Washing Biodiesel

Hi folks,
so here's a bubblewashing 101 I've been working on for my zine. I 
haven't found much comprehensive info anywhere that really addresses 
the ins and outs of emulsification, and I've gotten to see lots of 
emulsification since a bunch of us here have been working with some 
horrid oil that gets incomplete reactions and causes problems washing 
(now the problems are no more as I am back to using restaurant oil 
for the moment till I figure this one out) 
anyway I think this monster article thing needs graphics BAD to 
really get some of the points across, but here it is. feel free to 
pass it around to whomever, just let me know if it gets posted on the 
internet or anything like that.


Bubblewashing 101

Maria "Mark' Alovert
Wrench@tinkersworkshop.org

There are a number of water-soluble impurities left in biodiesel 
after the reaction and initial settling is complete. They are found 
in everyone's homebrew biodiesel, regardless of the quality of your 
reaction or of your oil. The contaminants include (primarily) soaps, 
a small amount of leftover methanol, a small amount of leftover lye, 
and some free glycerine.  Washing the fuel with water removes these 
nasties. Washing  also has two additional advantages- it stops the 
very slow remaining reaction that sometimes occurs (in unwashed fuel 
you can sometimes see glycerine settle out over a period of weeks or 
months, and washing and removing methanol or lye will stop this) and 
it provides you with some quality control feedback. 

Washing is commonly done using the U of Idaho bubblewash method or 
the `mist wash' system developed by members of the infopop biodiesel 
discussion group. Less common is bulk washing. All washes can be done 
after the initial settling time (18 hours) and after draining the 
glycerine layer, but you can get better results by waiting at least 
36-48 hours or longer before the first wash, so that soaps and free 
glycerine have more of a chance to settle out.
 
Bubble washing involves gently adding 1/3 water to 2/3 biodiesel, 
adding a cheap aquarium air stone and air pump setup to the water 
portion (water sinks to the bottom and biodiesel will float on top) 
and bubbling air through the water. The air bubbles allow a sort of 
indirect agitation of the two fluids- they pick up a tiny amount of 
water and gently carry it through the biodiesel, picking up soaps and 
other contaminants. When the bubble bursts at the surface it drops 
the water which picks up more of the soaps and contaminants on it's 
way back down. After about 6 hours of this (low-wattage) `washing', 
the air is stopped, then the water is drained, more fresh water is 
added, and the process repeats. These `wash water changes' are 
repeated about 3 times on average- until the water measures the same 
pH of your tap water, and is perfectly clear. The wash water can be 
reused a number of times to wash succeeding batches (explained below 
as `Counter Current'). 

Advantages are that bubblewashing uses less water than others, and 
that it uses very cheap equipment (aquarium air pump is under $10 
new, very easy to find `used' at flea markets and thrift stores, and 
uses 3 watts of electricity). It is easy to leave the wash unattended 
(especially with a cheap timer to make the process even more 
userfriendly). 
The disadvantage: if you have made poor quality biodiesel, or are 
washing a very small batch, bubblewashing can agitate the water and 
the biodiesel too vigorously- causing emulsification of the two 
liquids. Emulsification is the quintessential `wash problem'- but it 
is also a form of quality testing and  feedback  on your process. 
Once you know what causes it, it is also easily avoided.

Mist washing was developed as a way to address emulsification issues. 
It uses more water and more complicated equipment. It also masks 
quality problems- you can get a `good wash' but will have much less 
of an idea of what you've made. This system uses a very fine mist 
head from a garden supply store, suspended over a container of 
biodiesel, with a way to drain the water after it falls through the 
fuel.  The mist stirs up the fuel less than in bubblewashing, and 
removes soaps gradually. The gentler agitation gives less of an 
opportunity for the soaps and any mono and diglycerides to form 
emulsion. Many mist-users also do bubblewashing as a final step- 
after the soaps are gone the mg and dg alone won't usually emulsify 
the fuel and water as much, so bubbling after misting 
usually `works'.  
The disadvantages of misting are higher water consumption, more 
complicated equipment, and the masking of potential problems- 
Unfortunately, the mg and dg will still be there regardless of 
washing method- they are not watersoluble and won't wash out- and in 
misting, you won't know if your biodiesel was the result of an 
incomplete reaction  you're doing some other form of routine testing.

