Make your
own biodiesel (contd.)
 Back
to Page 1
How
much methanol?
Ethyl esters -- making ethanol biodiesel
Reclaiming excess methanol
More about lye
Using KOH
How much lye to use?
Basic titration
Better titration
Accurate measurements
Joe Street's titrator
pH meters
Phenolphthalein
pH meters vs phenolphthalein
High FFA levels
Deacidifying WVO
No titration?
The basic lye quantity -- 3.5 grams?
Mixing the methoxide
Stock methoxide solution
Poor man's titration
How much glycerine? Why isn't it solid?
Separating
glycerine/FFAs
PET bottle mixers
Viscosity testing
How the process works
What are Free Fatty Acids?
Iodine Values
-- High Iodine
Values
-- Talking about
the weather
-- Summary
Hydrogenated oil,
shortening, margarine
Oxidation and
polymerisation
Which method to use?
Why can't I start with the Foolproof method?
Quality
Quality testing
Cetane Numbers
National
standards for biodiesel
-- standards and the
homebrewer
-- standard
testing
Biodiesel in gasoline engines
Storing biodiesel
Home
heating
Lamps and stoves
Other uses
Fats and oils --
resources
Diesel information
Identifying plastics
Ethyl
esters -- making ethanol biodiesel
Making ethyl-esters
biodiesel using ethanol is a tricky process, not as simple as making
methyl esters with methanol. But it can be done -- the following
technical papers are available online in our Biofuels Library, and
there's sound advice below from a master home-brewer who routinely makes
his own ethyl-esters biodiesel.
Optimization
of a Batch Type Ethyl Ester Process
-- a recipe for biodiesel from ethanol (which you can make yourself),
instead of methanol (which is toxic, fossil-fuel derived, and you can't
make it yourself).
Production
and Testing of Ethyl and Methyl Esters,
University of Idaho, Dec 1994.
Transesterification
Process to Manufacture Ethyl Ester of Rape Oil
by Roger A. Korus, Dwight S. Hoffman Narendra Barn, Charles L. Peterson,
and David C. Drown, Department of Chemical Engineering, University of
Idaho, Moscow, Idaho, USA (Acrobat file, 672Kb)
Making
and Testing a Biodiesel Fuel Made From Ethanol and Waste French-Fry Oil
by Charles L. Peterson, Daryl Reece, Brian Hammond, Joseph C. Thompson,
Sidney Beck, University of Idaho, Idaho, USA (Acrobat file, 2.4Mb)
Biofuels
mailing list member Ken Provost, who has much experience
making ethyl esters, sent us the following tips&tricks sheet.
Ethanol-based
Biodiesel
1. Get plenty of
experience making biodiesel with methanol before you try it with
ethanol. Get comfortable titrating
your oil for FFAs (free fatty acids); you’ll need to do that when you
use ethanol.
2. Try to find a source of KOH (potassium hydroxide) to use instead of
lye with ethanol. Lye (NaOH, sodium hydroxide) will work, but it
dissolves VERY slowly in ethanol. You'll need to use more of either one
-- 7g per litre of clean oil with NaOH, 10g per litre of clean oil with
KOH. More as required per your titration level.
3. Your ethanol will have to be EXTREMELY dry. 199-proof or higher.
"Absolute" ethanol. Any more than one half of one percent
water can kill the reaction. Denaturants like methanol, isopropyl
alcohol, MIBK, etc., are fine. But no water. Ethanol that dry is
difficult to find cheap, especially in the US. If you want to try to
make it yourself, you'll need molecular sieve, quicklime, or something
else to do a "chemical" drying. Distillation alone can't get
the water under 5% -- still way too much. (See below -- Anhydrous
ethanol)
4. If you're interested in ethanol for environmental reasons, be
careful. Even if you find anhydrous ethanol, it may come from fossil
fuel. The denatured alcohols used by painters, or in other industrial
applications, may be anhydrous but still derived from petroleum. In
fact, since fermentation uses water, it's cheaper to make 200-proof
starting with petroleum. The only way to know is to call the original
manufacturer of the formula. Ask if the ethanol is "synthetic"
or "fermentation". One type of denatured anhydrous ethanol
that is almost always fermented is "fuel-grade", which is
199-proof denatured with gasoline. It's what they add to gasoline to
make gasohol.
5. Your oil will also have to be EXTREMELY dry. Heat the oil to 120 deg
C (248 deg F) and hold it there until you can turn off the flame and see
the bubbling stop almost immediately. You might want to throw in some
clumping cat litter (bentonite clay) and/or silica gel to scarf up any
remaining water, let it settle half a day, and take the oil off the top.
Sometimes that's still not dry enough. Remember -- any more than 0.5%
water can kill the reaction.
6. Your oil will have to be fairly low in FFAs. You'll want to do a
titration on every batch to make sure. Anything over 2 ml titration
(using 0.1% w/v NaOH solution) can cause failure of the glycerine to
separate -- under 1 ml is a good idea. Most waste oil is too high, and
either needs to be refined with
NaOH first, or cut with clean oil to neutralize FFAs.
7. You need to use more ethanol to get full conversion. Somewhere
between 275 and 300 ml per litre of oil is about right for most oils.
Coconut oil will need more, maybe 350 ml. Theoretical is about 180 ml
per litre, and the rest is excess to drive the reaction all the way.
8. Even when you do all the above, getting the glycerine to separate is
a matter of good luck and fervent prayer. Sometimes separation occurs
just like a methanol batch. Other times you won't start seeing a
glycerine layer for 3 or 4 hours, or maybe overnight. Then again,
sometimes it NEVER separates. Until you get separation, you haven't made
biodiesel. I've heard of folks who don't wait for separation -- they
just pour the whole mess right into the tank, or do some kind of water
wash and think it's good biodiesel. It might burn, but it's not
biodiesel. It must separate.
9. If it doesn't separate, you can sometimes force it by adding some
methoxide mix. You can also make it more likely to separate by including
some methanol with the ethanol right from the start. For example, you
could try using an initial mix of 5 to 7 parts ethanol and 1 part
methanol. Give it a few hours to separate. If it doesn't, add some
straight methoxide to the kettle, using enough methanol to bring the
alcohols ratio down to 3:1 eth:meth, and containing another 2g of KOH
per litre of oil. That usually initiates separation within an hour.
Fresh, refined edible oil might even work the first time with straight
ethanol. If you use a mixture of ethanol and methanol, you can get away
with 275 ml of initial mix per litre of oil.
10. If you're not scared off yet, Good Luck!
-- Ken Provost
Anhydrous
ethanol
To make ethyl esters the
ethanol must be anhydrous, 99%+ pure -- with less than 1% water content.
The purest ethanol that can be produced by ordinary distillation is only
95.6% pure, the rest being water, which interferes with the
transesterification reaction in making ethyl esters. More common for
home distillation is 170-190 proof -- 85-95% pure.
Members of Journey to Forever's Biofuel
mailing list have succeeded in making ethyl esters using
85% ethanol they've distilled themselves, by removing the excess water
with quicklime (CaO). See "The Manual for the Home and Farm
Production of Alcohol Fuel" by S.W. Mathewson, Chapter
12 -- Drying the Alcohol,
Drying with lime.
An easier method is to use 3A zeolite molecular sieve. Biofuels group
member Ken Provost reports: "Zeolite (aka 'molecular sieve' ) works
BEAUTIFULLY to suck the last bit of water out of distilled ethanol. I
got a sample of Type 3A Molecular Sieve from Adcoa
in Southern California.
"I got a can of the 4-8 mesh -- little balls of rock about
1/8" diameter. They absorb about 20% of their weight of water over
the course of a few hours. Take a litre of 95% ethanol, throw in 250g of
the stuff, swirl occasionally, filter out the next day through a
strainer, and presto! Anhydrous ethanol. Not expensive either -- US$2.05
a pound in 10 lb quantities, and reusable indefinitely. You drive off
the water under a broiler for an hour."
Adcoa
in Southern California":
http://www.adcoa.net/molecula.htm
Or do a search for "3A Molecular Sieve" at Thomas
Register:
http://www.thomasnet.com/nsearch.html?
cov=NA&which=prod&what=3a+Molecular+Sieve
&navsec=search&heading=75170753
or at Google or any good search engine.
An alternative is to run the ethanol vapours through 3A molecular sieve
in a column during the distillation process.
Cornmeal Adsorber for Dehydrating Ethanol Vapors
-- by Michael R. Ladisch et al., Laboratory of Renewable Resources
Engineering, Purdue University. About half the ethanol now produced in
the US is dried using corn grits. When the corn's drying capacity is
worn out, it can be fermented and distilled to make more ethanol. This
1981 paper is the original work on the subject.
