Jump to content

meerkat

Members
  • Posts

    215
  • Joined

  • Last visited

  • Days Won

    8

Everything posted by meerkat

  1. The pumped system is far more common than the individual drains in my experience. As Silk has suggested, for conversions between the two systems Paul ( @Southernhighlander) is definitely the "goto" guy.
  2. For spirits over 600 mg/100 ml use 30.31 (C)(1) If the evaporation method indicates more than 600 mg/100 ml then you can estimate the true proof from the apparent proof using the Liqueur Hydrometer Temperature Correction calculator in AlcoDens LQ. This is obviously not TTB approved but gives values that you can use in blending calculations. The final product proof will have to be verified once you replace the lab still.
  3. I don't think that you can generalize that packed columns are shorter than trayed ones. There are so many factors that go into determining the Height Equivalent to a Theoretical Plate for packings. And with trays there is always a trade-off between diameter and height. You can minimize the carry over from a tray to the one above either by spacing the trays more widely (i.e. making the column taller) or by increasing the column diameter (i.e. decreasing the vapor velocity). I have no personal experience with ProPak, but with all those perforations I would expect it to be good. The price that came up when I searched for it was rather high, but maybe that includes shipping to South Africa?
  4. I agree with your observations in terms of columns used in the craft distilling world. Packed columns have a narrower range of (efficient) operation than bubble cap trays. This means that if you are running a batch still where the rates and compositions are constantly changing then it is better to use trays. But if you are venturing into continuous operation where the column can be designed for a specific and constant capacity then it would be quite feasible to use a packed column. I think columns up to about 12" diameter have been mentioned on this forum?. A disadvantage of packed columns in this application is that it is not so easy to take off side draws. It is also difficult to predict at the design stage where the side draws should be located - I'm sure you have seen the trayed columns that look like hedgehogs with off-takes on every tray. Sometimes you will see a continuous column with trays in the bottom half where most of the side draws would be, and then a packed upper section. I am currently involved in an expansion at a large vodka producer where some of the columns are 2500 mm (8 ft) diameter, but these are valve tray columns. You should probably get a recommendation on the distributor requirements from your packing vendor. I would guess that a single center mounted inlet is OK up to about 4". If it is a clean feed you can put a solid cone spray nozzle on it. For larger than this, but less than 2 ft a simple spider made of standard piping components should be OK. Above this it would be best to use a proprietary design.
  5. I have worked with packed columns of up to about 5 foot diameter, but I wouldn't be surprised if columns of 3 times that diameter are used. Structured packings are often used in vacuum columns and these tend to have very large diameters to keep the velocities down. A rule of thumb with packed columns is that the packing pieces should be not more than 8-10% of the column diameter. Rings are available up to 3.5 inches, so a column using that type of packing should be at least 40 inches diameter.
  6. Hello Pete. A good cross check on these numbers is to measure the solids content by evaporation. This is an easy test that can be done using fairly basic equipment. The 44 % ABV that you measured with the Snap 41 is the apparent ABV. As you said, the true ABV would be expected to be higher than this because the dissolved solids would increase the density of the spirits. It looks like the initial lab result of 43.1 % ABV was definitely wrong because it should not have been less than the 44 you measured. Whatever the solids are that have been dissolved out of wood, their density should not be too far from that of sucrose. Wood is made up of cellulose and hemi-cellulose which are just polymerised sugars. It would take very little dissolved sugar to change the measured ABV by the small difference you are seeing between the lab result and the apparent ABV. To cause a spirit with a true ABV of 44.7 % to give an apparent ABV of 44 % would require only 3.4 gram/litre (or 0.36 Brix) of dissolved sugar. Measuring the actual dissolved solids will give a good indication of the expected obscuration.
  7. It is hard to add anything to the excellent write-up by Silk City in the earlier thread linked by Thatch. I can only elaborate on why vacuum distillation is different, rather than automatically better. We know that in distillation the more volatile components find their way to the top of the column, while less volatile components are found lower down. So a highly volatile component like methanol migrates to the top faster than ethanol, while a less volatile component like fusel oil stays nearer the bottom. The ratio between these volatilities is known as "relative volatility", and relative volatility changes with temperature. So if you distill at a lower temperature the relative volatilities will change and the profile of flavor components in the column will change too. This is not good or bad, unless you are looking for a particular profile. This is a small effect, and the effect of temperature on degradation reactions and the impact of low pressure on the column diameter requirements (both discussed in detail by Silk City) are generally more important.
