I had an idea for improving the safety of Grant's inline heater assm. If you inserted a T in the heater, then clamped a piece of SS foil or very thin sheetmetal between two flanges, you could vent a potential explosion to the outside, rather than have something burst. Any thoughts?



heater_illus_copy.jpg (39 KB, 245 downloads) heater

Instead of the metal foil why not just use a standard pressure relief valve?
I don't think you will be able to make a reliable system that can vent an "explosion" without some serious engineering and a huge pipe/blast route!
The small capillary tube you see in GL's design is to prevent a rupture, I just used a 30 PSI relief valve like one of THESE.
Jon

Hi. Most pressure relief valves flow very limited amounts with alot of restriction. If you used say, a 2" pipe for your heater pipe, and used a 2" vent, the relatively small volume of vapors in the pipe might be enough to rupture the diaphram, but with a 2" vent pipe, I very much doubt the pipe would be affected. If you had oil in the feed and return lines, I think an explosion in the heater would be pretty well isolated from the reactor itself, which as Martha would say, is a good thing.
Bryan

Hi Bryan,
I would suggest that there is no concern of an explosion during heater operation unless the reactor is empty or there is an airspace around the heater. Oxygen is required to have an explosion so if the heater is submerged the risk of explosion is eliminated. That being said any explosion in the heater pipe would instantly be transferred through the pipe work to the reactor full of fumes... BOOM!
The reason for the pressure relief on my heater is strictly to vent fluids in case the pipe plugs and the oil stops flowing past the heater creating pressure. I also have a flow switch that prevents the heater from coming on unless fluid is exiting the heater...
Jon

On further reflection, I think I will plumb the burst disc on the top of the reaction tank. I intend to incorporate several other safety mechanisms, but if the worst happens, the vent pipe should prevent a tank rupture, I think.
I picked up a disc on ebay for $20.
Bryan

Interesting Bryan.
How big (diameter) of a vent are going to build?
What kind of disc did you get?
Jon

Here's the spec's on the one I bought...

BS&B Rupture Disc 2" Model RD01040C
Type JRS Material 316
85psig at 72F

The ebay item number was 110301200988 if you want to see a photo of the unit.

My thoughts right now are to put a 2" T at the top of my reactor tank, with the rupture disc at the top of the T, the foam trap to the side. I think I will put a 2" to 3" adapter (black pipe) after the rupture disc, then convert to sched 40 3" PVC to vent the flames and boiling oil thru the roof and in a relatively safe direction...

Bryan

quote:

Hydrochloric acid is pretty much water anyway.Saturated with hydrogen chloride gas. That is why 34% is very concentrated HCL not dilute HCL.

Hi, I am a complete newbie. Have read this thread to page 15 and much of the acid/base process thread. Researching the chemicals used as I go. the MSDS for HCL says: However for constituent Hydrogen chloride:
Acute Oral LD 50 (ingestion) (rat) >900mg/kg
Acute Inhalation LC 50 (rat) 3124 ppm/1hr
Inhalation lowest LETHAL concentration (human) = 1300 ppm/30
min and 3000 ppm/5 min

Why are you playing with this HCL in a domestic environment.

The acid/base thread makes the point that the acid process is slow and will be stopped in its tracks with water. 50% h2so4 (battery acid) is too weak to do the acid/base process so I presume battery acid will not cause reverse actions on your completed single base process. H2so4 and water should separate at the bottom of your reactor immediately after you newtralise and stop pumping. Dain it off and recover the methanol. Drain the by product raise the temp and dry the BD. Is there any validity in this concept?

Add 100ml HCL or h2so4 to 100L very slowly (20 min). This sounds like a major manual headache. If dilute h2so4 will do the job it can be added to the push-pull processor just like methoxide.

This thread suggests variations in mix throughout the reactor.
If this is true how does one get an accurate titration. Take 1ml out of 100L and titrate. Maybe this is a major cause of the unexplained errors described on the forum.

I am impressed with the GL Push-Pull processor. Am expecting bad oil to process so leaning towards the acid/acid/base process so will probably have to recover methanol separately. Acid byproduct + base byproduct moderate heat and vacumm methanol recovery.