The last method, `bulk washing' is not usually used by homebrewers 
now. It involves mixing equal amounts of water and biodiesel, 
agitating gently, letting separate, draining water, and repeating 
again a number of times. It is more work, uses more water, and isn't 
an automated unattended process like bubblewashing.

I am going to focus on bubblewashing for the rest of the article:

Equipment: washing container, valve (optional) or siphon to drain 
water from under the biodiesel, aquarium air pump of at least 3 watts 
(if washing a large- 25 gallon or bigger- batch), aquarium air stone- 
the bigger 6" ones sink to the bottom easily, the small ones will 
need to be weighted down with some washers, and some airline tubing. 
Optional: cheap timer to turn off air bubbler automatically after 6 
hours. 

The container:
You can do washing in your processor or in a separate container. I 
use a 55gallon closed head drum which has been turned upside down 
(bungs are now on the bottom) and the end without bungs is cut open 
(sawzall works great for this). I then attach a ball valve to the 3 / 
4"  bung and mount the whole thing on a stand (or some cinderblocks, 
or just even set it on top of a two fullsize American milk crates 
stacked together to a bucket-clearing height) . Using a separate 
washing container rather than your processor frees up the processor 
for making and settling the next batch, keeps any residual water out 
of your next batch, and keeps any glycerine that didn't drain well 
from contaminating the wash water (which can lead to emulsification)  

The rest of the equipment is fairly straightforward.

Add 1/3 water to a container with 2/3 biodiesel. Dribble the water 
into the tank gently- don't spray it hard from a hose under high 
pressure.
Drop a bubblestone into the water, turn on the air pump, and let 
bubble for a few hours. The air pump should be mounted (I use thick 
wire and sort of tie it to the outside edge of the drum) above the 
level of the water so that if the power goes off there's no chance of 
liquid siphoning back into the pump.
The exact amount of time you bubble is not crucial. We use 6 hours 
for the first wash, and more for the others. The water will become 
somewhat saturated in relation to the soaps content of the biodiesel, 
and will stop taking in any more soaps/contaminants.  At this point 
(6 hours or overnight), turn off the bubbler. It is useful to let 
everything sit for an hour or several  so that the water and 
biodiesel separate well (optional). This is where the timer is useful-
 to turn bubbler off and let settle before you get there to drain it 
You will then drain the water as much as possible, add more water, 
turn on the bubbler. This wash water change and overnight bubbling is 
repeated three or more times, until the wash water is clear and 
measures the same pH as your tap water. 


Emulsification:  If you have made less-than-perfect biodiesel, the 
dreaded "emulsification" is most likely to rear it's ugly head in the 
first wash:

An explanation of emulsion: 
 If you mix water with some clean washed 100% biodiesel, the two 
liquids will separate quickly into two layers- oily biodiesel on top 
and water on the bottom, with nothing in between (try this with a 
commercial biodiesel sample- shake it up in a jar with an equal 
amount of water).  Well-made and well-washed biodiesel will act just 
like salad oil and water if they were mixed this way. Unfortunately, 
with the soaps (or free glycerine, or mono and diglycerides) present 
in unwashed fuel, the emulsification factor comes in, and if there is 
a lot of these soaps or other emulsifiers, the liquids don't separate 
easily.  This is more like what would happen if you mixed salad oil, 
water, and lots of dish soap, and shook it up in a jar- there would 
be more of a homogenous mixture instead of two layers.

Emulsification will also happen if these soaps/mg/dg are present, and 
too much/too vigorous agitation of the mixture occurs. This is the 
reason for the mist washing method- to take vigorous agitation out of 
the picture. To me it seems that this masks the actual problem- which 
isn't agitation, it is fuel quality. I would rather produce fuel that 
contains less soaps and less emulsifying mg and dg instead of 
decreasing the agitation- because the amount of agitation produced by 
standard bubbledrying isn't so strong that an average homebrew fuel 
shouldn't be able to handle it without emulsifying.  Two-stage acid-
base process, while a little complicated for beginners, produces much 
much less soap than the usual single-stage, and gives a consistently 
easier wash. Try it when you get experience handling the singlestage.