  8. @PeteB Here is a similar discussion we had a few years ago and how a customer of mine solved the same issue. Even the commercial lab could not get consistent, accurate results
  9. It is quite a calculation to convert the lower flammability limit figure of 3.3% into kg or litres of ethanol. At only 3.3 % we can regard the 1600 m3 as pure air to make the calculation easier. At 25 C and atmospheric pressure the density of air is 1.18 kg/m3 and the 1600 m3 will contain 1888 kg of air. The 3.3% is in terms of molar percent (same as volume % at this temperature and pressure) so we need to convert the mass of air to moles. The molecular mass (molecular weight in the US) of air is 29 so the 1600 m3 contains 65 kgmols of air. The 3.3% would be 2.15 kgmols of ethanol - or 98.7 kg of ethanol. I don't have a full chemical simulator, but a rough eyeball interpolation tells me that if the room were sealed, the barrels open and the spirit at 64 % abv the 3.3 % level would be achieved if the spirit was above 18 C (64.4 F). Of course the room is not sealed and the barrels are sealed so this is a worst case situation. I agree that the ethanol vapour would mix quickly with the air, but it is very difficult to say how quickly so it is safest to assume that the ethanol vapours could sink into drains etc. I know of men being suffocated by CO2 where it was denser than air because it was very cold and the CO2 sank into a pit - even though it would eventually all be mixed.
  10. The Snap 41 measures SG only and will be confused by the sugar. Ignore the results you get from this unit. The Alex 500 is designed to measure solids as well as alcohol but the web page does not say how it does it. The fact that it does not work well for high (> 50 abv) alcohol levels suggests that it is using Refractive Index as its second measurement because I know that the mathematics for combining SG and RI does not work well for high abv. You should give Anton Paar a call and ask them if the Alex 500 is suitable for your application - I believe that it is.
  11. For spirits containing alcohol and water only, you can determine the alcohol content by measuring either the density or the refractive index of the spirit. Adding sugar will impact on these readings and you cannot use the standard tables. The sugar changes the density in the opposite direction to the alcohol, but it moves the refractive index in the same direction as the alcohol. These 2 opposing effects can be combined mathematically to determine the abv and the sugar content at the same time. There are machines from Anton Paar and Rudolph Research that do this, but they are not accepted by the TTB. Only the lab distillation described by JailBreak is accepted. What 2 machines are you using currently?
  12. If you are in the US and subject to TTB regulations then the only accepted way to proof something with 5% sugar is by using the TTB lab distillation procedure. There are no chemical, instrument or software methods that I am aware of that would be acceptable. The software will certainly speed up the process of getting to the target proof, but the final product will always have to be gauged using the TTB procedure.
  13. Pumping the bottoms from the second column to the top of the first is a closer match to a single tall column - but whether it is better or worse than the double defleg arrangement I do not know.
  14. I only have experience with one column where the tray spacing was deliberately made very tight. It was a very difficult column to operate and never achieved its design capacity. Unless you have a very skilled designer with a proven track record I would stay away from tight tray spacing. Split columns are common and there are many vendors who can offer you these. I would be interested in reading how pumping the bottom liquid from the second column to the top of the first to provide its reflux compares with using a dephlegmator to provide the reflux on the first column.
  15. Something I didn't mention before, but which will apply to both @mors9 and @Skaalvenn, is that horizontal piping that can contain vapor and liquid at the same time should always be sloped down in the direction of liquid flow and never be truly horizontal. This is well understood by civil engineers and domestic plumbers, but we process people sometimes forget it. Ideally we want the liquid level in the base of the column to be a couple of inches above the base (which would make the drain pipe full of liquid), but because loads vary this cannot be guaranteed and at some stage the level will be in the horizontal pipe. This is always a recipe for problems and causes the spurting and cyclical behavior we often see. Even if we try to make it truly horizontal there are often high and low spots which aggravate the problem. It is best to give a generous slope of around 2% (2 cm per metre, 1/4" per foot). This ensures that the vapor will not impede the liquid flow. Over the last 40 years all the worst piping problems I have experienced have involved gas-liquid flow in horizontal pipes and this has caused me to rather err on the large side when sizing these pipes. An 8" column might have a return flow of 0.6 to 1.0 gpm and a 12" column a bit more than double that. I would prefer to use a 1" drain on the 8" column and a 1.5" drain on the 12" - even if this can only be done for the horizontal sections. I have discussed this problem with @Southernhighlander before and I know that he understands the problem well and has a practical working solution. A self draining system is nice, but if you just want the problem to go away then get a quote from him.