Separating
Ethanol From Water
-- by Renaldo V. Jenkins of Langley Research Center, Hampton, Virginia,
USA. More economical methods of separating water from ethanol to produce
anhydrous ethanol, using sulphur or castor oil. Provided by F. Marc de
Piolenc.
Absolute
Alcohol Using Glycerine
-- Mariller-Granger Processes, from E. Boullanger: Distillerie Agricole
et Industrielle (Paris: Ballire, 1924). Mariller's absolute alcohol
production process by dehydration using glycerine, various systems
examined and explained. Translation from the French by F. Marc de
Piolenc.
Reclaiming
excess methanol
Depending on the kind of
oil you're using, it takes from 110-160 millilitres of methanol per
litre of oil to form the methyl esters molecule. But you also need to
use an excess of methanol to push the conversion process towards
completion -- the total used is usually 20% and more of the volume of
oil used, 200 ml per litre or more.
Much of the excess methanol can be recovered after the process for
reuse, simply by boiling it off in a closed container with an outlet
leading to a simple condensor.
Methanol boils at 64.7 deg C, 148.5 deg F, though it starts vaporizing
well before it reaches boiling point.
Unlike ethanol, methanol does not form an azeotrope with water and
relatively pure methanol can be recovered -- pure enough to reuse in the
next batch.
The methanol can be recovered at the end of the process, or just from
the glycerine by-product layer, since at least 70% of the excess
methanol collects in the by-product and it's that much less material to
heat. See Methanol
condenser.
Recovering the methanol at the end of the process seems convenient
because it's already hot, and you can get the methanol back from both
the fuel and the glycerine by-product at the same time. But it can push
the reaction into reverse, converting the biodiesel back into glycerides.
Prof.
Michael Allen
says:
Removing methanol from
the reactor product is a no-no because it reverses the reaction you
have just carried out. Without the excess methanol, the ester and the
glycerol react to provide an equilibrium mix which is less than
desirable. That excess methanol serves a very real purpose in pushing
the equilibrium yield up. But once the glycerol is completely
separated, the reverse reaction cannot take place and we can get back
the excess methanol.
Some people do it before
separating it anyway (they often use a vacuum). We'd rather not risk
ending up with anything less than good process completion. It's best to
recover the methanol separately from the unwashed biodiesel and the
glycerine by-product.
For the by-product, heat to 65-70 deg C (149-158 deg F). As the methanol
evaporates, leaving an ever-lower proportion of methanol in the mixture,
the boiling point will increase, so you'll have to keep raising the
temperature to keep the methanol vaporizing. It goes up to 100 deg C
(212 deg F) or more, and then it starts to froth and you have to stop or
you'll get frothy brown by-product in your methanol condensate. But the
bulk of the methanol should have been recovered by then.
If you have a market for potassium fertiliser and/or industrial-grade
glycerine (about 80-90% pure) which makes it worth the cost of the
phosphoric acid you'll need, it's best to separate
the by-product into its components first. Adding the phosphoric acid
separates the soaps (FFAs) from the glycerin and frees the catalyst, but
it won't separate without the methanol. After separation the methanol is
left in the glycerine fraction and can be removed then.
It's probably more efficient to use flash evaporators to recover the
methanol from either the biodiesel or the separated by-product.
More
about lye
The catalyst used in
transesterification of vegetable or animal fats and oils is lye --
either sodium hydroxide (NaOH, caustic soda), or potassium hydroxide
(KOH). NaOH might be easier to get and cheaper to use, KOH is easier to
use and it mixes with methanol much more easily. KOH does an all-round
better job than NaOH.
From
a faulty batch of lye (rare) -- the semi-translucent half-pearls
are fresh, but the chalky white lump on the right is carbonated.
After only a second or two exposed to the air the fresh
half-pearls are already absorbing a sheen of moisture.
|
Keep
it dry
Lye is hygroscopic -- it absorbs water from the atmosphere. So make sure
you get fresh lye, and keep the container tightly sealed.
When weighing it out, don't leave it exposed to the air for any longer
than necessary. In humid weather we weigh it out into plastic bags, one
on either side of the scale to equalise the extra weight of the bag. As
soon as it's weighed out, close the container, close the bag, and add
the lye to the methanol as quickly as possible.
If the weather's really damp, you can do the whole weighing process
entirely inside a big clear plastic bag. If you feel it still might be
too moist even inside the bag, put some lye in an open container inside
the bag to absorb the moisture first.
We buy 85% KOH in 20 kg lots, 44 lb, and transfer it from it's tough
airtight clear plastic bag this way:
Everything goes inside a large clear plastic bag -- the sealed bag of
KOH, 10 empty 500 g HDPE plastic containers with wide mouths, bungs and
tight-fitting lids, a little KOH in another 500 g HDPE container with
the lid off to absorb moisture, some tough cord, and a pair of scissors.
Then use two knots of cord to seal the large bag closed.
We cut two holes in the side of the large bag and taped a resistant
plastic glove into each of them. With your hands in the gloves, it's
easy to work efficiently and safely transferring the KOH inside the
large bag, with no water present.
When it's done, press all the air out of the KOH bag and tie it closed
with two knots of the cord. Make sure all the filled HDPE containers are
properly closed. Take your hands out of the gloves, cut the knots
closing the big bag, take out all the HDPE bottles, the scissors and the
remaining cord. Press all the air out of the big bag, and tie it closed
with two knots of cord.
Lye also absorbs carbon dioxide from the atmosphere and becomes
carbonated if not stored properly. Carbonated lye is chalky white, fresh
lye is almost translucent. You can still use carbonated lye if it's not
too severe, but you'll have to use a bit more. Add about 25%, or better,
check it against a titration with fresh lye.
Where
to get it
You can get both NaOH and KOH from soapmakers' suppliers and from
chemicals suppliers. NaOH might be more readily available and cheaper
than KOH, but try to find KOH, it's worth the effort.
Lye usually comes in flakes and 5mm pearls or half-pearls, or small
pearls (1-2 mm), which with NaOH are usually 99%+, but might be more
expensive. Any of these will do if the purity is about 96% or better for
NaOH. In the US, Red Devil lye is pure NaOH, if you can still get it
(it's no longer manufactured).
Don't use Drano or ZEP drain-cleaners or equivalents with blue or purple
granules or any-coloured granules, it's only about half NaOH and it
contains aluminium -- it won't work for biodiesel. It's best to get your
lye from soapmakers or chemicals suppliers. Suppliers for schools might
have smaller quantities.
Using
KOH
KOH is not as strong as
NaOH -- use 1.4 times as much KOH (actually 1.4025 times).
Titration is basically the same.
You can use either KOH solution or NaOH solution for titration. Unlike
NaOH, your KOH probably won't be pure, but as long as you use the same
strength of KOH for both the titration solution and the reaction it
won't make any difference.
Mix the KOH solution the same as for NaOH: add 1 g of KOH to 1 litre of
water. Or do it the Better
titration way. For each 1 millilitre of solution used in
the titration add 1 g of KOH to the basic amount.
You can also use the usual 0.1% w/v NaOH solution and convert the end
result for KOH, see below.
Instead of the basic 3.5 grams of NaOH per litre of oil, use 3.5 x 1.4 =
4.9 grams of KOH (4.90875). So, if your titration was 3 ml, use 3 + 4.9
= 7.9 g KOH per litre of oil.
One more complication: KOH is generally not as pure as NaOH. KOH is
usually 92%, 90% or 85% pure -- check the label. We use half-pearls
assayed at 85%, with good and reliable results. KOH is available at 99%
strength, but it's hard to find and it's expensive, and anyway 92% or
85% work just fine.
Adjust the basic lye quantity according to the strength of the KOH: the
basic 4.9 grams of KOH at 100% strength would be:
To substitute KOH for the
3.1 grams of NaOH per litre of oil used in the acid-base process:
To convert NaOH quantities
for KOH:
For 85% KOH use 165% of
the NaOH amount (x 1.65).
For 90% KOH use 155.83% of the NaOH amount (x 1.56).
For 92% KOH use 152.45% of the NaOH amount (x 1.52).
Example:
You're using 85% KOH and the titration needed 3 ml of KOH solution to
bring the pH to 8.5.
Number of grams of KOH required for the reaction:
Basic amount: 5.8 g KOH
(5.775)
Titration result: 3 ml
Add: 3 g KOH
Total: 8.8 g (8.775) of 85% KOH required per litre of WVO.
Using 0.1% NaOH solution
for titration and the same WVO as above: the titration result is lower,
it only needs 1.8 ml to reach pH8.5. (Actually 1.82 ml, according to the
calculation.)
Basic amount: 3.5 g NaOH
Titration result: 1.8 ml (1.82)
Add: 1.8 g NaOH (1.82)
Total: 5.3 g NaOH (5.32)
Convert for 85% KOH:
5.3 (5.32) x 1.65 = 8.8 g (8.778) of 85% KOH per litre of WVO.