Mono and diglycerides sidebar:
The additional factor in `wash problems' is the presence of mono- and 
diglycerides, which result from an incomplete reaction in a less-than-
perfect batch of homebrew biodiesel.  Mg and Dg also emulsify oils 
and water, are soluble in biodiesel, and do not wash out.  Mg and dg 
are common ingredients in junk food- as emulsifiers in cheap ice 
cream and sauces and the like. They are unwanted guests in homebrew 
biodiesel, however. Mono and diglycerides can cause injector clogging 
and corrode some metals when burned as part of poor quality fuel. 
When making fuel, it's usually not hard to correct for the appearance 
of mg and dg- use the right amount of lye, higher temperature (I use 
130 F during reaction), enough agitation (go longer if in doubt), 
more methanol. Different oils need differing amounts of the above to 
produce good biodiesel. Skimping on any of the above factors, or 
inaccuracy in measurement, is the first thing to look for if you 
suspect poor conversion in your fuel.
Home testing for mg and dg:
 One test for mono and diglycerides is to re-process a liter of your 
fuel, and to see if more glycerine drops out. 1. Take a liter of 
biodiesel (prior to washing, or after washing, but not in the middle 
of the washing process somewhere as there will be water present that 
can disrupt the process and yield soap). 2. Re-process the liter of 
biodiesel as though it were new oil instead of biodiesel- 3.5 g lye, 
200 ml of methanol, and 3. see if more glycerine drops out, 
indicating that some mg and dg was present and finished converting to 
glycerine in the reprocess test. Be aware that some of 
the `glycerine' dropping out will be watered down by the methanol- so 
the total glycerine quantity isn't as large as it may seem. Good fuel 
won't drop any glycerine at all- again, try this test on some 
commercial fuel. (end sidebar)


What emulsification looks like:
 
Normal, no emulsion: Biodiesel made from WVO should stay relatively 
dark or golden-brown even during the first step of washing. A sample 
will be hazy/cloudy.  It could turn slightly orange and lighter, and 
this is still normal. The orange color is from some small amount of 
emulsification occurring- there is some water emulsified into the 
biodiesel layer (probably held there by soaps), but it will probably 
go away in the next wash water change or two. 

Not-so-normal: emulsification:
Emulsification can happen two ways: either all the way through an 
entire container of washing fuel and water (really bad 
emulsification) or in a less stubborn form- an emulsified middle 
layer in between the biodiesel and the water.  It's been described as 
looking like cream or mayonnaise or `chicken soup' (funny how it's 
all food references, must be the fryer oil exhaust smell that makes 
biodieselers hungry)


1. all the way through (bad):  If you look into the top of your wash 
tank and instead of brown biodiesel see the contents turn a creamy 
whitish-yellow, opaque color, with a thicker viscosity (like a thin 
milkshake or very runny pancake batter instead of the biodiesel 
viscosity)- then you've got emulsion all the way through your 
container.

2. emulsified middle layer (more common, less of a problem): If on 
the other hand the wash has formed an emulsion layer between the 
water layer and the biodiesel layer, you will come across it when you 
drain your first wash water. Normal `first wash' water should be an 
opaque white color, and of ordinary water-like viscosity, looking 
just like milk sometimes. If you hit some emulsion the water will 
look more creamy (slightly yellow) color and can sometimes also 
change viscosity- the emulsion can be more like that runny milkshake 
rather than like milk or water. 


Dealing with it:

If you`ve gotten yourself the middle layer variety of emulsification- 
don't despair- its somewhat normal and chances are that you can just 
let it sit and break on it's own.
If you're in a hurry, drain the wash water until you hit emulsion, 
then divert the emulsion into another bucket or container. Then wash 
the rest of the biodiesel. Separating emulsion from the rest of the 
biodiesel lets you keep washing the majority of the biodiesel in the 
wash tank while the emulsion sits and separates. If it does not 
separate by itself in time (half a day to a couple of weeks) you can 
then experiment with breaking emulsion by using salt, heat, or acid, 
and you will not compromise the quality of the rest of your washed 
batch with the salt or acid. Not removing the emulsion and continuing 
to just try and wash it out doesn't seem to work as well as removing 
the emulsion and it's associated soaps/emulsifiers- you'd have to 
wash the whole batch longer if you just recirculated the emulsified 
matter.
 
