there to provide some cush in case the main fuel tank runs dry. I've got a site tube on the side of it to tell me how much oil is in it, if I see the level dropping I've got a few hours worth of heat to get the main tank refilled. :) I haven't talked much about the vent on the heater yet, seems like a good place right here-
Waste oil tends to composed of a lot of stuff, there's no telling for sure exactly what's going to be in it. A lot of people tend to dump *everything* in the holding tank, besides crankcase oil you can pretty much count on gear lube, solvent, brake fluid, tranny & power steering fluid, gas, antifreeze, cigarette butts & the occasional dead body. Some of this stuff wants to boil off as things heat up. From experience I can tell you there's a whole lot of burping & spitting out the vent tube when we hit 100C. There's also something coming out at about 70C. Without the vent this stuff would be forming a vapour bubble inside the heater tank, eventually taking up enough room that the fuel level would be pushed below the actual pump's input port. You'll note in the diagram I've run the vent up higher than the day tank "full" level. As oil heats up it expands, about 15% at the temperatures I'm running. As it expands, it's density obviously drops, as it density drops the expanded oil gets pushed up the vent tube by the pressure exerted from the denser oil in the daytank. All this means that the oil level in the vent tube is gonna be higher than the level in the day tank. If the vent tube isn't high enough, we have A Fountain, which Is NOT A Good Thing. :( I've actually run into a problem in that I can't get my vent as high as I would like, dear ol'dad went & put the roof in the wrong spot. :( Under some circumstances the oil actually starts siphoning out the vent tube & back to the day tank. Oh well.

I built the tank out of an old air pig on which the expiry date had er, um, expired.
As I mentioned earlier I'd been running the furnace on more or less regular fuel for two years using this setup without problem. For the conversion I wanted a low fuel switch on the day tank. I built it out of pretty much stock stuff from Princess Auto. A vented tank cover, a surplus level switch, some brass rod & a pill case full of insulating foam for a float.

Ta Da!

Here we are with our first official waste burn.
And our first official steam.
And our first official overflowing overflow collector.

This is when I decided to add a vapour separator,seen here rust-proofing the roof's plywood sheathing, to the vent with a return line to the day tank, .

Oh Well.

But the thing is, It's up, Its burning waste oil, Its heating the garage. :)

One Year Later

As I right this web page it's been about a year that I've been running the furnace on waste oil. I can't say its all been roses-

Now, during the testing, everything worked fine. During actual operations, when the burner's been mounted to the furnace eight feet off the floor & is a royal pain in the butt to maneuver up & down the ladder in tight quarters, everything failed.

Go Figure. :/

Things That Go Bump In The Night, Or Morning Or Afternoon

Tight connections mysteriously leaked.
The flame detector started shutting it down. I eventually figured out that the 1/4" copper pipe between the filter & heater couldn't deliver fuel fast enough when it was cold, causing the burner pump to literally run itself dry. :| I replaced it with 5/16" pipe- No more problems with that.
The flame detector started shutting it down. For some reason I got it in my head that the CAD cell wasn't getting a good view of the flame, I spent literally hours trying to tweak its position. It finally dawned on me to try a new cell- No More Problems.

Some crud plugged up the orifice in the solenoid valve, effectively eliminating the warmup cycle.

The drive coupler between the blower motor & pump self destructed. This time I got smart & just removed the motor instead of bringing down the whole burner. ;>

The pressure guage melted.

The thermocouple wiring came loose, causing the PID to think things were about 40C cooler than they really were. The overtemp shutdown cut in but not before the oil in the heater transmuted itself into something resembling black vaseline.

Due to a miss-adjusted nozzle position there was crud building up in the end of there burner tube affecting the spray pattern & combustion, until I got smart & adjusted the nozzle I had to knock it off about once a week.

At one point I was running a little low on waste oil & so dumped a fair amount of old solvent into the tank. The filter plugged in no time flat & the oil in the day tank got so thin the float on the low fuel switch sank shutting the heater down. :|

The regulator started not regulating, the rubber washer I had used as a disk got pretty baked. I tried replacing it with a brass piece but I just couldn't get it to seal with the seat, causing oil to leak back through & dribble out the nozzle into the firebox. This made for some noisy, exciting moments every time it fired up. :/ By this time I had settled on 100 PSI as the operating pressure so I cried uncle & bought a regular 100PSI relief valve. It screwed right into the regulator base like it was made for it. A little work on the lathe to make a cap to fit over it & it's worked perfectly ever since.