We usually use NaOH
solution for titration and convert the result for our 85% KOH, but in
practice NaOH solution and KOH solution both work equally well.
One reason for preferring
NaOH solution is that it's usually the standard used in describing FFA
content of different oils -- 2 ml titration WVO or 5 ml titration WVO
invariably refers to ml of 0.1% NaOH solution, unless otherwise
specified. In the example above, the WVO would be described as 1.8 ml
titration WVO.
KOH dissolves in methanol much faster and more easily than NaOH does,
and doesn't "clump" together as NaOH can do. When you use KOH
the glycerine by-product is liquid and won't solidify.
KOH is easier to use than NaOH. It's more flexible and adaptible and it
gives generally better results. We seldom use NaOH (except for
titration).
How
much lye to use?
It requires 3.5 g of NaOH
lye or the equivalent of KOH (see above) per litre of oil as catalyst to
transesterify new, unused oil.
This makes standard-quality biodiesel within the usual processing
parameters of methanol quantity, agitation, processing duration and
processing temperature.
Fresh cooking oil contains a standardised amount of Free Fatty Acids (FFAs),
which can slow or stop the transesterification process. The lye is
alkaline and neutralises the FFA by turning it into soap, which drops
out with the glycerine layer, and with 3.5 grams of NaOH per litre of
oil there's enough left over to catalyse the biodiesel reaction.
Used cooking oil (WVO) needs more NaOH than new oil, not to serve as the
catalyst but to neutralise the extra FFAs formed in cooking the oil. The
hotter and longer it's cooked, the more FFA it contains, and the more
lye is needed to neutralise it.
You have to titrate the oil to determine the FFA content and, from that,
how much extra lye will be required. Titration measures the pH of the
oil, that is, the acid-alkaline level (pH7 is neutral, lower values are
increasingly acidic, higher than 7 is increasingly alkaline, or
"base"). From this you can calculate how much extra lye will
be needed to neutralize the extra FFA.
Too much lye will make extra soap, with very alkaline biodiesel that's
difficult to wash, with loss of production, or it can ruin the reaction
when the ratio of soap to biodiesel reaches a point where the whole
batch turns into "glop soap".
Too little lye will mean some of the oil is left unreacted. See below:
How the process works
The basic lye quantity -- 3.5 grams?
Good-quality oil that hasn't been cooked too much or
overheated is quite forgiving, but with poor-quality WVO with a high
Free Fatty Acid (FFA) content, accurate titration is more important. The
higher the FFA level, the more sensitive the reaction, the more precise
you have to be with titration and everything else, the more reactive
agents you'll need -- and the lower will be the production rate.
Basic
titration
An electronic pH meter is
best, but you can also use pH test strips (or litmus paper), or
phenolphthalein solution (from a chemicals supplier).
Dissolve 1 gram of lye in 1 litre of distilled or de-ionized water (0.1%
w/v lye solution) (weight to volume).
In a smaller beaker, dissolve 1 ml of dewatered
WVO oil in 10 ml of pure isopropyl alcohol. Warm the beaker gently by
standing it in some hot water, stir until all the oil dissolves in the
alcohol and the mixture turns clear. If you're using phenolphthalein,
add 2 drops of phenolphthalein solution.
Using a graduated syringe, add the 0.1% lye solution drop by drop to the
oil-alcohol-phenolphthalein solution, stirring all the time. It might
turn a bit cloudy, keep stirring. Keep on carefully adding the lye
solution until the solution stays pink (actually magenta) for 15
seconds.
Take the number of millilitres of 0.1% lye solution you used and add 3.5
(the basic amount of lye needed for fresh oil). This is the number of
grams of lye you'll need per litre of oil.
With a pH meter or test strips, use the same procedure without adding
the phenolphthalein. Add the 0.1% lye solution drop by drop as before
until the pH reaches 8.5.
Better
titration
Unless you have a very
accurate scale, it's not easy to measure exactly 1 gram of lye. It's
much easier to measure 5 g accurately than 1 g, so mix 5 g of lye with
500 millilitres of distilled or de-ionized water to make a stock
solution.
Before titration measure out 5 ml of the stock solution, add 45 ml of
distilled or de-ionized water. This makes a 0.1% w/v lye solution.
It's also not easy to measure exactly 1 millilitre of oil. Instead of
the usual 1 ml of oil and 10 ml of isopropyl alcohol, mix 4 ml of oil in
40 ml of isopropyl alcohol in a glass beaker.
Warm the mixture gently by standing the beaker in hot water, stir until
all the oil disperses and it becomes a clear mixture.
Then titrate as usual, measuring millilitres of stock solution used.
When it reaches pH8.5 count up the number of millilitres used as usual
and divide by 4. This will give a much more precise measurement.
To save on isopropyl alcohol, use 2 ml of oil in 20 ml of isopropyl and
divide the results by two -- still twice as accurate.
Accurate
measurements
When "newbies"
have problems making their first test batches and it doesn't pass the
quality tests it's most often because of inaccurate measurements, or
they didn't follow the instructions closely enough. Here's some general
advice on being more accurate.
Weights
Truly accurate scales are expensive, unless you can pick up a
second-hand set in good condition. Even then, it pays to check your
scales for accuracy.
A good way of doing this is with new coins. Find out from your bank, or
the Central Bank, what the weight is (in grams) of the coins in your
country. If you get a full set of new coins you can use them in
combination to check the accuracy of a wide range of weights.
With a balance-type scale (two sides with a fulcrum between them), get
two full sets of coins, and figure out different combinations to put on
each side; you should also be able to use this to achieve smaller
gradations than your scale allows: aim to measure a tenth of a gram, 0.1
grams, preferably less -- 0.01 grams is best.
At 0.1-gram accuracy, with a one-litre test batch, measuring the 3.5
grams of NaOH required might give you 3.4 grams or 3.6 grams instead, or
anything in between the two. With fresh oil this margin of error or even
more won't matter, if you get everything else right you'll have good
results.
That margin of error won't matter with test batches of used cooking oil
(WVO) either, IF you heed the advice to avoid oils with titration levels
higher than about 3 ml of 0.1% NaOH solution for your first test
batches.
WVO with higher titration levels and higher levels of Free Fatty Acids
will not be so forgiving, there's much less margin for error -- leave
them for later when you're more experienced.
Small test-batches of high-FFA oils with titration levels higher than 6
or 7 ml will be easier with scales accurate to 0.01 grams and might need
bracket tests as well, and you'll
probably do better with the two-stage acid-base
process, which is also not for novices.
Volumes
If you have a standard millilitre measure that you know is accurate, use
it to check all your various measuring flasks, syringes, pipettes, etc.
Otherwise, check them against each other. With syringes or pipettes, or
whatever you use to add the 1 ml of oil and the 0.1% lye solution to the
titration mixture, you should be able to measure 0.1 ml accurately or
better.
Smaller syringes are more accurate for titration, 5 ml syringes are a
bit big, 2.5 ml will do. We use 1 ml clear plastic syringes, long and
narrow, marked in tenths and tenths-of-tenths of 1 ml -- 0.01 ml. The
measurement scale is 2-1/4" long (5.5 cm) so it's easy to read.
They're cheap, you can get them from hobby shops (you don't need the
needles).
Work with good lighting and on a white surface. First draw the syringe
plunger back about 1/8" (2 mm) to take in some air. Then insert the
end into the oil or NaOH solution and fill the syringe. Hold it up level
with your eye, preferably with a well-lit white wall in the background,
keep it vertical, and carefully empty a few drops, drop by drop, until
the bottom of the surface meniscus is level with the 1 ml mark.
When emptying the syringe into the titration vessel, don't empty it
completely -- the one millilitre volume ends at the end of the scale,
which leaves a little extra in the spout. Empty the syringe only to the
end of the scale, with the bottom of the surface meniscus level with the
0 mark.
Use measuring flasks the same way. When filling them hold them vertical,
with a well-lit white wall in the background, and with the gradation
mark for the quantity you want level with your eye. Measure from the
bottom of the surface meniscus.
Always keep your glassware and other measuring equipment meticulously
clean. Barely visible smears of dried caustic residue or other chemicals
can upset your results.
Joe
Street's titrator
Constant
heat, constant agitation of the titration sample (note the test
tube at the right end of the box)
|
Joe Street of Ontario, Canada, made a titrator kit that he says makes
the process easy and repeatable even in the field, by incorporating a
small heater and magnetic stirrer in a small portable low-cost unit.