Aggregate tank: Since I have the space and make a lot of fuel, I now 
use an `emulsion aggregate tank' to which I add any of the middle-
layer emulsion I've drained out of the wash- it's a small drum (a 
15gallon would work well for someone washing in a 55gallon drum) with 
a drain at the bottom. Most emulsion will break or at least release a 
lot of water very rapidly. I drain out the released water 
periodically. Eventually the aggregate tank builds up 
enough `released' biodiesel that I pump that fuel into a wash tank 
and do a regular wash on it.  Before I had the aggregate tank I just 
stuck the emulsion into buckets or restaurant oil jugs, let it sit, 
and then pumped or siphoned off the biodiesel when it released 
(usually a day or so).  I had a lot of emulsion to deal with because 
I have for the past 6 months worked with some feedstock oil that 
produced exceptionally bad biodiesel. I got poor conversion no matter 
what factors I and other local biodieselers changed- but now that I 
am back to using regular restaurant fryer grease I rarely have much 
emulsion to deal with, so the tank isn't very necessary and may not 
be necessary for you either.


Severe Emulsion Formation:

If your whole batch of washing fuel turned light yellow and 
emulsified, you have a number of options:
In order of preference, try: time, gentle heat, salt, acid.

 Time: 
turn off the air bubbler and let it sit. It might `break' in a day or 
several. 
Drawback: time, and your wash tank being `tied up', which may or may 
not be a problem. 

Gentle heat:
heating an emulsion can break it, but it is somewhat dangerous if the 
heat approaches boiling point of water- someone told me that he once 
did this, accidentally overheated the water, the weight of the 
unbroken thick emulsion layer above the water layer held down all the 
steam that was forming, and the guy got a nasty steam explosion and 
blew the lid off the tank.  Obviously this can be done safely, now 
that you know how to avoid the unsafe method. 
Drawbacks: 1) the obvious danger above if you overheat it, and 2)the 
additional energy required, unless you're using solar hydronic with a 
heat exchanger 

Salt:
Adding salt will break an emulsion. The water and salt molecules have 
more affinity for each other than water and soaps do, and the water 
portion of the emulsion to drop the soaps and take on the salt 
instead. With no soaps bonding to the water, the oil (biodiesel) 
drops out of emulsion with that water.
Drawbacks:
1.Salt is a toxin depending on how you dispose of your wash water- ie 
don't just run this stuff over your garden!
2. I have found that if you salt, you will then need to use more wash 
water changes to get to the finished fuel point- ie where your wash 
water is clear and of the pH of your tap water.
My theory is that in between wash water changes, there is still quite 
a bit of water present in the biodiesel, and that this dissolved 
water holds the salt as well as the soaps.
While salt shouldn't dissolve in perfect biodiesel ordinarily, 
remember that during the intermediate steps in a wash the biodiesel 
layer contains a lot of water and other impurities.
     The salt then prevents subsequent washes from being as effective 
as normal. White color of wash water is caused by soaps, and if you 
salt out an emulsion you will find a water with very little white to 
it- and the next few washes will also have less white color to them 
than normal. 
3. Salt of course accelerates corrosion of metals. It seems very 
important to make sure that all the salt is washed out. 

Acid:
The University of Idaho and many published instructions about washing 
used to recommend using acid to help make washing easier.  People 
used to use it routinely in their first wash water- which acid to use 
doesn't matter very much, and household vinegar or citric acid was 
routinely used. 
You can also add acid to an emulsion and might see the emulsion 
break. However it isn't the best method to use as it can compromise 
the quality of your finished fuel:
 Acidifying a wash or an emulsion works by breaking up the soaps into 
their constituent parts: forming a salt and a free fatty acid. The 
ffa will end up in your biodiesel and is indistinguishable from it. 
Drawback: FFA content is a concern of the ASTM and other 
specifications for biodiesel.  FFA promotes some corrosion of metals. 
Raising the FFA level doesn't seem to be a good thing to do.
I also believe I've had a harder time washing fuel that I have 
acidified- same situation as salt- a harder time getting clear 
water/neutral pH for some reason- but I could be wrong on this one as 
I don't routinely do the acid treatment.
Some people recommend breaking really stubborn emulsion with acid, 
then drying the biodiesel (big energy input there), then re-reacting 
the biodiesel portion to remove the ffa (and to get a better reaction 
if mg and dg were a problem). The re-reaction will again form more 
soaps- using two-stage acid-base process in this case seems like a 
good idea. But really, really stubborn emulsion isn't all that 
common, and one of the other methods should fix it without resorting 
to acid. 

