Silk City Distillers

VWR 1109 low temperature lab chiller, glycol at 50%, I need to use a gear pump at that temperature.

I'm working on a post about the vac setup that'll go up on SD first.

Vacuum is fun.

i have a gin still that runs the product condenser at -20f.

I can easily distill at room temperature with no heat input at all to the boiler.

I'm not an organic chemist, but the most probable source of the "minty" aroma is going to a salicylate ester (this is how most commercial mint flavors are made).  Camphor - not likely (sometimes described as minty, but more menthol or eucalyptus).

Salicylic Acid (SA) is found in both Grape and Wheat.  Looking through the literature - SA appears to sometimes be used as a preservative.

Ethyl Salicylate has a minty aroma and flavor.

Methyl Salicylate has an even stronger mint aroma and flavor than ethyl.

Other 4-methyl compounds could be the cause as well.

Why the reaction is favored, increased, etc - beyond my pay grade.

Try attempting to hydrolyze the esters (if they are salicylate esters) with NaOH during distillation in a small lab still.

Posted this two years ago over at Stilldragon.  This was the last time I experimented with baking soda for ester hydrolysis.

-----------

Ok - I wasted a lot of time and alcohol yesterday running through trials to determine what the appropriate amount of baking soda is. I didn't want to clutter up other threads this is being discussed in. The game plan is to run a set of trials with a number of different ester hydrolysis/acid neutralization approaches.

Here's what I got - Day 1 - The Sodium Bicarb Trials

First I wanted to understand the pH dynamics of sodium bicarb in a strip:

Sodium Bicarb - 2.5 grams per liter (approximately 1 tablespoon per 10 liters stripped wash) is sufficient to neutralize acids out of a typical strip and push the pH near the equivalent of a water saturated solution. Starting pH was 4.3 - by 2.5g/l sodium bicarb this was pushed to pH 8.3. Adding any more than this amount did not push the pH any higher.

Repeating the same test with a clean hearts cut (approximating a very clean strip), even less, 1 gram per liter (approximately 1 tablespoon per 25-30 liters of clean stripped wash) was sufficient to push the pH to 8.3

Then I wanted to understand what the upper limit would be, given a higher than typical acid level:

Repeated again adding 5g/liter of citric acid to a 50% ethanol solution - to simulate a very high acidity strip - probably unrealistically high, but not quite vinegar (an awful infected run perhaps). pH was 2.6. Even a small amount sodium bicarbonate was enough to turn it into alka-seltzer - with significant co2 production - you would absolutely recognize the amount of fizz. By 5 grams per liter the pH had moved up to approximately 5.1 (one tablespoon per 5 liters of wash). By 10 grams per liter the pH had moved up to 6.2 (1 tablespoon her 2.5 liters of stripped wash) . By this time there was very little citric acid flavor remaining in the sample. By 15 grams per liter the pH had moved up to 6.7 (1 tablespoon per 1.25 liters, by 20 grams per liter (1 tablespoon per 0.75 liters) the pH was up to 7.2 (and tasted very salty). Between 15 and 20 grams per liter I believe I passed the point of saturation, as I was unable to dissolve some of the sodium bicarb and it was settling at the bottom. Realistically, 1 tablespoon per 2.5l is probably the upward max, and only if your strip was awful - or distilling something with very high volatile acidity (faulted wine maybe?).

After running these tests, I realized I should have done this with acetic acid, so that I could run each of these through a distillation and organolepically assess the ethyl acetate produced during distillation. I'll try this next time, but I need a balance with a 0.1g precision so I can run smaller trials. Next time.

Next, I wanted to understand the time factor, how long did the hydrolysis reaction actually take to complete.

To understand the time factor - I distilled an "ester bomb" with the lab still. 500ml (300ml tails, 200ml hearts) - along with (50ml)acetic acid, propionic acid (10ml), lactic acid (10ml), and butyric acid (10ml), sulfuric acid to bring the pH down to about 2.4. Refluxed for one hour. Worked great - took the sickeningly sweet juicy fruit hearts cut (50%) and split it into two containers (250ml).