"What I needed was a way of warming the mixture of test oil and
isopropyl alcohol, which is required for an accurate titration, and I
thought it would be nice if agitation was included so that it would not
be necessary to stop and shake or stir the solution with a stick while
simultaneously trying to carefully control the flow of 0.1% caustic
solution used in the titration. A laboratory grade stirring hotplate
would be really nice but I could not justify the cost of $500.00 to
$1000.00. Necessity is the mother of invention..."
Joe's titrator only cost him a few dollars. "It was quite simple to
put together and was done in less than a day," he says. "Now I
have both hands free and can dribble the solution smoothly without
interruption just watching for the indicator to go off. I made my life
easier by fixing two of the variables: the heat is always consistent,
and so is the agitation."
Full how-to instructions at Biofuel
cooperative,
Joe's website:
http://www.nonprofitfuel.ca/Titrator.html
pH
meters
It's said you can't
reliably use an electronic pH meter for titration, nor to check the pH
of biodiesel, because biodiesel is not an aqueous solution. Not quite
true -- biodiesel is hygroscopic and will always have about 1,200-1,500
ppm water content absorbed from the atmosphere, if from nowhere else.
With titration, you're adding water with the 0.1% NaOH solution. And
laboratory-standard titration equipment uses electronic pH meters.
We have three pH meters, one of them rather expensive, and we did some
comparisons, with phenolphthalein, fresh from a laboratory supplies
company in Tokyo, and with various test strips. We used WVO from several
sources, and fresh oil as a check. The results were checked with test
batches. In each case, the three pH meters agreed with each other and
produced good test-batch results. Phenolphthalein results were
consistently slightly higher, but the test-batch results were still
good. The test-strips came a poor third -- not as precise as pH meters
and phenolphthalein. People do use them with apparent success, but we
don't think test strips are precise enough for titration.
See Technical
tips for
advice on using and maintaining pH meters.
Phenolphthalein
Phenolphthalein is often
confused with "phenol red" (phenolsulphonephthalein), obtained
at pool supply stores and used for checking water. It's not the same
thing, and phenol red won't work for titrating WVO, its pH range isn't
broad enough. It ranges from about pH 6.6, at which point it's yellow,
through orange, to a maximum of pH 8.0, red. It's at its most accurate
at pH 7.4. For accurate titration you need to be able to measure pH 8.5.
Phenolphthalein is colourless up to pH 8.3, then it turns pink
(magenta), and red at its maximum of pH 10.4. When it just starts
turning pink and stays that way for more than 15 seconds, it's measuring
pH 8.5.
With good-quality oil with low FFA levels you might just get away with
using phenol red for titration, but for higher FFA levels it isn't
accurate enough. Use phenolphthalein -- specifically, 1% phenolphthalein
solution (1.0w/v%) with 95% ethanol.
Phenolphthalein lasts about a year. It's sensitive to light, store it in
a cool, dark place.
pH
meters vs phenolphthalein
Which is better? There's
lots of argument about it. Biofuel
mailing list member Jim
MacArthur (aka
"the skapegoat"), who teaches chemistry at Three Rivers
Community College in Poplar Bluff, MO, had this to say about it:
I can shed some light on
using pH meters vs. phenolphthalien (PHTH) for titration from a
chemist's perspective.
As acids become weaker, it is important to become more careful in the
selection of a chemical indicator (such as PHTH), as the change in pH
is much slower than it is for a strong acid titration. PHTH is the
most commonly used indicator for strong acid titrations because it has
such a clear color change. Fortunately it appears to change color over
the same range as the equivalence point for FFAs (Free Fatty Acids).
The difficulty in either method is knowing what the equivalence point
is going to be. This will vary based on the type of oil and the degree
of hydrogenation. I don't suspect it will vary much, but it will vary.
If your goal is a pH of 8 or 9, then using PHTH is going to be as good
as a pH meter. If you have a better idea of what the equivalence point
is going to be, then a pH meter works better.
The point is there are a number of reasons why PHTH may not work as
well as a pH meter in some cases yet works fine in other cases. This
is something I think everyone will have to decide for themselves. --
the skapegoat, Jun 2005
High
FFA levels
Most people find their
used cooking oil generally gives a titration of 2-3 ml, but some used
oils can have much higher FFA levels than this -- we've seen horrific
titration levels of 9.6 ml. "Horrific" because FFAs are not
good for you -- it's a very bad idea to eat food from a restaurant that
does that to their cooking oil. Another biodieseler reported titration
levels of 16 ml -- black stuff with the consistency of sump oil.
We did succeed in making biodiesel with our 9.6 ml oil. It's not easy to
process oil like this with the usual single-stage base process. You're
likely to end up with about 50% production half the time, and maybe not
a very good product, and glop the rest of the time. If you're really
precise with everything you can do it -- we managed to get a consistent
75% production with the single-stage base process, good product, easy
wash.
The oil has to be thoroughly dried first -- traces of water make a
bigger difference with high FFA levels, because there's more lye for the
water to react with. And the reaction itself releases traces of water,
especially with high levels of lye.
The better answer is to use the Foolproof
two-stage acid-base method, which effectively handles high FFA levels
and still produces high production rates with low levels of reactants
and easy washing.
You can also deacidify the oil.
Deacidifying
WVO
In commercial oil refining
this is done with lye (NaOH), which saponifies the Free Fatty Acids,
converting them to soaps which can then be removed, but it usually
requires a centrifuge to separate it. This is an easier way, no need for
a centrifuge.
Use the titration amount of NaOH -- eg, 9.6 grams per litre of oil for
our high-FFA 9.6 ml WVO (see above). Mix the NaOH with 40 ml of water
per litre of oil. It gets hot. Use a stainless steel container, mix it
outside (by stirring), and take care! This is very corrosive stuff, take
full safety precautions, have running water handy.
When the NaOH is fully dissolved add the solution to the oil (room
temperature), stir gently by hand until thoroughly mixed. Be gentle!
Settle overnight. This leaves soapstock at the bottom. The water stays
with the soapstock.
Filter to remove the soapstock -- no need for fine filtering, fine steel
mesh will do (like a fine tea strainer). Again, do it gently.
Now process the filtered WVO as usual for fresh oil -- 3.5 grams NaOH
per litre of oil, but use 25% methanol, process at 55 deg C (130 deg F),
with good and prolonged agitation as usual.
In our tests the product was good, the production rate was 80%. With
high-FFA oil like this, it's a a much easier process than the normal
single-stage, and it's nice not to have to make such strong methoxide as
a straight single-stage process would require with this oil, 13.1 grams
of lye per litre oil, or more like 13.6 grams (it needs a bit of excess
lye).
It's an alternative -- better than straight single-stage base for oil
like this, and while it won't get as a high a production rate as the
acid-base method, and it uses more catalyst and gives you more
co-products, it's very quick and simple.
This is also useful if you're making ethyl
esters biodiesel, using ethanol rather than methanol: the
ethyl esters process doesn't work well with oils with more than about 2
ml titration.
As always with a new process, try it first with a small sample, say 1
litre of oil. Be gentle when mixing it -- if you agitate it too much it
won't separate easily. If that happens, try heating it, and be more
gentle next time.
You can add the soapstock to the glycerine layer after separation and
neutralize as usual to separate
catalyst, glyc and FFAs.
The soapstock can be used for producing soap, or turned into calcium
soap, which is something like Dubbin and has an extremely low water
solubility. Useful stuff. "Thus an equimolar amount of calcium
chloride may be directly added to the soapstock and prompt separation of
the calcium soap by precipitation from a relatively pure saline (NaCl)
solution will ensue. Calcium soaps are useful industrial ingredients,
for instance as demoulding agents." -- Chemical Reactions of Oil,
Fat and Fat Based Products -- Neutralization (chemical processing)
http://alfa.ist.utl.pt/~fidel/creac/sec34b.html
Mix some calcium chloride in a little water (careful, it gets hot, don't
splash) and add it to the soapstock a little at a time, stirring it in,
until it separates.
No
titration?
There are three ways of
avoiding titration:
- Use the two-stage
base-base method;
- Use the two-stage
acid-base "Foolproof" method;
- Do a series of test
batches using graded quantities of lye and compare the
results. Start off with maybe 6 g of lye per litre (3.5 g for the
transesterification and 2.5 g for extra FFAs). If that works really
well, then go ahead. Otherwise try more tests, at 5 g, then at 7 g;
if, say, 7 g is better, try 6.5 g and 7.5 g, and so on until you get
satisfactory results.
See also Stock methoxide solution
and Poor man's titration,
below.
Satisfactory results mean
that you get a good, clean "split" (ie separation), that it
settles well, leaving a clear product with not too much soap formation,
a good production rate, and, most important, that it washes easily
without frothing.