  16. That all sounds fine to me. The last thing that I can think of checking is if there are anti-siphon holes drilled either in the elbows just inside the boiler or just below them. If there are anti-siphon holes but no U-bend seals in the drain pipes immediately below the columns it could cause a problem. If you have anti-siphon holes then you should also have U-bend seals. From the spurting behavior and the vapor bubbles in the returning wash it really looks as though there is vapor getting into the return lines. If there are no anti-siphon holes or other openings above the liquid level then I am stumped.
  17. How far do the return pipes project under the liquid level in the boiler? Are there any U-bends or seals in the return piping inside the column skirt (i.e. in the section of pipe that cannot be seen in the video)? How far above the boiler liquid level is the entrance point to the drain pipe at the bottom of each column? Do you have any pressure measurements?
  18. This is a common problem in still arrangements like you have. The solution proposed by @Stumpy's will solve the problem but you may get away with a less complicated arrangement. In the photo posted by Stumpy's the base of the second column is below the liquid level in the boiler and it was essential to use a pump in his case. If your columns are sufficiently above the liquid in the boiler you might not need a pump. A similar problem was discussed in https://adiforums.com/topic/11969-return-line-to-pot-question/ where the photo shows the column bases higher than the boiler. In that thread I mentioned that sometimes small holes (3 mm, 1/8 inch) are drilled into the return pipes inside the boiler but above the liquid level. This is especially necessary to prevent air locks if there are gooseneck (U-bend) seals in the return piping. Never use an S-bend seal in the return piping. Although the return piping seems very simple there are many things that can go wrong in gravity driven drains. If you can post a photo or sketch of the piping it would likely bring out better comments.
  19. A rotavap is a very useful piece of kit, but I have had customers (of my software) get confused by the readings. By pulling a vacuum the boiling point of the solution is lowered and it is important to realize that this also means that the condensing point is lowered. If the cooling water on the condenser is not sufficiently cold for the vacuum you are using some alcohol can be lost in the condenser. It has no effect on the solids left in the evaporator bulb so the solids analysis will be correct, but you may lose alcohol and get an incorrectly low proof when proofing the condensed spirit.
  20. In the US the TTB distillation method is the only officially acceptable method of proofing. But as PeteB has pointed out, if you have two separate measurements of two independent properties (i.e. density and RI in the case of the Rudolph machine) you can combine them with the applicable math and get to an accurate, but unofficial result. You can do a similar procedure using much cheaper, but also much slower, equipment than the Rudolph AlcoTest. With a simple drying oven and an accurate weighing scale you can measure the solids content after the infusion. Once the sugar content is known it can be combined with the SG of the spirit to calculate the proof. Here is an example done in AlcoDens LQ.
  21. I think @PeteB is correct regarding the reactions being faster in the liquid phase. I have seen large continuous vodka columns which were made of stainless steel, but the reflux was passed through a bed of copper raschig rings before returning to the column.
  22. I have been supplying engineering software (including a units conversion program) to the US for 30 years, and alcohol blending software for 15 years. Over this period I have seen a gradual shift towards metric units in the US and the eventual total switch over is inevitable. You have already switched to 750 ml bottles to match the rest of the world, and the other changes will not be much more difficult. I think that being proactive about it is the right way to go. Apart from the actual units, the basis of the ABV measurement will have to be changed from 60 °F to 20 °C as the reference temperature does change the ABV very slightly.
  23. Check the regulations CFR 30.31 "Determination of Proof" and CFR 30.32 "Determination of Proof Obscuration". Unless you are doing an extreme maceration there is a good chance your solids will be below the 600 mg per 100 milliliter limit. You only need to do an evaporation test to confirm the solids levels. If you are below the limit then you do not need the fancy lab distillation setup or the expensive electronic meters, and the proof determination is basically the same as for a normal vodka.
  24. To move the apparent ABV from a true value of 42.52% down to the 41.97% that you are reading would require only 0.26 mass % of sugar/solids. The only way to be sure of what is happening is to do as Silk City has recommended and do the evaporation test. This will tell you the quantity of solids present. If your column was sufficiently overloaded to carry over that quantity of sugar you would not achieve any rectification at all, so I believe you can rule out sugar from distillation being the cause. If there are solids present they would have to be leached out from the maturation barrels, but I don't have the experience to judge whether this is possible. Anyway, the first step must be to determine if there are solids present, and then the next step will be to track down the source. A word on the results from the external lab. The top-of-the-range digital density meters sometimes incorporate a refractometer and with some fancy math the density and RI can be combined to give you the alcohol and sugar content. But this is not very accurate and is not accepted by the TTB. Is you external lab doing it this way, or with a proper TTB approved lab distillation?
×
×
  • Create New...