I have recieved and tested 6 samples of biodiesel from home brewers so far this week. Only one measured less than 500ppm. The other 5 were between 700 and 1500. All of the samples looked very clear. Just as I used to do, these brewers are depending on visual clarity to guage dryness. This simply does not work.
Diesel engines, particularly newer ones and common rail types will not tolerate that level of water in the fuel. Note also that fuel filters with water traps are not an answer to this problem. These will only trap free water. The water in these sample was in suspension and thus will pass right through the filter.
I have continued to experiment with methods of drying oil feedstock before processing. Two weeks ago used a fan to blow cold air over hot oil and got a reduction fron 2400ppm to 700ppm in one hour.
This time I used hot air from a paint stripper heat gun to blow hot air over hot oil and got a reduction fron 2400ppm to 50ppm.
Obviously hot air is more effective but at a cost of more energy. No suprise there really
I now feel ready to begin the next phase of water removal. The 33 litres of methoxide that I use to process 150 litres of oil contains .65 litres of water. I am going to atempt to eliminate that water by adding calcium oxide.
As this is a bit off subject the progress of this experiment will be reported on another topic " How to get rid of water from Methanol"
Since my last post I have received 8 more samples from Ireland , UK and USA. All looked clear at room temperature although one was so dark it was hard to tell if it was clear or not. Two tested below 500ppm, in fact one showed a result of 40ppm which is really at the limit of my measurement. The others were between 600 and 1800 ppm which is not acceptable. As the weather gets colder we all need to look at our drying systems and make sure they are operating properly.
I installed a vacuumpump for quicker drying my bd and wvo.
I remarked that after drying the classical way as you do, you can stil extract a lot of water.
I therfore think that after 2 hours of 250°C your wvo is not dry at all and that your zero does not correspond with zero water.
what is your opinion about this?
When I calibrated my manometer I heated the reference bio to 125 degrees C for two hours. I admit that I assumed that this would be sufficient to remove all or very nearly all of the water present. It is possible that a very small amount might remain but I only claim that the manometer is accurate to about 10% so I felt the assumption was justified.
Using a vacuum pump to lower the pressure in the drying vessel is an interesting proposal. Graham Laming does something similar using a venturi. The low pressure reduces the boiling point of water so that the water can be extracted more easily. However it also reduces the boiling point of biodiesel significantly and it is important to keep the temperature lower than the reduced boiling point of bio or you will end up with distilled bio which incidentally looks almost perfectly clear like water.
It is possible that your previous drying method was not working perfectly and this could explain the extra water extracted under vacuum conditions.
i have a waterjet venturipump. it is very inexpencive (plastic edition) and simpel.No motor needed.
it gives very high vacuum. the only disadvantage is that it uses very mutch water. but as I can dispose over undefinit quantities of water, this is no inconvenient for me.
the results are spectaculor. ex: newly fabricated bio becomes clear and translucent in 4 hours.
when drying wvo you see the water bubling out of the oil. since i use it systematiquely i found out that i can always extract water from oil that has been dryed the classical way.
a suggestion: to define the exact zeropoint why not bring an example to a official labo. they can perform the absolute exact watercontent by a system caled carl-fisher, which is out of read for the amateur( to complicated, to expensif and mutch experience needed).
have a nice week end.
I did have my reference solutions checked by a laboratory using the carl fischer process. These solutions were prepared by heating 2 litres of biodiesel to 125 degrees C for 2 hours and then adding .5ml of water to the first litre (500ppm) and 2ml of water to the second (2000ppm).
The first tested at 468ppm and the second 2081 so both results were within my 10% accuracy target.
I did not have the zero water sample tested because the tests were quite expensive but it must have been very close to zero otherwise my test results would have been way out.
It occurs to me that if you were to bring a larger quantity of oil or biodiesel, say 200 litres, up to over 100 C for 2 hours that would certainly not remove all the water. In fact ,depending on how much water was present it could take 10 or 12 hours. I only used that method to prepare a small quantity for reference purposes.
We have been having a cold spell here in Ireland for the last couple of weeks, It didnt rise above ) degrees C for over a week which is unusual here. I keep a glass jar of washed and dried 200ppm) biodiesel near where I park my car so that I can see if there is any danger of the fuel crystalizing in my tank. Beside it I have a jar of washed but not so well dried biodiesel (800ppm). The difference in cold water performance is very distinct.