Love the knowledge being disceminated here. Thanks to all the experienced members for contibuting. Hope to be there one day myself.

Hi Graham,

Glad to hear that your push-pull method is coming along well.

I am working on a humanitarian engineering project designed for bringing biodiesel to the developing world. I was very interested to hear that you power a vacuum pump with a bicycle?

"I used a bicycle pump with the piston reversed to suck rather than blow"

As cost is an issue with these projects, your idea sounds ideal. If you could throw me a link with any more information I would be very grateful. Thank you.

~Ryan Johnson

Just successfully completed my first batch on the GL push-pull processor.

I heated the oil and removed 500 ml. of water under vacuum.

One thing I discovered is that my two horse pump will cavitate when subjected to over 17 inches of vacuum and stop pumping but this did not seem to effect methanol recovery.

Processed 345 liters of oil and recovered 8 gallons of methanol out of 22 gallons.

The methanol recovered is 98% pure.

It almost passed the 3/27 with only a pinhead sized drop showing and did well on the pHLip test which indicated it was slightly under reacted.

I'm almost certain this was because my method of mixing the methoxide was poorly done. I have modified my mixing tank to insure a better mix next time.

I recovered about 23 gallons of glycerin. I waited almost ten minutes after recovering the methanol to let it settle before opening the drain valve.

This was a mistake as the glycerin was so thick that it only oozed out.

I had to put compressed air in the tank to force it out.

I have not made biodiesel before so I'm not sure how thick the glycerin is supposed to be but mine is in five gallon buckets and is so firm I can stand on it without sinking in.

I guess there is no way I can use it to pretreat the next batch.

I'm really excited and can't wait to do the next batch this weekend.

Dave

Congratulations!
Your recovered methanol amount sounds consistent with what I get with my 100L push/pull setup.
I don't run the circ pump when I dewater but I do when doing the push stage. I am able to drain the glycerinafter about 10 minutes with out too much grief through a 3/4" hose.
You should do the 3/27 before you stop the mixing to be sure its fully reacted and then again once complete the recovery stage to determine if a back reaction occurred.
You can heat up the glycerin to make it liquid again and then use the vacuum to suck it in for a pretreat.

How did the batch titration go to determine when to stop adding the acid?
How much acid did you add?
What was your initial feedstock titration?

I look forward to your reading about your next batch!
Jon

Hi Jon

My initial titration was 3.6.

I had a very difficult time seeing the color change. I had three people watching and nobody was in aggreement.

I used 98% sulfuric acid and from reading this thread determined that I would probably need less than 150ml.

While sucking the acid into the processor the suction tube popped out of the jar and spilled a few ml. onto the floor so I think I put in about 130 to 140 ml.

I did a titration afterwards and it took three drops of titrating fluid to turn it solid pink.

I should mention that I grabbed an oily rag to mop up the acid,wearing gloves of course, and the rag caught fire.

That acid is scary stuff.

Dave

quote:

Originally posted by Hummerbk:
ops of titrating fluid to turn it solid pink.

I should mention that I grabbed an oily rag to mop up the acid,wearing gloves of course, and the rag caught fire.

That acid is scary stuff.

Dave

acid spills are a bit of a problem there are lots of ways of dealing with them from special chemical absorbent pads and powders available commercially (these I expect to be expensive) to simple neutralisation with the likes of baking powder (which will result in potentially harmful gasses) but for a small spill in this type of set-up flushing the area with a large amount of water will dilute the acid to a negligible level.

trying to mop up a strong acid with anything other than propose designed spill kits as you found out is not a good idea.

it is worth thinking about in advance! if you are going to use a process like this then you really should think about how you will deal with a problem before it happens.

sounds like your reactor is a success though, good work!

quote:

I had a very difficult time seeing the color change. I had three people watching and nobody was in aggreement.

Ya I feel your pain! I am hoping a better way comes about!
That being said, I bet 120 to 150Ml of muriatic per 100l batch would produce quality results but that is just a gut feeling and wild ass guess!
Jon

I think next time I will add most of the acid, maybe 100 ml. and then pull a liter out and while on the stir plate add acid a drop at a time.
I think I'll be able to see the color change easier that way.
I can always do a titration to confirm I have not gone too far.
After a few times I should know how much to add.