Now, besides the unplanned for stuff there was the stuff I knew I'd be facing going into it-

Nozzles- they plug up. I'm getting 40-80 hours per nozzle, thats roughly one per month for my usage pattern. At ~ $10 per nozzle I can live with that. Delovan makes a special nozzle that is supposed to handle dirt better, I may give this a try but I suspect the actual problem is the oil baking in the nozzle after the burn cycle ends, once the oil stops flowing through (during the burn cycle) the radiant heat from the firebox (before it cools down) could be warming it up pretty good (Bad). At some point I may try adding a cool down cycle where oil gets pumped through the nozzle for a minute or two to keep it from heating up, same sort of idea as the pre-heat idea I did initially do some experimenting with different nozzles, both in capacity & spray pattern, to see what would work best. What worked best turned out to be exactly what was called for on the data plate.

Filters- I've been using standard 25 micron fuel filters, I did try going to an oversized high performance 3 micron filter, it plugged solid in about a week. I've recently installed a 15 micron water blocking filter, we'll see how that works out.

Crud- As I mentioned I did have a problem with this stuff on the end of the burner tube. Since I fine tuned the nozzle position it hasn't caused me any more problems but it does build up inside the firebox & exchanger which can reduce the warm air output. I opened up all the access ports for the fall cleaning/tuneup, it really wasn't as bad as I feared it would be & it cleaned up easily. The fire box had a good coating of fluffy grey stuff, inside the exchanger it was more of a gritty black compound. The stack & chimney were pretty much regular soot.

On the bright side (after I'd bought all the parts to build it :( ) I didn't find any need for an air heater. The flame ignites as soon as the pre-heat ends & the nozzle starts spraying, down at least as far as -20C/-5F. The plenum I put on for the heater did get used though, it hit me one day that as the furnace is up by the ceiling when it wasn't running my nice warm air must be running out through the burner & up the chimney, with freezing cold air coming in from outside to replace it. When it was running, it was sucking *really* sucking up the warm air as well, which was being replaced by that cold stuff. So, I got some dryer ducting & hooked it up to the plenum. Now the burner can only draw air from floor level & the warm stuff stays inside :)

Another change I've now made is a better vapour separator on the vent line. The original was simply a tee fitting with a piece of pipe sticking up a few inches. It usually worked well but if there were some exception rumblings happening it would spray hot oil out the top making a mess all over the place. After the checkup the second time I fired the furnace up it started pulling a gusher act. It appears that the return line to the day tank either had met its match or had just plain plugged up. I replaced it with the design shown in the fuel system diagram above, a cup made out of the upper end of a propane cylinder. The oil shoots down through a baffle (made of 1/8" steel mesh) into the cup. Steam & whatnot escape out the top, the oil drains back to the day tank.
Yes, I increased the drain tube diameter).

Be all that as it may, It's at the point where it starts without problems & runs very reliably, heating the garage to a comfortable temperature & holding it there despite the lack of insulation & somewhat drafty nature of the building.

What Would I do Different?

Is a hard question to answer. Most of the things that were performing marginally have been sorted out as I worked through the design & build anyhow. In it's current state there's really not anything that I feel needs to be changed. Of course, what I've ended up with is neither simple nor cheap although a large proportion of the expenses ended up on stuff that didn't work out. If I just counted money spent on what's actually still attached to the Mk. 2 then the conversion cost me somwhere's between $300 -$400, & probably a quarter of that was those SS braided hoses.
Design wise I would have preferred an simple inline heater between the pump output & the feed pipe, or even a heated feed pipe however with the tank there's a benefit in that the water is boiled off out of the stock before it gets to the nozzle. The fuel return line on the feed pipe may also be extending the nozzle life, keeping the nozzle cooler & cutting down on oil baking inside due to induced heat being radiated into the nozzle heating it above what's needed.

I built this pretty much in a vacuum, which is to say there was very little input from others on how to do it so the decisions I made & paths I took were pretty much all my own. There may well be (in fact I'm pretty sure there are ;) ) better ways of going about this. When I started on this there was very little info to be had by Googling "waste oil furnace", a few links to the previously mentioned drip burners & some sites for commercial systems, that although were obviously based on existing burner heads, were very short on details of how they work. While finishing up this web page I actually stumbled across a group dedicated to these sort of conversions that seems to have started up about the same time I started working on mine.


It's interesting to see how different, & how the same some of the solutions people there came up with compared to what I've done.

So that's it. For anyone contemplating this sort of project my only advice can be "Do The Research", read everything you can get your hands on. The more you know going into something like this the better the odds you'll come out of it alive. :)>

Howard Eisenhauer
Toasty Warm
In the Garage
Feb. 2008

p.s.- Please mention me in your will. :)

If you-

a. Have any other info or links on information about this that you think I'd like to hear about-
b. Think I've made a mistake on the facts-
c. wonnt two pointe out taht im a terribul spieller
d. Would like to call me a Plagiarizing B******D

Please feel free to contact me Here :)