Spent the last couple of days thinking about my results from previous batches and why the no titration formula doesn't work for me and probably others too. Lets do a hypothetical batch with the following. Batch 100l, Methanol 16l, Base 5.2 NaOH and use a 14/126 test tube for drop out measurement. Drop out measures 2.1ml which is 15% or 15l unracted oil. (2.1 / 14). Here is the flaw in the published formula. If you continue to use the same base (5.2) for the second stage, you will end up with still 15% of the 15l unreacted. (2.25l in this example). This requires (theoretically ) an infinite number of stages to complete the reaction. Not two.
Also, since the first stage reaction is incomplete and a significant amount of methanol remains, the formula demands additional methanol to excess. Here is what I have so far.
Use the formula for calculating the first stage and measure drop out. As you will come to realize, the drop out measurement accuracy is the single most important step. Next, I added 5 iterations of calculating the base required and added them up. i.e. first stage unreacted X base,plus second stage unreacted X base plus.....a total of 4 times. The fifth simply adds the same amount as was calculated for the fourth stage as we are getting some relatively small amounts at this point.
I reviewed my batch history and as stated in previous posts, second stage was calculated from the published formula then a fudge factor was applies (largely a guess) adding additional methoxide. And yes there were occasions where I would guess wrong and end up with a third stage.
I used data from my batches applied them to my new formula and compared calculated total NaOH to the total actually used and the totals correlated reasonably well. Worst case was off by 7.5%. (remember the actual amounts used were arrived at by a swag.)
Additional work need to be done for the methanol calculation. I had been using a processor from which I recovered the methanol after second stage. I did this to reduce the amount of methanol remaining so I could effectively bubble dry the methanol out prior to my dry wash. What I noticed is there was a close correlation to the second stage methanol added and the amount recovered indicating most of the second stage addition was not required. The theory states that 12.5% of the methanol will be consumed by a complete reaction. Our first stage is not a complete reaction hence considerable methanol remains available for further reaction.
One additional foot note. Drying the oil is also important. A high moisture content will consume some base quantity therefor stopping the reaction sooner. Presumably, this amount of moisture will not be present after first batch draining of glycerol. This will result in a drop out percentage that will be higher than needed for subsequent calculations leading to excessive methanol and base additions. My current threshold for moisture content is 500 ppm max.
I am moving to new equipment that will not allow demething so working with the minimum amount of methanol will be an important detail for me. Will go back and review the second stage methanol calculation shortly. I will post my spreadsheet calculator when this is done if anyone is interested. Hope this helps.
Jeff