There are people who claim titration isn't necessary, just use 6 or 6.5
g per litre and you'll be fine. Don't listen to them! They might have
always done it that way and they've driven 20,000 miles already in their
diesel without any problems, but 20,000 miles is nothing in the life of
a diesel motor. Oils vary considerably from place to place -- even the
fabled "stable source of supply" can't be all that stable,
unless it's a food processing factory with a standardized operation. If
it's the usual restaurant or canteen, that would mean they cooked
exactly the same number of identical meals in exactly the same way,
every day. It just ain't so.
Sometimes the "no titration" folks point to Aleks Kac's
two-stage methods, neither of which use titration, though they're both
based on an "average" lye requirement of 6.25 g/litre. But
two-stage processes work in a different way, and this cannot be applied
to a single-stage process. In fact even with two-stage processes we do a
titration, just to know what sort of oil we're working with.
They also point to Mike
Pelly's statement that he usually needs between 6 and 7 g of lye --
but Mike also says titration is the "most critical" step in
the process: "Make your titration as accurate as possible."
And: "It's a good idea to do this entire process [titration] more
than once to ensure that your number is correct." In fact Mike has
a stable source of supply for his WVO, but he checks it regularly just
the same, by titration and/or with test batches.
Titration won't tell you everything about your oil, but it will tell you
more than anything else will (short of Gas Chromatograph analysis), and
the more you know about your oil the easier your life will be. It is a
basic technique in making biodiesel, not to be avoided.
The
basic lye quantity -- 3.5 grams?
This is the amount of lye
(NaOH, sodium hydroxide) required as catalyst to transesterify 1 litre
of fresh, uncooked oil. For used oils, titration determines the amount
of lye needed to neutralize the Free Fatty Acid (FFA) content, and this
quantity is added to the basic figure of 3.5 grams per litre.
In fact 3.5 grams is an empirical measure -- an average. Different oils
have slightly different requirements, and even the same type of oil
varies according to how and where it's grown. Other estimates are 3.1 g,
3.4 g, and some people have set it as high as 5 g.
Here is what we've found. For most fresh oils and low-FFA used oils
(with titration levels less than 2-3 ml), 3.5 grams works just fine. For
high-FFA used oils, use more lye -- up to about 4.5 g instead of 3.5 g.
Do small test batches to see what works best.
Different oils also require different amounts of methanol -- see How
much methanol? For oils and fats requiring more methanol --
coconut, palm kernel, as well as tallow, lard, butter -- again, use more
lye, up to 4.5 g, even with new oils, and especially when it's used.
Once again, do small test
batches first.
Mixing
the methoxide
See Methoxide
the easy way. You can use the easy method with 4-gallon
HDPE carboys or similar containers with screw-on caps (preferably with
bungs as well). First the methanol, then add the lye gradually. Swirl it
about from side to side rather than shaking it up and down.
If you shake it a lot, and often, it can be ready a lot sooner than 24
hours -- in just a few hours or even as little as half an hour, some
people say. But DON'T use it until ALL the lye is thoroughly dissolved.
If you use a white translucent HDPE container you can see any
undissolved lye at the bottom of the container.
KOH -- potassium hydroxide -- dissolves in methanol much faster than
NaOH, sodium hydroxide. It can be ready in as little as 10 minutes.
We use HDPE carboys with two screw-on caps and an aquarium air-pump to
transfer the mixed methoxide to the reactor vessel via plastic tubing
(the braided translucent type), with no exposure at all. Clean, safe and
simple. See Methoxide
transfer. We transfer the methanol from its tank to the
carboys the same way.
For HDPE, see below, Identifying plastics
Stock
methoxide solution
Stock methoxide solution
is very useful for making test batches, with a series of tests made in a
blender using
different amounts of lye for each. Rather than measuring tiny amounts of
lye for each half-litre (or whatever) test batch, make a stock solution
using one litre of methanol and 50 grams of lye. Then you can dilute
quantities of the stock solution to whatever strength each test batch
requires. If you're making half-litre test batches and using 20%
methanol, measure out the methoxide this way:
If titration was, say, 3 ml, you'll need 3 + 3.5 grams of NaOH lye (the
basic amount) for the reaction -- that's 6.5 grams per litre of oil. For
half a litre of oil, that would be 3.25 grams and 100 ml of methanol at
20%.
It's easy to calculate that 65 ml of the stock methoxide solution will
contain 3.25 grams of lye. So measure out 65 ml of the stock, and top it
up with 35 ml of pure methanol to make 100 ml (20%).
For a test of 6 grams per litre of oil, you'll need 60 ml of the stock,
top it up with 40 ml of pure methanol to make 100 ml. For a 7 grams per
litre test, measure out 70 ml of the stock, top up with 30 ml of pure
methanol. And so on.
Didn't figure the calculation? For a half-litre test at the rate of 6.5
grams of lye per litre of oil, divide 6.5 by 2 = 3.25 grams. The stock
solution is one litre of methanol plus 50 g of lye. To calculate the
amount of stock solution needed, multiply 1000 (1000 ml per litre) by
3.25 divided by 50 = 65 ml. For 20% methanol, 20% of 500 ml (half a
litre) = 100 ml. 100 ml minus 65 ml = 35, so add 35 ml of pure methanol
to the 65 ml of stock to make up 100 ml containing 3.25 grams of lye,
equivalent to 6.5 grams of lye per litre of oil with 20% methanol.
Once mixed, methoxide won't last forever, but it's good for a few weeks.
Don't make large amounts -- one litre is good for a dozen or more tests.
If in any doubt, make up a fresh batch. Include what's left of the old
mixture in the methoxide for your next full-sized batch of biodiesel.
(With thanks to Todd Swearingen of Appal Energy.)
Poor
man's titration
Essentially, the
"poor man's bracket method" of titration is lining up a
half-dozen pint jars of the same type, placing 200 millilitres of oil in
each, adding methoxide of differing concentration to each and observing
the reactions.
The easiest way to alter the concentrations is to prepare a stock
solution of methoxide. Mix up a stock solution of 20 grams of NaOH with
400 ml of methanol. Your oil volume of 200 ml is 1/5th of one litre,
meaning that you should reduce the usual 20% methanol volume
proportionately, to 40 ml. However, it is suggested that you reduce it
only to 50 ml to ensure that there is no shortage of alcohol.
For instance, if you want to see what 6.5 grams of NaOH per litre of oil
will achieve in the way of a reaction, you need 1.3 grams of NaOH for a
200 ml test batch (6.5 divided by 5 = 1.3). The stock solution is 20
grams of NaOH per 400 ml of methanol: it's easy to calculate that 26 mls
of stock methoxide will contain 1.3 grams of NaOH (400 x 1.3 / 20 = 26).
Combine 26 mls of the stock methoxide with 24 mls of pure methanol (=
50). With 1/5-sized samples, that amount is exactly what you want, the
same ratio as 6.5 grams of NaOH to one litre of oil.
Number your six identical test jars from 1 to 6. For this example,
they'll contain the equivalent of 5.0 grams, 5.5 grams, 6.0 grams, 6.5
grams, 7.0 grams and 7.5 grams of pure NaOH per litre of oil processed.
You'll have 1/5th of one litre of oil/fat in each jar (200 ml), with 50
ml of methanol, however, each jar will have a different concentration of
lye, ranging from 5.0 grams to 7.5 grams. Calculate the proportions of
stock methoxide solution and pure methanol for each of the samples as in
the paragraph above.
Add the following to each sample:
1) 20 ml stock solution + 30 ml pure methanol (5.0 grams/litre of oil)
2) 22 ml stock solution + 28 ml pure methanol (5.5 grams/litre of oil)
3) 24 ml stock solution + 26 ml pure methanol (6.0 grams/litre of oil)
4) 26 ml stock solution + 24 ml pure methanol (6.5 grams/litre of oil)
5) 28 ml stock solution + 22 ml pure methanol (7.0 grams/litre of oil)
6) 30 ml stock solution + 20 ml pure methanol (7.5 grams/litre of oil)
When all of the methoxide portions are prepared, add them to their
respective jars of oil. Seal all lids tightly. Vigorously shake the
contents of each jar 50 times. Repeat the agitation several times over a
period of 10 minutes.
Let all samples settle. If settling can be conducted in a shallow water
bath at 120 deg F (49 deg C) all the better.
Observe. The reason for using jars of the exact same type is to assist
in more accurate visual comparison and conclusions. If possible transfer
the contents of the jars to a 250-500 ml graduated cylinder for a more
precise comparison. The appropriate amount of lye can be reasonably
estimated based upon the volume of glycerin cocktail that settles out in
each sample.
To fine-tune it, try one test batch 0.25 g on each side of the batch
that had the most glycerin cocktail. Say that was with the 5.5 gram
sample, do three more tests, for 5.25 grams, 5.5 grams and 5.75 grams.
(With thanks to Todd Swearingen of Appal Energy.)
How
much glycerine? Why isn't it solid?