The dry sample begins to cloud at -3 degrees C and thickens to a creamy consistency by -7. The less well dried sample begins to cloud at 1 degree and thickens at -2. At -7 it resembles butter, if you up turn the jar nothing pours out.
Im not sure why this should happen but it does and suspect that the water crystals which form at 0 degrees act as seed crystals on which biodiesel crystals can begin to form.
Whatever the explanation it reinforces the need to make sure fuel is properly dried before use in cold weather.
I am posting this to make-biodiesel.org. If I can ask those that have mastered this technique to check it out (click the link above) and offer suggestions. I mostly cut and pasted Imake's posts with the lightest edits I felt I could get away with. If something needs changing, please let me know.
Hello Rick, thank you for posting the info on the carbide manometer to your site. In a private message you queried the use of volume rather than mass to measure the water content. Im having trouble logging in to private messages so Ill attempt to answer your question here.
The measurement of parts per million can refer to volume or mass. In scientific notation the number is followed by ppm(v) or ppm(m). A Carl Fisher analysis would read in ppm(m).
I realize that the two figures are not the same because the specific density of biodiesel is different to that of water.
I chose the volume method because it is possible to make fairly accurate reference solutions with equipment that most home brewers already have. A jug for measuring a litre and a 5 ml syringe to measure half a ml.
To go the mass route an ordinary kitchen scales would be ok for measuring 1 kg but you would require something much more accurate and expensive to measure half a gram with any accuracy.
Both methods have limits to their precision but if one has access to a very precise scales accuracy could be improved by making up reference solutions by mass.
Im not proposing the carbide manometer as an alternative to Carl Fischer analysis but with careful calibration you can achieve consistent results to within 10% accuracy which makes it a very useful tool for making high quality biodiesel.
In your introductory paragraph you say that " he puts a measured volume of biodiesel..". It should read " he puts a measured weight of biodiesel...". However the instructions further down the page are correct.
After a lot of thinking about it, I think I understand why you use volume to make the reference solutions and mass for the actual tests.
Using my carbide manometer I recently did some comparisons of the water content of oil feedstock before processing and came up with some interesting results.
My normal oil feedstock after settling in a warm place for two weeks contained 840 ppm of water.
The same oil after spray drying for 2 hours prior to processing contained 180 ppm.
A very similar sample of oil which had been pretreated with glycerol but not spray dried had 410ppm.
Another sample which had been pretreated with glycerol from a previous batch which had used the 5% prewash during processing had 760 ppm.
Although the samples were not identical and there were some other variations I think I can safely say that the glycerol pretreatment method is an effective way to reduce the water in oil feedstock but its effectiveness is severely reduced by using glycerol that has come from a process involving a 5% water wash.
IMAKE- Your SPRAY DRYING seems very effective..
By spray drying I assume you mean pumping the oil through the spray/drying head of your processor, do you heat the oil ?
Spray drying is not really the correct description but yes its the same as your processor.. I heat the bio up to 50 degrees C and then pump it into a fitment that produces a large bubble shaped fountain. I blow hot air at the fountain from a hot air gun from a distance of about 1 foot, (300mm). It normally takes 2 hours to dry a 150 litre batch of either bio or oil.
A couple of points on the manometer idea, and I haven't yet read all the previous eight pages thoroughly so don't kick me too hard if someone has already suggested this.
Make the 'output' end of the U tube a much bigger bore than the input side. This will allow a proportional decrease in the height of liquid pushed up the tube.you will need quite a long tube on the input side though. Or even use a large bore U section connected to the small bore that runs from the sample under test.
Or you can use a short straight section of pipe wit a balloon tied across the end. The diameter of the balloon will give you the volume of gas produced. As it will be under some pressure this is only approximate, but good enough for comparison.
Who's also having trouble with lots of wet wvo.
This is an interesting thread.
It seems that almost no one is using "Dry" biodiesel as described by the ASTM/EN standard.
Has this use of wet biodiesel over the years led to thousands of failed IP/s?
I can not recall hearing about it.