Began adding sodium bicarb to one sample, keeping the other sample untouched for comparison purposes. If bicarb can reduce the esters, I should be able to discern a pretty linear reduction in ester odor in the test sample, no?

Tell you what, not much of a difference, it was relatively minor. Ran all the way through 20 grams per liter, and beyond the point at which I could dissolve additional sodium bicarbonate, and the ester odor was nearly as prominent as the control sample. You could very faintly discern the butyric and acetic acid odor at the 2.5g/liter mark (1 tablespoon per 10 liters), but not much change after that. Even past saturation, but the ester odor still overwhelmed all else.

I then heated the sample, to test if I could accelerate the reaction with heat, perhaps it was just too slow. During heating I liberated a bit of vapor, which strongly smelled of volatile acids. Hmm, not what I expected. Cooled it down to see if perhaps some time under heat was sufficient to hydrolyze the esters, nope, nearly identical to before.

Gave up around 1:30am - let the samples sit overnight. Fast forward to now, roughly 9hr later - little to no change since last night. Now, it could very well be the amount of ester created was far beyond the ability to easily hydrolyze, but not sure that makes any real theoretical sense.

Decided to distill the sample this morning rather than letting it sit, I don't believe it would have done much to wait, especially since heating it made little to no difference.

Distilled into 4 equal fractions using only minor passive reflux, 3 distilled through and the final being the remaining boiler contents.

NOW WE ARE TALKING

Heads fraction - still very similar overall odor, however I expected it to come through very strong as it did during the initial esterification - significantly reduced odor.

Hearts fraction, ester odor nearly nonexistent, not that I should be surprised, we know this works, I just didn't expect the very significant reduction in esters - especially since the sample going into the distillation was still very, very strong.

Tails were similar (I did start with a hearts cut) - No acidic aroma - which is typically of the tails cut during the esterification. Absolutely no ester odor.

Remaining boiler contents, shockingly free of acidic odor, as is VERY common during the esterification process. Absolutely no ester odor.

Need to run a few more trials, since what went into the final distillation test here was a sample with excess bicarb - past the point of dissolution. I'd like to try this with significantly less.

DAY 1 CONCLUSION - I strongly feel that the 1-2.5g/l amount is more than sufficient to provide for ester reduction in low wines (this is 1 tablespoon per 10 liters to 1 tablespoon per 30 liters). I do not feel that any waiting time, past the point of dissolving the bicarb in low wines, is necessary at all - the reaction was almost nonexistent cold - 9 hours no change - but clearly takes place rapidly during distillation. For the average distiller, plain old baking soda is going to be hard to beat - really, it works that well.

Fair.

CCR equipment looks like eye candy for sure.

I suspect you are right on the tails chase - it's not going to be ideal for a spirit run.  I would imagine you'd see a huge swing from start to end ABV as a result.

At a minimum, I'd imagine you would want to work with a boiler charge well above what you typically set, and use the condensate from heating to provide whatever dilution to your usual charge abv.  For example, if your strip typically ends up at 50% and you dilute to 30% for your spirit run, you might want to just charge at 50%, knowing that you'll probably get closer to that 30% ballpark during the run.  Perhaps some combination of pre-heat and steam to reduce the overall liquid addition?

There were one or two people on here who were going to try it for a spirit run, but they never posted their results.

Just to be clear, my comment was directed towards the OP, and not intended to start a fight.

Pretty basic setup.  We have a steam line going to a steam filter, which then goes into the tank from the top.  At the bottom of the pipe, we have an elbow and the 3/4" eductor is screwed on.  In between there is a simple ball valve to turn the steam injection on or off.

It makes a hell of a racket, sounds like the roar of a jet engine.

The assembly is mounted with a union, and we pull it apart to clean when necessary.

We couldn't go straight through the wall, because we're using a jacketed and insulated tank - jacket only used for cooling.  But, in a tote, you could just go right through the side with a weld-in fitting.

The benefit the fancy eductor brings, is that it uses the motive power of the steam to act as an agitator and keep the tank mixing.  During startup, it does a nice job heating up quick, but once you get the corn in and gelling, it's not enough to replace a real tank mixer.