Just
made 20 minutes earlier, and still settling.
|
Newcomers to biodiesel making their first batches sometimes think it all
went wrong because the glycerine didn't go solid.
Messages sent to the Biofuel
mailing list:
"I did my first
test batch of wvo biodiesel over the weekend. Although I appear to
have formed a layer of glycerine on the bottom of the flask, it is not
congealed, but is still liquid at room temperature (24 hours later).
Did something go wrong?"
"The glycerin I got
from my first batch is thinner than molasses at room temp. Why would
it remain so thin? The information I have found suggests that it
should be solid, or close to it, at room temperature."
Others think it
"didn't work" because there wasn't "enough"
glycerine:
"I did a first test
batch of 3 litres WVO, adding 600ml methanol and got only 350ml of
glycerine. I didn't really know what to expect, but 10% relative to
the original stock seemed kind of low to me."
"I did a blender
test batch of biodiesel last night. I titrated to 2ml, so I used 5.5g
of lye and mixed it in with 200ml of methanol. I then heated the oil
up (1 litre), and put it in the blender, and mixed for 15-20 minutes.
It quickly began to separate and the biodiesel at the very top inch of
the blender after the first 15 minutes or so was quite clear. I woke
up in the morning and only 125 ml or glycerine settled out. Shouldn't
there be at least 200 ml of glycerine settled at the bottom? There are
only two layers, the top, light and slightly cloudy, and the dark
glycerine. Where is the extra 75 ml of glycerine?"
In fact in all four cases
the tests worked just fine.
There is no "set" amount of by-product, such as 200ml per
litre, and there is no rule that the by-product must be solid at room
temperature.
What's much more important is that in each of the cases above, the test
batch produced a good "split" -- the glycerine separated and
settled to the bottom, and, if they'd followed the directions carefully,
the rest would have been good biodiesel, needing no more than settling
and washing. It "worked", it's just fine, move on to bigger
and better things!
How much
glycerine?
The rule of thumb is 79
millilitres of glycerine for every litre of oil used (7.9%). And crude
glycerine is not solid at room temperature. But the so-called "glycerine
layer" is not just glycerine, it's a variable mixture of glycerine,
soaps, excess methanol, and the catalyst (lye).
The quantity varies according to the oil used (more with heavily-used
oil), the process used (less with the acid-base
two-stage method), the amount of excess methanol used (most
of the excess methanol ends up in the by-product layer).
Why isn't
it solid?
It's mainly the soaps
combined with the glycerine that can cause it to solidify. Soaps made
from saturated fats such as stearin are harder than those made from
unsaturated fats such as olein, so the type of oil used makes a
difference (see How
much methanol? for compositions of different oils).
More important is how much soap there is -- the more soap, the more
likely the by-product layer will solidify, no matter what oil you used.
Other factors:
- Excess methanol makes
the by-product layer thinner
- Too much lye creates
excess soap
- Using potassium
hydroxide (KOH) as the catalyst instead of sodium hydroxide (NaOH)
produces a liquid by-product that won't solidify. (Using
KOH.)
See: Glycerin
PET
bottle mixers
Despite claims to the
contrary, this is not a good way of making biodiesel. It is useful if
you're doing a demonstration, though it makes it look simpler than it
really is.
PET bottles are transparent plastic soft-drinks bottles, generally 1/2
or 1 or 2 litres. "PET" stands for Polyethylene Terephthalate
-- see below, Identifying plastics
Here's how the PET bottle fans do it: warm up the oil (or maybe not),
funnel it into the PET bottle, add the (pre-mixed) sodium methoxide,
screw on the cap, shake the bottle up and down 10 times or maybe 40
times, leave it for an hour, and it's done.
Well, maybe. Or maybe not -- it's a good recipe for an incomplete
reaction.
For test batches, here's a better way: warm the oil to 55 deg C (131 deg
F). Pour it into the PET bottle, add the methoxide, screw on the cap,
shake vigorously up and down 40 times or more. Then stand the bottle in
a bath of hot water to maintain the temperature -- keep the water in the
bath at around 55 deg C. Shake again every 5 minutes, for up to two
hours, maintaining the heat all the time. This will have a much better
chance of not leaving unreacted and partly reacted material in the mix.
See below: How the process works.
Better still, make one of these instead: Test-batch
mini-processor
Biodieseler Greg Yohn, who developed this technique in the first place,
along with Steve Spence and other seasoned biodieselers who've used it
for years, all say it's useful ONLY for demonstration purposes, and so
do we. Use new oil, and best do it in private first so you don't risk
making a fool of yourself. The value of the "method" for
demonstrations is that it doesn't matter much if the reaction is
incomplete and leaves unreacted material in the mix, just as long as you
get separation and the glycerine drops out -- you're demonstrating the
process, NOT making fuel, and in a transparent PET bottle you can
clearly see it all happening. And it's non-messy.
Viscosity
testing
Viscosity levels are a
comparative indicator of biodiesel quality. Unfortunately, and despite
claims to the contrary, that's all they are, a comparative indicator:
this batch is better than that batch. Even at the laboratory or
industrial level, viscosity testing alone cannot tell you if the process
has gone far enough before reaching equilibrium and that there are not
unacceptably high levels of harmful unreacted and partly reacted
materials in your fuel.
Unconverted monoglycerides (MGs) and diglycerides (DGs) are fuel
contaminants that can cause injector coking and engine damage. MGs and
DGs are very similar in viscosity to biodiesel and stay in solution with
it after an incomplete reaction, they can't be washed out. The allowed
maximums are low: less than 1% for DGs and less than half that for MGs.
Viscosity tests might get you within 5% accuracy, not nearly close
enough for a useful quality check.
The same goes for density -- specific gravity measures (SG). Even both
viscosity and density together can't assure you that the reaction has
gone far enough towards completion. About the only sure way to know that
is with a Gas Chromatograph or expensive laboratory tests which few
biodieselers can afford. Short of a GC the best indicator of a completed
reaction is the wash -- easy washing and a crystal-clear product. See Quality
tests. See below: How the
process works.
Nonetheless, viscosity can be a useful indicator, especially with test
batches. You can check viscosity with a 100 ml pipette and a stopwatch
-- time exactly how long it takes 100 ml of your fuel to empty from the
pipette. Or use a viscosity
meter. Excess methanol in the fuel will render the results
meaningless, so you must wash the biodiesel first. Measure some
petro-diesel for a comparison. Remember that viscosity is sensitive to
temperature -- try it at two or three different temperatures. See the
various quality-specification tables here
for some guidance.
Aleks Kac offered this advice to a Biofuel mailing list member
struggling to get his home furnace working with
biodiesel:
Kinematic viscosity is
measured in "Stokes". You cannot measure it at home without
a viscosimeter. There is a comparative way, though. Take a liquid with
a known viscosity value (petroleum heating oil, look the value up in
an engineering manual) and let a known volume flow through an
upside-down plastic water bottle with a drinking straw glued in a hole
in the screw top. Stop the time with a stopwatch. Do the same with
your biodiesel (same volume) and compare the results. Generally a
smaller diameter straw will produce more accurate results. If the time
of your sample is 1.5 the time of your control sample (petroleum oil),
this means its viscosity is roughly 1.6-ish that of the control
sample. -- Aleks
Kac, 6 Nov 2001
You can use two straws,
the second one to let air in for a smooth flow of oil -- see how we do
it with small test batches of methoxide: Adding
the methoxide
Measure specific gravity (SG) by weighing a specific volume of the fuel.
Remember that volume is also sensitive to temperature. A litre should
weigh about 880 g at 15.5 deg C.
See Quality
testing
How
the process works
What is meant by
"completion" and "equilibrium"?
First, vegetable or animal fats and oils are triglycerides (TGs),
composed of three chains of fatty acids bound by a glycerine molecule
(see diagram in next section below).
Triglycerides are esters. Esters are acids, such as fatty acids,
combined with an alcohol, and glycerine (glycerol) is a heavy alcohol.
The transesterification process converts triglyceride esters into alkyl
esters (biodiesel) by means of a catalyst (lye) and an alcohol reagent,
usually methanol, which yields methyl esters biodiesel -- the methanol
replaces the glycerine.
In transesterification the triglyceride molecule is broken into three
separate methyl ester molecules plus glycerine as a by-product. The lye
catalyst breaks the bond holding the fatty acid chains to the glycerine,
the glycerine falls away, the fatty acid chains then bond with the
methanol.
It happens in three stages (this has nothing to do with the single-stage
or two-stage processes). First, one fatty acid chain is broken off the
triglyceride molecule and bonds with methanol to form a methyl ester
molecule, leaving a diglyceride molecule (DG) -- two chains of fatty
acids bound by glycerine. Then a fatty acid chain is broken off the
diglyceride molecule and bonded with methanol to form another methyl
ester molecule, leaving a monoglyceride molecule (MG). Finally the
monoglycerides are converted to methyl esters -- completion.
