Just going back to Johns first post on here and the explaination of the different results between Sandy brae and Karl Fischer.
I am fortunate to have both types of equipment. I have done many tests with both for comparison, there is correlation and it is very accurate.
I cannot see how the Sandy Brae people can say any difference
is due to the SB not being able to detect chemically bonded water, if water was chemically bonded it wouldn't be water.
Surely they mean physically bonded water, ie dissolved water. And for this the SB is virtually as accurate as the KF.
There is an anomoly here which could explain the reason the results were so far apart.
No way will the SB result be readable within the timescale the instructions state. A typical test on water in biodiesel takes around 6 hours, most certainly NOT the 20 mins or so that the instructions state.
This was very confusing when I first used my SB until I realised the correct timescale by experiment. 'Drum' a forum member on here has experianced the same thing.
If your people in France John tried this then they may get similar results.
"The water must be in the form of free molecules or droplets and not be chemically or physically bound."
This is incorrect Tilly.
A recent thread on here where Graydon was involved talking to the SB people did indeed confirm that the SB does detect bound (dissolved) water.
Goodnes me Dgs,
You already have put me on ignore- Twice!
Ignore button engaged and locked!- Jan 22 2015
Thats better, He's now on my ignore list.- Oct 12 2014
If the quote is incorrect then you need to contact Sandy Brae Laboratories and tell them that they have incorrect information in their Sandy Brae Water Test Kit Instructions.
From the top of page 2 (third page down) of the instruction manual-
"The Sandy Brae Water Test Kit can be used to test for water content measured as Parts Per Million or as a percent (%) of volume. The test can be performed on any liquid which is not based on water
Any fluid can be tested as long as the two following requisites are met:
a. the water must be in the form of free molecules or droplets and not be chemically or physically bound;
b. the fluid should not attack the water vessel components which are anodized aluminum, viton gasket, and brass"
Perhaps there is a difference between bound and dissolved water. https://en.wikipedia.org/wiki/Bound_water
Water in vegetable oil and biodiesel exists in three phases. One is dissolved water, which is characterized by individual water molecules dispersed throughout the oil. Dissolved water in oil is similar to humidity in the air - the water is there, but it is too small to see because it is dispersed molecule-by-molecule. Thus SVO or biodiesel can hold a significant concentration of dissolved water and still appear clear. Most vegetable oils can easily hold 200 to 600 PPM of water (0.02 to 0.06 percent) in the dissolved state depending on the temperature and age of the oil. Used oils can hold three to five times more water in the dissolved state than new oil.
Oil becomes saturated with moisture when the amount of water exceeds the maximum concentration for it to remain dissolved. In this second phase, the water is suspended in the oil in microscopic droplets known as an emulsion. This is like fog on a cool, spring day. In that case, the amount of moisture in the air is greater than the saturation point, and a suspension of small droplets of moisture or fog results. In SVO or biodiesel, this “fog” is often referred to as haze and the fuel is called 'cloudy' or 'hazy'. Haze usually indicates moisture above 1500 PPM [0.15%] at 'room' temperature. Therefore SVO or biodiesel that appears clear may not necessarily be dry enough for use as motor fuel.
The generally acceptable amount of water in SVO or biodiesel fuel is 500 PPM or 0.05 percent.
Adding more water to an emulsified mixture will cause a separation of phases to produce a third phase of free water as well as the dissolved and suspended water. This is similar to rain falling when the amount of moisture in the air exceeds the dew point. This free water is usually found on the bottom of containers of used SVO, where it's mixed in with the fatty sludge and food particles. If free water is evident then it's likely that dissolved and suspended water is also present in the SVO or biodiesel.
Over the years you have posted this many times.
Were you trying to make a specific point concerning this particular discussion?
Thank you John for reminding us of this. Over the years I've forgotten more than I remember.
In the link I provided there are further links which discuss water being physically bonded (through hydrogen bonding) with various other molecules of biological origin - proteins, lipids etc.
It seems fairly safe to say that finely dispersed water can be dissolved and not physically bound, and physically bound but not dissolved. In our discussions we tend to try and see things simplistically, whereas the reality is usually quite complex.
If we accept what I am saying, then everyone is correct.
The important questions are:
◙ what forms of water in our fuel will cause damage to our vehicles?
◙ how can we measure the damaging forms accurately enough?
◙ what quantity that we can measure is the acceptable limit?
John's development of the carbide manometer is one of the most significant contributions to biodiesel 'home-brewing', right up there with the 10-90 test and it's variants.
Hi John, been busy for a week or so and just got back home. I should say that it was not the Sandy Brae company itself that said this but their distributor in the uk and I have my doubts about their explanation. The carbide manometer works on the same principle as the Sandy Brae and I have found it to be pretty close to the Carl Fischer results. Your explanation of the time required to get a result is more likely but if that the case its not really a useful field test.
With the carbide manometer the secret to getting a quick result is grinding the carbide to a very fine powder. A glass marble helps mix the carbide and oil thoroughly. With biodiesel I find 4 to 10 minutes is enough to get a final reading and 15 is enough for oil. Is that your experience?
We have had an interesting if exasperating experience with commercial ion exchange washing. Its not something I have a lot of experience with but was insisted on by my client. We found that if we skipped the caustic processing stage and processed oil with a titration of 4 ml Koh. We got good conversion quickly and easily but the soap levels in the unwashed biodiesel was highish. The ion exchange resin removed the soap but pushed the TAN total acid number above the permitted 0.5. As an experiment we neutralized the acid levels with a very small dose of KOH. The sample now passed TAN but failed Metal Ion Content . In the end we abandoned that batch and caustic stripped every batch.
