hi folks,

I'm working on the anaerobic co-digestion system and the problem that i faced to, is testing the COD of Glycerine in the lab. because glycerine is the biodisel waste and is really dens and viscose and is not completely soluble in the DI water, so the final solution ( DI water + Glycerin) is like a suspension solution that may not give me the correct COD at the end.
do you have any idea that i can dilute the Glycerine for COD test?

PS: I've tried methanol as a solvent but i had some problems at the spectrophotometry of the vials which came out of range.

cheers....

ya, as i noted, it's the biodisel waste and the glycerione is impure, that is, the glycerin purity was 72.2% by mass, 2% methanol, 1% moisture, and 5.4% ash+ other unknowns. so, i used 10ml methanol+ 1ml of the sample and diluted with the 10^4 times to get a good result at the standard curve limitation. when i used 10 ml methanol, i got a good solution but after digestion it didn't give me the absorbance in range. it was out of the assumed COD for glycerine which should be around 1100g/L.

the mothod is; closed reflux ( colorimetric method). APHA, 1998,standard methods.

i also tried to test other ways like mixing for for 1 day and heated up in the bath at 37oC when i used only DI water as a solvent, but those options didn't work. then i came up to use methanol as a solvent and first i used 5 ml MeOH + 1ml sample in 100 ml dilutted with DI water but it didin't work, so i used 10ml MeOh with the same procedure and the solution looked good but i still have problems when digest it and want to read the absorbance at 600 nm. it is out of range even though i calibrated the spectrophotometry by methanol or digested blank.

no i can't do titration method, because the assumed method for COD is closed reflux in this project and it has done before exactly by the same method.

the only problem is dilution the sample in the proper way. because it doesn't solve in the DI water and it goes like a suspension. it means my problem is mainly a chemistry problem rather than an engineering matter, so we need to find out a good method to dilute my sample to the order of 10^4 to be in the range of my standard curve.

as i know the COD of my sample should be around 1100g/L( it has been reported before), so i need to dilute my sample up to 10^4=10000 times.

the procedure that i do is: grab 1ml sample solve in 10ml MeOH then the volume reaches to 100ml by DI water. again i collect 1mi of this solution and diluted to 100 ml by DI water and i do all my COD test with the last solution. the problem is as soon as i add DI water to the solution the suspension particles produce in the solution and its not possible to do COD.
i was wondering that if i use 1ml sample in 99ml MeOH and solve it entirely then grab 1ml of this solution and dilute it to 100 ml with DI water. does it work?

what's your idea if i use other compounds, if exists, that have close COD to the DI water and are non-polar solvent. is it a good idea to use some other non-polar solvent instead of DI water ? i need to find a compound which has COD close to DI water and dissolve the glycerine.

i'm totally confused. mixing and heating up the sample do not work either.

Using methanol as your initial solvent is giving you an insanely high COD contribution.  I think the COD of Methanol is somewhere around 1-1.5 g COD/g MeOH.  If I do my calculation right, your MeOH contribution in your final dilution is 1000 mg/L.  That is pretty close to the upper limit for your COD curve anyway.  None of your solvents are going to be good for dissolving the sample.  They will all contribute very significant COD values to your final dilution.  The best approach, scientifically, would be to do an initial 1000x dilution in a volumetric flask.  Add the full volume of DI water to the flask and use a pipette to remove exactly 1.0 mL.  Use a dropper to add the sample into the flask to come back up to the line and then just pour the entire contents of the flask into a large beaker or flask.  Stir the sample in that to maximize your surface area and then withdraw the final volume to do the last 10x dilution while it is stirring.  Then, stir the working dilution amount and withdraw the portion that you will add to your COD reaction containers while it is stirring.  That should give you the best homogeneous mixture you can get.

It isn't so important what you use to transfer the original 1 mL of sample.  Since you are taking exactly one mL out of the volumetric, just bring it back to the line.  Even if you just get some of it on a glass rod and let it drip off it will work fine.  It will probably be tedious work.  Can you bill extra for your trouble?  

I would do an additional order of magnitude dilution (another 10x) to compensate for the MeOH.  I would also do a solution blank.  Start with 10 mL of MeOH and do the same dilution on it that you are doing on the sample.  Then you can subtract that from your sample result.

i've done the COD and the interesting part of results is that, the COD of the blank whic was only MeOH is higher than the COD of sample+methanol.

i checked all the dilution and everything twice and it seems ok.
i was wondering that could it be because of the bacteria in the sample and they use MeOh as a source of carbon, so the reaction between the sample (glycering sludge) and MeOH could have affected on the final result?

or is there any specific reaction between MeOH and Glycerine?

I think that I would just do a COD on my digester sludge which should be in solution or the digester is in trouble. Then I would calculate how much I think I am going to feed ratio wise to the digester. I would add this amount to make a dilution with digester sludge and run a COD of the spiked digester sludge. The COD of the feed material can be calculated based on difference in COD and dilution. If it will not mix with the digester sludge you would not want to feed that amount anyway. Ammonia in the digester sludge will help dissolve the mixture and will not add to the COD value.

Go ahead and make a dilution with DI water. Homogenize the dilution and pipet out the sample for the COD analysis just as you would any other sample. You can make a series of dilutions and after digestion some of them will be in a readable range. Average the readable samples. This should work. The solutions should be clear after digestion unless ther is too high a concentration of chlorides.

vahid wrote on Jun 7^th, 2010, 2:59pm:


How about trying a very small sample of the glycerin that is weighed rather than a water dilution.  You would need to calculate an "order of magnitude" sample size, say 1 drop (which would be about 1/20 ml if the sample was water).  Add the digestion reagents, and do the test as  specified in Standard Methods.  You will get a result that is NOT in milligrams per liter, but milligrams/milligram (or, something similar).  You can correct the calculation by determining the specific gravity of the glycerin.

vahid wrote on Jun 7^th, 2010, 4:17pm:

How about calculating the required sample size (weight); and, adding about that amount to the digestion flask/tube/?  You can weigh a fine needle and sample tube before and after.  Just add a tiny amount from the tip of the needle to the inside of the digestion tube.  You should be able to weigh out a very small sample if you are careful.

vahid wrote on Jun 7^th, 2010, 5:51pm:


I think that its worth a try.  The "diluted" samples have been a very small amount of glycerin with water or water and methanol.