 
Counter-Current Wash Water Reuse:
If you have the space and do a fair amount of biodiesel production, 
it's good to know that you can reuse the wash water for subsequent 
batches, up to a point. This is called a counter current system.
The water needs to be changed because there is reached something like 
an equilibrium between the amount of soaps in the water and the 
amount of soaps in the biodiesel, where the water won't take in any 
more from that batch of fuel. But if you then saved this water and 
reused it on ANOTHER batch, one step BACK in the process (ie for fuel 
that's soapier than the first batch you used it for), it will again 
absorb more soaps. The example is, to reuse the water from Wash #2 in 
the Wash#1 step of the next batch you bubblewash. So you only need to 
dispose of wash 1's water, and only after it's already been reused 
for three other washes in previous batches. 
Drawback: keeping this much water sitting around isn't practical for 
everyone, given space limitations.
At our co-op, we don't have any room for any more storage of water.  
However we try and have two wash tanks `going' at once. The wash on 
one batch is started one day before the other. We change the wash 
water on both tanks at the same time- draining the water out of the 
older batch into the newer one. So even though we dispose of water 
before it's been reused as often as it could ideally be, we DO reuse 
it at least twice. We also leave the last wash water in the tank, 
remove the biodiesel by pumping it off the top, and save that last 
wash water for the first wash of the next batch.

Disposal:
It is theoretically possible to use biological filtration- a 
constructed wetlands graywater system- to break down the soaps in 
wash water, cleaning it enough to reuse it many times. I don't know 
of anyone actually doing this yet, though many biodieselers are 
talking about it. See the book "Building an Oasis With Graywater" 
available through RealGoods , for a discussion of a few homebuilt 
graywater systems- in brief, a graywater system uses water plants- 
cattails and water hyacinth or the like- along with gravel, sand,  
baffles, and microorganisms, to break down impurities in water. This 
can be done in a trio of drums or old bathtubs set on a staggered 
stand (so water flows slowly from one to the others by gravity) and 
planted with the above plants. Remember, you're only trying to get 
the water clean enough to absorb more impurities, not to necessarily 
clean it enough to water a food garden with it.
 Barring graywater filtration,  you can do several things: add acid 
to the water, checking the pH, until it is neutral, then sewer it. 
Most of us currently sewer the wash water, figuring that it is no 
worse than the soaps, detergents, shampoos, laundry soaps, bleach, 
fabric softeners, drain cleaner, shower scrubbers, tile polish, floor 
cleaner, car cleaners, tire cleaners, windshield bug deflector 
cleaner, ammonia, and host of other nasties flushed by the average 
American household. The soaps in the biodiesel wash water are fairly 
watered down- you won't get `suds' from your first wash- and the 
amount of lye and methanol left to be washed out has been shown to be 
fairly low (most of your excess lye and methanol ends up in the 
glycerine layer, not the biodiesel layer we wash). One improvement 
could be to do methanol recovery on the biodiesel, but it's usually 
not worth the energy to do it to the biodiesel (as opposed to 
glycerine) layer.  I'm also waiting for someone to try reverse 
osmosis to clean up their wash water, but the units are VERY 
expensive.

Drying Washed Biodiesel:
The current conventional wisdom on dry fuel is that if it is `clear'- 
referring to lack of haze, not color- it does not contain much water 
(in reality biodiesel will absorb some small and harmless amount of 
moisture from the atmosphere).
Washed fuel will often look slightly cloudy or hazy. If you let it 
sit it will eventually drop or evaporate it's water content. You can 
also heat it (I don't do this as the MSDS implies that biodiesel 
vapors are harmful) . I have been experimenting with bubbling air 
through it, and it yields clear fuel, but I am not certain that 
actually indicates dry fuel.

More importantly than these external drying aids, is the fact that 
the better-washed the fuel, the faster it will clear. If there aren't 
any soaps left to hold on to water, the water will settle out 
rapidly. I have made fuel that cleared almost immediately, and I 
think the factor that helped this was the number of times I washed 
it. So like with everything in biodiesel, this `drying' seems to be a 
form of `feedback' on the process- if it refuses to clear in a few 
days I usually give it another wash (using a little less water than 
usual). 

A test for water content is to weigh a sample, heat it until it 
passes the boiling point of water, then re-weigh to see how much 
water it has lost. Biodiesel contains 1200-1500 ppm water normally, 
this is the water absorbed from the atmosphere naturally.

http://journeytoforever.org/biodiesel_bubblewash.html

http://www.veggiepower.org.uk/page207a.htm

http://webconx.green-trust.org/2000/biofuel/biodiesel_2stage.htm

http://home.earthlink.net/~galiagante/house-biofuel.html