Pics and info submitted by ECN Member Alan Belson.
Here's how I solved the unburned-oil and jet clogging I met with in my earlier machines,HERE, to make a reliable and efficient Vegetable Oil central heating burner. I’ve accepted the compromise of blending VO with diesel oil, though the fuel is no longer entirely 'free'. Features of this burner, built from a recycled pressure-jet unit, are- A dual-fuel system, rationalised electrical fuel-heating, a new exhaust gas re-circulating combustion chamber and additional electrical controls, the latter forming most of the modifications.
Schematic burner assembly. Burner ignites with residual neat diesel oil. Simultaneously, a VO/diesel oil blend is being made at the burner gear-pump. This fuel blend takes about 20 seconds to reach the nozzle, [depending on line volume and fuel consumption]. When the 'programmer call' ceases, [or the boiler thermostat is satisfied], the VO supply pump switches off. Diesel oil then flushes VO from the burner, because a delay timer prolongs the burn by 30 seconds. This only raises the boiler temperature by a few degrees C above normal. Residual neat diesel oil is thus present for the next start. With the combustion chamber fitted it solves the VO polymer clogging problem, giving assured ignition and superb combustion.
Finally, a fan-only sequence cools the 3 lb + weight combustion chamber to protect upstream components from inertial heat, by using a 50 second delay timer. This pic, taken at burn termination, shows why!
Detail drawing of the combustion chamber. 316 grade stainless steel was used for its high temperature qualities.
Combustion chamber, new gasket, spilt sleeve and fittings. The additional drillings are from trials experimental dead-ends. The combustion chamber is designed to re-circulate a percentage of hot exhaust-gas to improve combustion by the gasification of fuel. The band clamping method, [ex dairy supply house], proved very strong and rigid.
Burner on bench trials at c. 21kW. Two motors are fitted, [separating the air-fan and gear-pump functions], to let me use an existing Danfoss LE pump which unfortunately counter-rotated to the original motor & fan on my donor burner. It auto-depressurises the Protek nozzle valve [see later] on idle. A dummy firebox is used in trials to mimic furnace conditions. Recirculation of exhaust only occurs when combustion is so confined. The flame colour is an ephemeral pink, probably due to alloy traces from the firebox’s stainless steel sheet. Otherwise, the flame is clear, with no unburned fuel, smoke, smell of fries or visible soot from the start to finish of the burn. I settled on running c. 50% VO / 50% D during trials, a very sensible and conservative blend for this machine. I aim to increase the VO ratio in long-term trials on the boiler using a gas analyser and a few slight modifications. Nozzle fitted : Delavan 0.5 GPH 80º hollow cone, oil pressure 150psi with a 135psi Protek anti-dribble valve, depressurised on idle by the Danfoss LE pump.
Electrical circuit. The core control of this burner still resides with the original Burner Controller. A variable output peristaltic pump is programmed to start on ignition and injects VO directly into a 'Y' inlet manifold at the gear pump inlet. It receives both VO and diesel oil. The solenoid valve fitted gives a positive VO stop to prevent diesel oil scavenging from the VO supply. Because the peristaltic pump is of positive displacement, the blend ratio is varied simply by changing its speed. Diesel oil is drawn in to make up the burner gear pump's fuel demand, fixed by the gear pump pressure and the nozzle size. Flow settings were obtained and noted in trials. Sequencing is by 230v x 30A relays, breaking into the existing burner controller terminals for the logic as shown. I have used two remote trigger timers for the delay circuits, as these are both accurate and robust. The latching circuit is a simple 2-relay design, which requires a manual intervention in a power outage. Considering the complexity of the operations this is a very simple circuit, but it has a disadvantage. A power outage during a burn would leave a high VO blend in the burner line. I am considering UPS as the perfect solution but the circuit as-built un-latches and needs to be reset if power is lost, giving me the option of a manual flushing procedure. To be honest, I'm not quite sure how to go about choosing a suitable UPS circuit and advice would be welcome. All the burner safety arrangements are retained of course.
The electrical arrangements as built. Labels are gloss photo-paper & were created on the PC. A fourth relay may be seen in the pic at far left, [blue terminals], as I have aleady begun to install a further PTC heater to the fuel line to try raising the fuel temperature and the % VO in the blend.
Schematic fuel supply. VO is pumped wherever possible to avoid mess. The VO tank is a 60 litre plastic drum housed in an insulated box. The VO in the drum is warmed to c. 40-45ºC using four x 15W x 230v lamps, wired in series/parallel for long life, which generate about 30W. This is to avoid fats clogging the VO line. The lamps are mounted on a float riding on the oil surface, bearing a thermostat and a guard grill. The tank is suspended on a weighing scale -it's the fuel gauge. I made a scale disc from photo-paper, marked in litres. VO is drawn from just under the liqid's surface, where it's warmer.