We did have a bucket of water standing by when the rags caught fire.

Dave

Hi all,

I was wondering what type of vessels most people are using for the "rapeseed chamber". So far all I have found is GL recommends a 1-2 L vessel and someone posted that they are using a 5 Gal propane tank. Thanks very much!

~Ryan J.

quote:

I was wondering what type of vessels most people are using for the "rapeseed chamber".

That vessel is not pressurized or under vacuum, so, as long as it can withstand the temperature, you can use almost anything. I'd recommend you to use a metallic vessel for safety reasons.
In my case, I have a Heineken 5 liter mini keg waiting for it...

quote:

that's about 1 atm., (14.7 PSI)

also be sure to add another atmosphere (at sea level) acting upon the outside of the tank.
our systems were designed around it so we cannot feel it -- but the tank will.
if the air tank is used it will likely have corosion on the bottom creating a stress riser
to even lower whichever capacity it may have.

I've tried two experiments so far, titrating the esters and glycerin separately and calculating the amount of acid needed. I can't say this will work for everyone, but the concept seems to hold promise and may be worth pursuing:
My processor has a drain port just after the pump. When the mix has been running long enough to be sure of complete reaction, I drain a sample while it's still pumping and let it settle for a couple minutes. Using a syringe, I draw 1cc samples from the top (ester) and bottom (glycerin) layers. I know they're not completely settled, and that's one variable I haven't investigated.
I have a 0.1% solution of acid to titrate with. This allows the same calculation as my KOH titration, which we're all familiar with.
Blank some IPA (rubbing alcohol) with Phenolpthalein and the regular KOH solution in a beaker, in preparation for the acid titration.
Add the 1cc GLYCERIN to the IPA - it turns brilliant purple because it's strongly alkaline.
Add the 0.1% acid solution until the color just goes away. Record the amount used.

Repeat the process with the 1cc ESTER sample. It doesn't take much acid, compared to the glycerin. OK, still with me? I now have the numbers for acid neutralizing the glycerin and the esters separately.

I use Muriatic acid from the pool supply store. It's 31.45% HCL, and a little friendlier than Sulfuric Acid.

For both experimental batches I assumed (dangerous word, I know) 25% glycerin and 75% esters in my batches, based on yields of previous batches of the same oil. I Multiply the ESTER titration value by 0.75 (75%), and multiply THAT number by my batch size (175 liters). That gives me the amount of acid to neutralize the ESTERS in the processor.

I multiply the GLYCERIN titration value by 0.25 (25%), multiply THAT number by the batch size, to determine the amount of acid to neutralize the GLYCERIN in the processor.

Total the two acid quantities to determine the TOTAL amount to add to the processor, which is still running.

Results so far: after adding the acid and circulating for a while (2 hrs), I drained and processed the glycerin and esters normally (not "-pull"), then checked their alkalinity with phenopthalein - neither component were alkaline any more, but it only took a drop of 0.1% KOH solution to turn them pink. This indicated to me that the method would have worked well enough to use in a push-pull method.

The two batches were from the last and nastiest of the oil I collected over the winter. The first was a re-reacted failed batch that had a second KOH titration of 8.2, the other was a 8.5 (I'm used to seeing titrations less than 4.0). I later discovered my KOH solution was weak, giving grossly higher indications than normal. This caused an excessive degree of alkalinity in the mix, but it only really shows up in the glycerin. The esters only needed 0.1cc acid solution to neutralize, in both batches. The glycerin took 1.6cc (first batch) and 5cc (second batch, which prompted me to double check the KOH solution).

Total acid added to the first batch was 83.1cc. Total added to the second batch was 231cc (I'd badly overdosed it with KOH, but it made a good test batch).

If this method works well enough, then it could eliminate the need to add phenolpthalein to the whole batch.
Cheers,
JohnO

This message has been edited. Last edited by: johno,

JohnO, that is a very interesting method of determining how much acid to use. Can I ask how you made your .1% acid solution?
I've been just guessing how much acid (sulfuric) to add since I can't see a color change in the sight tube.
If I can get this method to work for me I've got about 500 grams of Phenolphthlein powder for sale.

Dave