We use a TLA tank heating eductor 3/4" with our 16hp boiler.

http://www.nciweb.net/Jacoby PDFs/Tank_Eductors.pdf

If you feel you must in some way treat your boiler feed water, Reverse Osmosis is far superior to treating your boiler feed with chemicals.

We have a mix of injection (mash tun) and jackets (stills).

Jackets are plumbed to return condensate, steam injection - obviously does not.

We don't use boiler additives.

I gotta ask.

Doing this by hand, it's the fastest process in the bottling line.

Looking at some of these machines, they look to be considerably slower.  Especially the ones that look like you need to manually load the slotted feeder.

What an I missing?

So what you are saying is, the entire history of American Whiskey is wrong?

Why not just use steam injection if you are going to use it to strip - no jackets required.

In another life at another job - the KPIs you are talking about are more in line with annual performance goals set for your employees - which would be reviewed at least annually - and would serve as a basis for any merit increase.  Generally, goals would be set in conjunction with the employee, and could be broader than their specific day too day activities (for example, learning and cross-training as a goal).  However, these are always SMART goals - Specific, Measurable, Achievable, Relevant, Time-Bound.  Your examples seem to fall in this category.

But in terms of bonuses, these would be higher level shared goals, for example - for the manufacturing team, might be profitability - which encompasses all of the things you are mentioning.  Sales goals, revenue goals, profitability goals.  You don't want to be too prescriptive about how the target is reached, as you risk stifling innovation and new ideas.

But don't, in either case, put a KPI on someone who has no authority or control to move the needle - this can be extremely frustrating.

Another thought - why not some additional caramel color to darken the bitters?  It might make the clouding less apparent, letting you preserve the flavor profile you worked hard to get.

Increase your botanical quantity or maceration duration to account for the flavor loss in filtration?

If this is something that really bothers you, why not use a sanitary rupture disc for overpressure protection?

No moving parts, all compatible materials.

Simple metallic disc that is in a triclamp adapter.  Kettle on one side, exterior vent piping on the other.  No moving parts, very simple design.  Some companies even make rupture discs integrated with a triclamp gasket - that can be installed directly into a clamped pipeline junction.

Unfortunately, they are not inexpensive, and it's one time use.  However, if it bursts, it probably saved your a$$, and is worth every penny.

Don't disappear - make sure you update the post with the outcome and any further observations you have!

If you are concerned about copper reactions and copper in your distillate, you could also consider replacing everything after the apex of the vapor path with stainless.  Copper Up - Stainless Down.

Scorching would be very obvious in the flavor, and it wouldn't cause the coloration in the distillate.

The puking needn't be obvious, if there was foaming in the still, that would probably be enough to carry over into the distillate on a fast/hard strip.

If you are just stripping bulk cider in smaller batches, just add it back into your next stripping run.

Still puke?

Roughly, very roughly.  For every 9,500 BTU added, you'll add a gallon of water.  Remember, simply, BTU is raising the temperature of 1 pound of water by 1 degree Fahrenheit.  So, let's say you had 100 gallons of water (834 pounds) and you wanted to raise it by 50 degrees, you would need 834*50 = 41,700 BTU Required.  That would take 41,700/9,500 = 4.4 gallons of total water/steam condensate added to get there, for a total volume of 104.4 gallons.  Again, very roughly - this doesn't take into account the net change in total temperature as a result of the additional water volume, but that should be minor.  It also assumes 100% efficiency.

Culinary steam does have a specific definition in the FDA, which is easy to meet if you use *NO* boiler additives or chemicals, and you can filter the steam such that 95% of particles 2 micron and larger are removed from the steam.  We use the Spirax Sarco filter that @Southernhighlander mentions.

Another possible option.

Wilden sells a diaphragm pump that uses an electronic/solenoid batching controller.

The pump is an air operated diaphragm, and each stroke dispenses a specific volume.  By counting the strokes, you can dispense a specific volume from one tank to the other.

This might be easier/more accurate than a flow meter.

Or, why not just use a scale on the bottom tank?