Another point in the mass versus volume argument is that temperature can effect the outcome if you use volume whilst it won't effect measurements using mass.
To quote Dieselbob, one of the most experienced diesel engineers who is not biased against the use of biodiesel...
"It is often thought that the most important factor to consider when using non fossil fuels is that fuel must be clean, whilst this is important it must also be understood that non fossil fuels are hygroscopic, that is they absorb water from the atmosphere. This creates two major problems.
WATER: water is very undesirable in the fuel system and measures must be in place to prevent the water reaching the delicate injection components.
ACID: when water is present in the fuel it creates a chemical reaction resulting in acid formation. Acid can wreck an entire injection system very quickly and will require all the injection components to be serviced and the fuel tank to be internally steam cleaned and flushed with detergent to eliminate all acidity.
Our information is based on several years of testing and having to rectify mistakes made by many past clients".
www.dieselbob.co.uk is a useful and informative site for anyone using biodiesel or SVO.
The manometer uses 3 metres of 5mm tube. A larger diameter tube on the output side of the manometer might make it more compact but it would reduce the read off scale and hence the accuracy of the instrument.
The manometer is affected by temperature but the compensation calculation is very simple. It is posted somewhere earlier on this thread. If the temperature variation is less than 6 or 7 degrees C it makes a negligible difference.
Posted 10 June 2010 10:44 PM
I just set up my manometer and was tinkering with it today. I noticed temperature (vapor pressure) is a significant factor. Do you always measure for your water content at "room temp"? 20 deg. C? Do you correct for barometric pressure? How does your calibration curve look? Have you added more points to the curve and checked the R2 value (the line might not be a straight line)? How much CaC2 do you add to the 100 g sample? How big are the pieces of CaC2? How many feet of the U-tube is filled with liquid on each side?
I'd appreciate the help because my manometer isn't doing much. Just a little liquid change based on the temperature of the liquid.
The above query was passed on to me from Rickdatech from Thor. There are a lot of questions here so Ill try to answer them roughly in order.
Temperature is a factor but not as signifcant as you might think. I calibrated my manometer in a cool workshop at 16 degrees C so if I do a test when the temperature is at say 22 degrees C I would have to compensate.
I think an example is the best way to explain this so lets assume I get a result of 550 mm on my manometer at 22 degrees C. I divide 550 by 297 (297 degrees Kelvin is the same as 22 degrees C) and multiply the result by 291. (291 degrees K = 16 degrees C) The adjusted reading is 538 mm.
Note that the temperature must always be in Kelvin.
If you calibrate a manometer at a constant temperature and barometric temperature the line on the graph should be straight. The volume of gas / pressure is directly proportional to the water content. Inevitably because this apparatus is only accurate to about 10% your line may have a slight bend.
Barometric pressure will have only a tiny effect and is not worth compensating for given the 10% accuracy limit.
The measure I use for calcium carbide is a metal screw top from a wine bottle. It should be filled to a heaped level. I cant remember how much that is in grams but somewhere earlier on this thread I calculated that it is about 3 times more than the minimum required.
The calcium carbide must be crushed to a very fine powder. Crush only sufficient stones needed for each test. Once crushed use the powder immediately as it will start to decompose very quickly. Keep your stones sealed in an airtight container.
There is a picture of the manometer a few pages back which illustrates the lengths of the tube etc.
The left hand side if the u tube which is attached to the pressure vessel is 1 metre long. The right hand side of the u tube which is open at the end is 2 metres long. There is 1.5 metres of coloured water in the bottom of the u tube, .75 of a metre on either side. This means that the level can rise 1.25 metres (1250mm)the right side.
Something isnt right with your manometer. When doing a test you should get a very distinct rise of liquid within minutes.
I will go back and crush my CaC2 smaller...and add more of it (i was only using 1g/test). I will also make sure everything is at the same temperature (no warming or cooling of the fluid before the test).
I'll let you know what happens.
When shaking the pressure vessel be careful not to warm it with your hand. I put two glass marbles in the vessel to help it mix more easily
I crush my stones with a lump hammer on a small anvil, a big flat stone would do instead.
|Powered by Social Strata||Page 1 ... 6 7 8 9 10 11 12 ... 16|