Float shows dimmed bulbs, [one's hiding]. Peristaltic pump, variable, 0.6 - 4.0 litres/hr. 'Santoprene' rubber pump-tubing - warning this rubber is not compatible with diesel oil. A disc thermometer aids setting the stat.
Burner assembled to the boiler using studs to space the combustion chamber and burner chassis at the correct position. A 6"diameter stainless heat shield covers the combustion chamber- it gets very hot [c. 500ºC ] near the boiler end but fan airflow keeps it cool near the burner. PTC heat tape [red cable, exposed here for the pic] is applied to the VO supply line, raising its temperature to about 55-60ºC prior to the Danfoss FPHB 5 preheater, which again raises the temperature close to 70ºC. The Danfoss preheater delays ignition till the oil is hot, with an integral thermostat linking Controller pins 3 & 8. The run is insulated with foam tube.
Boiler room view. I motorised an old dairy-cream separator to centrifuge the VO. 5000G in the rotor traps most of the crud & will process about 20 litres an hour. Note; these machines will not remove breadcrumb or any other floating debris, [such as dried herbs from my pizza-parlor guy!]. The crude VO is pre-warmed to about 50ºC with a deep-fat fryer element fitted in the bowl. This prevents animal fat, which is always present in small amounts in my waste VO, from choking the separator cones and it also gives warning of any water present by 'popping'. The final operation is to 'polish' the VO by pumping it through a 75 mesh strainer and a 20 micron wound cotton filter directly into the VO tank while it's still warm. I have a 150w immersion 'lance' to heat cold oil for the 'polishing' operation if necessary.
The burner works superbly, starts reliably and cleanly and stops without dribbles, with the combustion chamber running at a red heat [c. 600ºC] once up to temperature. The flame is longer but my boiler firebox can accommodate it without flame impingement on the heat-exchanger surfaces. I'm still on a steep learning curve with both the burner and handling the VO and will be still playing with 'improvements' as the winter progresses.
Alan, Very thorough job of engineering and fabrication! In the future, you might consider increasing your heat tape wattage to accomplish the VO heating, without the lamps. Many industrial sites use this technique to maintain fluids at elevated temperatures, so it is a proven technique.
I have installed primary and backup heat trace systems to allow for future failure and easy switch-over to the secondary system. Also a variable voltage control, such as a rotary motor speed control makes it easy to adjust for the need of additional heat.
Good luck with your alternative heating system! You set a good example of recycling the earth's resources.
Where are you getting your oil? Used cooking oil is getting hard to come buy here because so many people are using it. My wife used to have to pay to have it hauled away. Now she barters it out for services. Burger King was selling their oil as long ago as the 90s.
Greg, my oil comes from just a couple of privately owned restaurants, who'd otherwise be paying to have it disposed of at the Commune's décheterie. It's mainly fried "fish & chips" oil, with occasional fried chicken, breadcrumbs 'n herbs debris from a pizza joint. No supply problem as yet- just the opposite - I have over a year's supply in stock now but daren't lose my good suppliers...
Larry, the diesel tank's head has no effect on flow rate, which is controlled soley by the gear-pump pressure and the nozzle's GPH rating. The GPH is constant, regulated by a relief valve in the gear-pump, which bypasses surplus fuel back to [in this case] the gear pump inlet. [I have a 'one-pipe' diesel supply, which I omitted to mention in my efforts to keep the blurb to a minimum - see last pic but one, with the existing redundant return line plugged with a spare fire-valve]. The peristaltic pump has a constant flow rate once set, of about 1 liter per hour. It is only a 3.5 watt motor but [tip] it flipping hurts if you stick your finger in the rotor like wot I did!
Gear-pump failure scenario. Good point, thanks and a new note added to the instructions! No fuel then flows to the burner, which would 'flame out'. The Protek valve shuts at 135 psi, so there is a minimum oil pressure. The trachoidal gear set is virtually oiltight unless rotating, so VO would then be pumped toward the diesel tank. But a 'flame out' will be seen by the photo-resistance flame monitor, which will lock-out the LOA Controller, shut down the burner and wait for a manual reset. A lock out also shuts off the peristaltic pump and its solenoid valve. This usually takes a few seconds, but in this case, worst scenario, the photo-cell might 'see' the combustion chamber for half a minute or so while it cools down. Half a minute is only 8cc of VO, [ 2 teaspoons ], but it's enough to warrant a manual flush-out, which would be my normal reaction to a power outage. Flushing out is easy. I have a spare 3-pin Danfoss LE pump plug and can spin up the gearpump independantly to run diesel out of the finger tightened air-bleed nipple at the 'Y' manifold which is just visible under the pressure gauge, [a necessary chore with 'one-pipes', which don't auto-bleed.]
A non-return valve in the diesel line would risk damaging the peristaltic pump gearing &/or its motor windings - or a blown-off VO pipe- [it's only push fit tube] if it has 'nowhere to go', so that option is out.