The problem is that the process can run out of reagent or catalyst
before it gets that far, or agitation, temperature or processing time
may be inadequate.
The result is some unconverted or partly converted material remaining in
the biodiesel. Well, so what if the process isn't completed? SVO
(straight vegetable oil) is a good fuel anyway, so what's it matter if
some of it is unreacted? But it's not just unreacted material that's the
problem so much as the partly-reacted stuff. Diglycerides and
monoglycerides are bad things to put in your diesel. Diglycerides don't
burn well and lead to coking problems, monoglycerides can lead to
corrosion and other problems -- bad fuel.
"The level of glycerol, mono- and diglycerides at levels of 0.1% (a
factor of 1/1000 or less of the main ester components) or lower appears
necessary for optimum engine performance." (International
Conference on Standardization and Analysis of Biodiesel, Session 2,
"Interaction Between Engine and Fuel", Vienna, November 6-7,
1995 -- to be published.) -- From "Analytical Methodologies for the
Determination of Biodiesel Ester Purity -- Determination of Total Methyl
Esters", NBB Contract #:520320-l, Richard W. Heiden, Ph.D., R. W.
Heiden Associates, February 27, 1996
http://www.biodiesel.org/resources/reportsdatabase/
reports/gen/19960227_gen-221.pdf
So, either don't process it at all and use SVO (which can have its own
problems), or process it PROPERLY.
In fact the process never reaches 100% completion, it always reaches
equilibrium first, so there will always be some unreacted glycerides
left. The various national
biodiesel standards stipulate just how much is allowable,
and it's not very much: diglycerides range from less than 0.4% to less
than 0.1% by mass, monoglycerides less than 0.8% by mass.
The first part of the process happens rapidly, which is why some people
think it only needs a few shakes and that's it. Not so. If it takes X
minutes to convert half the TGs to DGs, it takes almost as long, another
X minutes, to convert half the remaining TGs, then a further X minutes
for the remaining half, and so on. So the process goes more and more
slowly, and never quite arrives -- there's always half left. Finally
comes a point when the remaining half is insignificant, and, indeed,
within the limitations set by the various quality standards. But it's
very easy to fall short of that point and end up with nasties in your
beautiful clean eco-friendly nice-smelling home-brewed fuel, and in your
motor.
See Kinetics
of Palm Oil Transesterification in a Batch Reactor,
by D. Darnoko and Munir Cheryan, University of Illinois, for what
actually happens during the biodiesel process reaction. (Acrobat file,
72Kb)
You CAN make high-quality biodiesel, all it takes is a little care. On
analysis, biodiesel made by home-brewers with no qualifications and no
special equipment using the methods detailed at this website has proved
the equal of professionally made commercial fuel. Professional mechanics
checking their motors have been amazed by the lack of wear and
corrosion. You can do it too. See Quality
Message to the Biofuel
mailing list from a biodiesel dealer in the US:
"Currently I resell
commercial manufactured biodiesel in Atlanta, GA. Over the past two
years I have seen the quality of this fuel vary greatly. Funny thing
about the "commercially manufactured" biodiesel... One of
the big arguments against backyard biodiesel (from industry folks) is
quality, yet every batch that I have made, and every batch I have seen
by a homebrew biodiesel maker has been much better than the
"fuel" I am reselling. Individuals with small-scale setups
seem to really care, take their time, and craft their fuel... after
all, most are using it in their own cars, not selling to the boiler
fuel market."
For beginners, start off
with good practices: follow the instructions carefully, be meticulous
with your titration, make sure your measurements are as accurate as you
can make them. Learn as much as you can. You'll soon get a feel for it,
and then, once you're familiar with the process in all its aspects,
you'll be able to decide what's best for you in your situation, with
your source of oil, on your budget, and just where you can relax a
little and take calculated shortcuts, based on your own experience.
NOTE:
It's a common misconception that biodiesel has lower viscosity than SVO
or WVO because the transesterification process shortens the carbon chain
length of the fatty acid molecules in the oil.
This is not so. The biodiesel molecule is indeed smaller and less
complex. Transesterification converts the triple-chain triglyceride
vegetable oil molecule to three single-chain methyl ester molecules, but
the chain lengths of the fatty acids themselves remain the same. The
fatty acid composition of biodiesel depends on the feedstock and is not
changed by transesterification.
What
are Free Fatty Acids?
Oil
molecule -- graphic by Jeff Welter
|
Again, vegetable or animal oils and fats -- what we make biodiesel from
-- are triglycerides, composed of three chains of fatty acids bound by a
glycerine molecule.
Free fatty acids (FFAs) are fatty acids that have become separated from
the triglycerides, leaving diglycerides, monoglycerides and free
glycerine. This is caused by heat, water in the foods cooked in the oil,
or oxidation. The hotter the oil gets and the longer it's cooked, the
more FFAs it will contain.
As glycerine is an alcohol (glycerol), a fatty acid attached to it (a
glyceride) forms an ester. A "transesterification" is the
conversion (switching) of one ester into another -- a glyceride ester
into an alkyl ester in the case of biodiesel, where methanol (or
ethanol) replaces the glycerine.
An "esterification" is the conversion of a non-ester into an
ester. FFAs are non-esters. FFAs are converted into esters by acid
esterification in the first stage of the two-stage acid-base biodiesel
process, but cannot be converted by the more common single-stage base
transesterification process. Here the FFAs must be removed from the
process, or they will dissolve in the biodiesel being formed, yielding
an acidic, poor-quality fuel that will not meet quality standards.
In transesterification, extra lye is used to neutralize the FFA content
of the oil, turning it into soaps. These soaps drop out of the process
as a by-product, joining the so-called "glycerine layer" at
the bottom -- often more of a "soap layer" as it may contain
more soap than glycerine.
The basic lye quantity used in transesterification acts as a catalyst,
not a neutralizer. Lye attacks ester bonds, breaking the bond, and the
alcohol drops off, leaving an open-ended fatty acid chain. With
glycerides the alcohol that drops off is glycerine. The affinity of the
replacement methanol or ethanol for the resulting open bond is strong
enough to prevent the glycerine reattaching to the fatty acid.
This is also why it is critical that a minimal amount of lye is used, as
lye will continue to attack ester bonds, even those of biodiesel. Too
much lye will break the biodiesel ester bonds; some of the broken bonds
will mate with the lye and form excess soap, and others will match up
with a water molecule to form FFAs, which dissolve back into the
biodiesel. It is this excessive formation of FFAs that the "acid
number" in the US ASTM and other national quality standards refers
to.
While it is unavoidable that some FFAs are formed by biodiesel ester
bonds being broken, excess lye increases the proportion.
According to the Fuel
Injection Equipment (FIE) Manufacturers (Delphi, Stanadyne,
Denso, Bosch), FFAs can corrode fuel injection equipment, cause filter
plugging and the build-up of sediments on fuel injection parts.
-- With information from Todd Swearingen of Appal Energy, and DieselNet/Ecopoint
Inc.
http://www.dieselnet.com/
Which
method to use?
Three main choices, all of
them here:
Single-stage
base -- Mike Pelly;
Two-stage
base-base -- Aleks Kac;
Two-stage
acid-base "Foolproof" method -- Aleks Kac.
What's the difference?
The single-stage base method is the place to start.
The two-stage processes are advanced methods, not for novices -- learn
the basics thoroughly first. Single-stage base is the original method,
and still the most widely used, tried and trusted. It's the simplest
method, especially for new oils which don't need titration.
A lot of beginners want to use WVO but they're put off by the titration,
thinking it's too complex. Actually it's simple enough -- just follow
the directions. However, the
single-stage process produces more and more uncertain results the higher
the FFA content gets in WVO, with lower production levels even when it
works well.
The two-stage base-base method avoids the need for titration and
produces good results even with higher FFA levels. It's the
method-of-choice for animal fats.
Growing numbers of biodieselers are now turning to the
"Foolproof" two-stage acid-base method, especially with high-FFA
oils. Here are some of the reasons:
- Less base catalyst
needed.
- Less soap production.
- Higher conversion rates
as a result of less soap formation.
- Less emulsion formation
in the wash.
- Less loss of fuel in
the wash as a result of emulsion formation.
- Less wash water as a
result of less soap formation.
- Less neutralizing acid
needed for the wash.
- Less acid needed to
neutralize base during glycerine recovery.
- High-quality product.
The negatives:
- A little extra
processing time.
Even with higher-FFA oils
the production rate should be 100% or more by volume (biodiesel has a
lower density than the original oil).
In fact the same advantages apply to new oil, although to a lesser
extent. Many biodieselers who turn to the "Foolproof" method
for high-FFA oils soon make it their method-of-choice for all oils.