I should say here that we were entirely focused on passing the tests and Im in no way recommending caustic stripping for every batch for home brewers. There is no reason to caustic strip ( or acid esterify) oil that titrates below 8ml KOH.
My apolgies for inaccurate description of these tests, Im translating test results I receive each day from French into English and my French was never that good. What I described as "Metal Ion Content" should be "Group 1 ( alkali) metals ( Na & k)".
It is me (dgs) that gave the explaination about the Sandy Brae.
I didn't realise they had a distributor here in the UK. I realise the manometer does give faster results than the s/b when detecting dissolved water. Even though I said 6 hours I have found overnight is better. As you say, no good for a quick field check.
I think your experiance re TAN is probably normal when the resin is new. Hardwood chips/shavings probably don't sound high tech enough for your client but in practice work probably better than resins.
Maybe if processing speed and low soap are important factors ASM as the catalyst may be a consideration.
I concur with Dgs.
The final 'polishing' stage of my biodiesel production is filtering through a column of hardwood shavings in a 20µ filter bag, then through water absorbing polymer such as
water absorbing crystals
...and into the storage barrel. It's pumped from the storage barrel through a 5µ filter into the vehicle tank. I'm in a very dry climate so water absorption in storage is not a problem.
Moisture content is checked with a carbide manometer.
Hi all long time reader first time poster.
A big thanks to John on helping us with our fantastic project in France.
Just to reply to a couple of points
1 it was indeed as pointed out in Saint etienne which is quite near Lyon
2 generators are typically used in most stadiums as they provide an instant high power output which would be problematic in getting from a grid connection
3 our client was a local government run association called Ondaine agro which helps unemployed people getting back to work the idea is they collect the wvo locally process it locally and use it locally
4 the Sandy brae tester we are using did I fact show up inaccurate we have independently had a comparison test carried out against the Parker titration in an energy specialist reasearch lab called cirad in southern France over numerous tests in consistently read just over half against Parker titration
5 finally as a long time home brewer it's great to move up to something of this size again a big thank you to all involved especially John
Hi Greentech bio,
Thanks for the info.
The Sandy Brae will not read low against a K/F or similar as long as the test is left long enough. Sorry if I sound arguementative, it's not intended.
As said I have done dozens of comparative tests between my s/b and k/f and the comparison is quite close as long as the s/b is left long enough.
This was the source of much confusion and head scratching when I first used the s/b as I expected the result in the 20min time limit as per the instructions.
Thanks for the post, most interesting.
Ive no doubt that both you and John Galt are right, you have a lot more experience that we have with the Sandy brae. However having to wait so long reduces its usefulness for two reasons. First and most obvious it is often used to decide when to stop drying and having to wait 6 hours is no good for this. Second and possibly more serious is the fact that a manometer reading will change according to changes in both temperature and atmospheric pressure. A carbide manometer if left reading about 500ppm makes a reasonably good weather barometer.
You are right that the TAN problem was probably made worse by the fact that the resin beads were brand new and would probably settle down over time. However that didnt help us at the time. Greentech and I have agreed that a better setup for the future would be a woodchip wash followed by a resin bead polishing wash.
Another solution that I am experimenting with at the moment would be a short fine mist wash starting 30 minutes after processing is finished while the glycerol is STILL IN THE PROCESSOR . This would physically transfer the bulk of the soap and residual glycerol from the bio into the glycerol. The attraction between water and glycerol is so strong that the bio is no wetter than before the mist wash 800ppm approx. After drying this leaves the resin tower with very little work to do.
This has been my set up for years; it was patterned after the commercial system set up by RolfQuo (for those who read the interview I did with them). It works great, however I did a demeth prior to the wood chips and resin column and finished it off through a 2 micron absolute fuel filter.
Have a nice day.
** Biodiesel Glycerine Soap - The Guide
- on 5 continents helping people make & sell soap from the Biodiesel Glycerine.
From the Sandy Brae Test kit instructions.
"9...take the final reading 15 to 20 minutes after starting the test."
The BOLD type is in the instructions.
It was your interview with Rolfquo that first got me started on woodchip washing, thanks for that. I both demeth and fully dry the bio before the woodchips and filter to 5 micron afterwards. For my own domestic use I dont use the resin tower but Ive never had that tested to EN14214 . It does reduce the soap levels to under 50ppm so Im perfectly happy with that.
In a commercial system my concern is avoiding the acid / alkali trap that we fell into. If the feedstock contains FFAs these are converted into soaps. These soaps must be removed or neutralized to pass the alkali metal test (soap test) The resin tower converts the soaps back to FFas and if these are too high it fails the TAN test. If you neutralize these free fatty acids with a very small amount of KOH it passes the TAN test but fails the alkali metals test again.
The solutions Im considering are
Remove the FFAs using caustic stripping. Every batch begins processing at titration <1 KOH. The losses in yield are not as great as might be expected.
Remove most of the soaps using a mistwash before drying and resin tower.
Neutralizing the soaps using a sequence of wood chips and resin tower.
One problem for a commercial system with woodchip is that it creates another waste stream which complicates waste processing permits and EPA approval. The mist wash just adds to the volume of the glycerol waste product. I realize that these issues are not of any great concern to domestic producers. Slightly elevated levels of either soap or FFAs will not affect the quality of homebrew fuel.
|Powered by Social Strata||Page 1 2 3|