Here's some advice from Aleks Kac on using the Foolproof method:
"Stick to the recipe, to the letter. There's two years of trial and
error research in this. Do not change, simplify or speed up anything. It
will take care of all sorts veg fats, even heavily used. The 'solid'
portion must be reduced to less than 50% because of the much lower
acid-stage temperature. Animal fats content is best at less than 25% for
pork or chicken and less than 10% for bovine. These fats at greater
concentration should be processed with the two-stage
base-base method."
Still, if your oil is quite good and usually titrates at 3 ml or less,
you might well be satisfied with the single-stage process.
Why
can't I start with the Foolproof method?
It says at the top of the Foolproof
acid-base process page: "NOTE: The two-stage biodiesel
processes are advanced methods, not for novices -- learn the basics
thoroughly first. The single-stage base method is the place to start.
Start here."
Here being here:
Where do I
start?
But novices sometimes take no notice and plow ahead with the acid-base
process anyway. Sometimes they also take no notice of the advice on Test
batches: "Whenever you're trying a new method, it's always a good
idea to make small test batches of a litre or less first to familiarize
yourself with the process before moving on to bigger batches."
All too often it results in a highly discouraging 40 gallons of glop.
"It doesn't work!" they wail. It works. It also works to heed
the advice of the many who've gone before you. "DO AS THEY
RECOMMEND and bio will come," said a recent novice at the Biofuel
list.
"But it doesn't sound difficult at all, in fact it sounds easier
since there is no need to titrate. So what am I missing?"
Quite a lot. First, we think it's worthwhile titrating your oil anyway
-- the more you know about it the better, and titration will tell you
more than anything else.
Avoiding learning how to do titration is a VERY BAD reason for using the
acid-base process! Titration is a basic skill needed for making
biodiesel. Anyway, it's easy enough:
Basic titration
Better titration
Second, it's not just a question of whether or not novices can get it to
work. Some do -- mostly they seem to get away with it. But that's a pity
just the same, because it's doubtful that they'll get the best out of
it. The acid-base method is very flexible, people bend it and twist it
every which way in adapting it to their preferences and needs, and you
just can't do that unless you have a good overall feel of the process as
a whole, not just of this one method.
Another novice at the Biofuel list said he had good chemistry knowledge
so he could afford to skip the "newbie" stuff. But:
"Well, my first test batch is done, and the end result is less than
spectacular..."
Suddenly he found himself facing an ocean of variables, and since he had
no "feel" for what was supposed to be happening he had no way
of finding out what he'd done wrong. Several things, as it turned out.
One answer:
"This is why this
isn't a good place to start. If you were more experienced you might
have had a better idea of how to translate the mixing instructions for
a full-sized batch to the small scale you're using. Maybe it
translates direct, maybe not -- I don't know how fast your drill stirs
it, nor what rate of agitation it gets with that paint stirrer, but,
comparatively, neither do you, and that makes it difficult for you.
Starting instead with single-stage base and fresh oil, you begin with
fewer variables and they're less critical, and it's a logical
progression from there. Now you're facing too many variables and you
don't have the experience to assess them -- and you're more likely to
make mistakes anyway because you lack a basis of comparison."
He relented and went back
to the beginning, processing 1 litre of fresh oil by the single-stage
base method. "Despite people saying that in different ways, I
hadn't heard it until now. All I was hearing was that you have to start
with the single-stage process then graduate to the foolproof process.
This just seemed like trying to learn something one way then do it in a
completely different way, which doesn't make sense."
"... so what am I missing?"
One answer:
"If everything goes
exactly right, nothing. The problem is when something doesn't, and you
have NO IDEA what's going on 'cuz you've never seen all the possible
quirks of even the basic process.
"...What's all this white stuff?... Nothing seems to be
happening... There's this weird layer, and I'm wondering if it's
biodiesel... etc etc."
Another answer:
"I can mention that
from a beginner's perspective, starting with pure vegetable oil and
single-stage base is a really valuable learning process. It gives you
an idea of what outcomes look like, and the shake
test provides feedback on the quality of the process. I am still
playing around with variables (processing time, %lye, %methanol, etc.)
using pure vegetable oil, before moving to waste vegetable oil.
"As many people continue to emphasize, process quality is really
important, and that seems to be best learned in small steps. My game
plan is also to get to the 2-stage acid/base with waste vegetable oil,
but I still have a lot to learn before getting there. I think what the
experienced folks are cautioning is that troubleshooting a more
complex process is extremely difficult (and perhaps frustrating) if
you don't have a solid grounding in the basics. I hope this helps.
Good luck with your experiments!"
The response:
"After reading the
links to the discussion groups that you attached I see that the basics
is to get a feel of the process and see the correct colors and
textures using the process that has the best chance of success, then
use that as a baseline for future mini batches using WVO. Then
progress to larger scaled processing, then full scaled process, and
then finally graduate to the two-stage method, with knowledge that the
resultant biodiesel is clean and pure enough to run through my beloved
TDI."
Right!
After all, it's you and your kitchen vs ExxonMobil -- and the kitchen
wins! Not only that but it's better fuel! Isn't that good enough? Why
take short-cuts? Just do it right.
Biodiesel
in gasoline engines
Biodiesel can also be used
in gasoline (spark-ignition) engines, but only as an additive. People
have had good results with it, but it's still experimental, there are no
guarantees.
Biodiesel
in 2-stroke gasoline engines:
Many people have used biodiesel as replacement 2T lube oil in two-stroke
gasoline (petrol) engines. Biofuel
mailing list member Martin
R. of
Australia uses it in his chain saw, at a mix of 1 to 20 with gasoline.
"It works fine," he says. "After using the saw for 2.5
hours in one go on dead Australian hardwood with no hiccups I was very
impressed to say the least."
List member Franklin
Del Rosario,
biodiesel maker in Manila in the Philippines, uses it in his motorcycle:
Biodiesel was first used
as 2T replacement in the Philippines by a group who added a
viscosifier and sold it as 2T replacement for tricycle taxi use at
several places outside Manila.
The biodiesel, pure or blended with oil additive, performed better
than 2T mineral oil as lube oil, according to the testimony of the
tricycle drivers who tried it. Their motors ran smoother and quieter
than before, the engine block did not heat up, others experienced
better engine performance, no more carbon build-up inside the cylinder
and muffler, no oil dripping on the exhaust pipe and most of all no
more smoke.
I use B100 biodiesel as lube oil for my 1983 Yamaha 125 cc two-stroke
motorcycle instead of Shell 2T lube oil which I previously used. I mix
biodiesel with gasoline at a ratio of 1:20. I replaced all rubber fuel
lines, carburetor gasket etc., cleaned the fuel tank, and changed to a
new spark plug. The motor is running smoothly on city driving, the
smoke has improved because no more 2T lube oil.
The Yamaha was retired before because of the smoke it produces using
mineral 2T oil. With biodiesel as a lube oil it runs better than
before and with less smoke at the exhaust pipe. I took my motorcycle
to the emission test center and it passed the gas standard limit with
better results than a new motorcycle which used mineral 2T oil.
The results were as follows:
Data reading
CO2 -- 2.90%
CO -- 3.39%
HC -- 5396ppm
O2 -- 13.29%
AFVR -- 20.95
LAMBDA -- 1.44
Standard Gas Limit
CO2 -- 20.00%
CO -- 6.00%
HC -- 10000PPM
LAMBDA -- 1.00 +/- 0.99
Result -- Passed
The main property of the lube oil is not the viscosity but rather the
ability to mix well with gasoline while maintaining its lubricity
(ability to provide a film of oil at the metal surface) to prevent
wear of the crank shaft, connecting rod & piston pin and cylinder
wall. Biodiesel's lubricity is enough to do the job and it can be
improved by adding compatible oil additives to increase the film
strength of of the biodiesel.
BD is a fuel with 11% of oxygen by weight, this property alone can
help gasoline to burn well reducing smoke emission and leaving no
trace of oil at exhaust muffler. -- Franklin Del Rosario, January 2004
Biodiesel
in 4-stroke gasoline engines:
List member Gregg Davidson wrote:
In June of this year, I
posed a question about "Biodiesel in gasoline engines". I
received a reply from list member JC in Taipei. He advised me that he
had mixed biodiesel with gasoline for his car, using no more than a
15% mix. I have had success following his example and had no engine
problems. One of the three vehicles I tested this in is a Chrysler
Town & Country mini van with a 3.3 L V-6 Flex Fuel Engine, the
other two are Jeep Grand Cherokees with 4.0 L I-6 engines. Even though
the van can run on E-85, I do not use ethanol blends because E-85 is
not available in my home state of Georgia. -- Gregg Davidson, October
2004
< |
