This series of posts has to do with the usefulness of Clinitest (a test normally used by diabetics to measure sugar content) as a brewing tool. Dave Burley doggedly continues to hail the merits of Clinitest, while I (equally doggedly) continue to write rebuttals when Dave posts in favour of the Clinitest. Mind you... I don't say that Clinitest cannot be used or is not accurate, but rather that it's accuracy across a wide range of beer types is unproven.
In Homebrew Digest #2333, AJ deLange writes:
AJ> Date: Fri, 31 Jan 1997 14:08:40 -0500
From: email@example.com (A. J. deLange)
Subject: Reducing Sugars/HBD
Dave Burley suggested the use of diabetic products to determine residual sugar content towards the end of the fermentation. This is workable but also quite approximate. [snip] Tests based on the reducing properties of aldoses (glucose) and ketoses (fructose) [such as Clinitest] are also valuable but suffer from a similar flaw in that when dissacharides are formed in which the anomeric carbons of both monosacharides are involved in the glycoside bond (such as sucrose) the reducing power is lost. Thus sucrose is not a reducing sugar and is not detected. Furthermore, polysaccarides and oligosacharides which are not fermentable can have reducing ends which will be detected. This doesn't mean that these products can't be used with some benefit. Clearly a reduction in reducing sugar level to a stable value indicates that fermentation is nearly complete.
Al> I'd like to point out that this is not the way in which Dave suggests that we use the Clinitest... Dave suggests that we can safely bottle when the Clinitest reads less than 1/4% reducing sugar.
In Homebrew Digest #2576, I responded to Dave's post in a recent HBD:
Dave> My posts stretching back to when I first suggested to HBDers that Clinitest was useful in finding out when a brew was finished, clearly stated that Clinitest was useful for reducible sugars as a direct indicator of the sugar content and should be used for this purpose in favor of that nearly useless object of your affection called a hydrometer.
Al> You are presuming that I use a hydrometer to detect the end of fermentation and this is very rarely what I expect to get from it. I use the hydrometer to measure OG (as you later suggest in your post) and the SG at bottling time, and to then later figure out what the apparent attenuation was. Twice, it has saved me from bottling a batch suffering from a stuck fermentation (both cases which would have been also detected by the Clinitest, but because I record the SG at bottling as a rule, I have to make this measurement anyway). Once was a Barleywine whose yeast pooped-out with only 60% apparent attenuation and the other was a Tripel that was fermented far too cold and the yeast went into hibernation once their activity (and therefore the heat they generated) decreased, resulting in a 50% apparent attenuation. In both cases the beer was saved (in the former, more yeast was added; in the latter the fermenter was warmed and swirled; the fermentation restarted in both cases).
Al> I bottle (or keg) when the beer clears. I double-check that I got a reasonable apparent attenuation before bottling, but I (usually) know my yeasts, pitch BIG and use oxygen, and therefore I rarely need to rely on the hydrometer to tell me that the apparent attenuation was lower than it should have been. Since I'm taking an SG reading anyway, I would not benefit from the use of the Clinitest even if it were 100% accurate.
Dave> This antediluvian object responds to sugar content, dextrin content, bubbles on the instrument body, the height of the meniscus and the temperature. The reading is obscured by the foam and beer color, takes 80 mls or more of beer and the reading is dependent not only on the FG but the OG because of the alcohol content. Cleanliness of the instrument as well as the sample holding tube can affect the reading.
Al> I don't have bubbles on the instrument because I don't even think about using it until fermentation is over. The proper way to read a brewing hydrometer is to ignore the meniscus and look at where the beer level *would* cross the scale if the meniscus were not there, so that's a moot point. I have made Imperial Stouts that are impervious to high-beam halogen headlights, but yet did not have any difficulty reading the hydrometer, and since I transfer the beer into the hydrometer test jar gently, I don't have foam on top -- the beer is uncarbonated and if it creates a head, you were either too rough on it, or it is still fermenting. I measure OG and FG. I'm aware that FG is influenced by alcohol, but that does not change the fact that I'm reading the FG. I am also aware that I'm calculating the apparent attenuation and not the real attenuation, so I simply ignore the alcohol and think of it as one of the compounds that contributes to the FG (albeit negatively).
Dave> The hydrometer is hardly a dependable instrument, since it has all these errors which can easily show up in the reading. Something as simple as a 10 deg F difference from the calibration temperature will cause an error larger than the day to day difference being read to determine the end of a fermentation. Bubbles of course can make the hydrometer rise right up and give you the impression that your SG is going up as the fermentation ends! ( and a past HBD contributor has commented on his consternation when observing this). This makes the hydrometer totally useless for distinguishing a stuck fermentation from a finished fermentation - *especially* if it is a new recipe. And particularly for an inexperienced user.
Al> Again, I don't recommend using the hydrometer to determine the end of fermentation. I use a table which tells me how much to compensate for the measurement temperature, so that's a non-factor. Again, I don't measure the FG of a fermenting beer, so there are no bubbles. If it's a stuck fermentation the fermentation will be over, right?
Dave> All fermentable sugars of interest to us are reducible sugars, except sucrose. Lactose is a reducible sugar, but not fermentable by S. Cerevisiae. So what does this mean? There are potential errors in the use of the Clinitest method if used improperly,as I have often pointed out and as in any method.[snip]
Dave> I have never had a beer test out at more than 1/4% sugar when it was finished fermenting even for very high dextrin beers which I routinely make. This 1/4% represents less than a division on a hydrometer and is much better than can be reliably be read by the brewer using a hydrometer under field conditions. Most often the Clinitest results are none or <1/4%. This is a practical estimate of the error.[snip]
Al> Again, you are presuming I'm using the hydrometer to detect the end of fermentation. I *only* use the hydrometer to measure the OG and for calculating apparent attenuation... I don't take daily samples and try to determine when the attenuation has stopped. I calculate the apparent attenuation, consider the recipe (how dextrinous the wort is or if I've added malto-dextrin or lactose), check the range of apparent attenuations for the yeast I'm using and either bottle (or keg) or take corrective action (because the fermentation is over, the yeast has settled and if the apparent attenuation is not high enough, I need to do something).
Dave> in fermentable content because the sucrose is not detected by Clinitest. Use it *incorrectly* and your error could be as high as say 10%.
Dave> Sounds horrible, but this is just about the size of the error used in estimating alcohol content with an hydrometer, anyway, isn't it? (5%+/- 0.5) i.e 0.5X100/5 = 10%.
Al> I don't follow where the 0.5 came from. With the precision, narrow-range hydrometers I use (something like $20 each at Cole-Parmer or other lab supply), I can read 0.0005. Still, let's say I'm within 0.001 on both the OG and FG. If the OG was, say 1.050 and the FG was, say 1.012, if I were calculating the alcohol content (ABV is then roughly 1.050 - 1.012 times 131), the actual alcohol content would be about 4.99% ABV. Had I misread by 0.001 high on the OG and 0.001 low on the FG, I would have calculated 5.24% ABV, which is only 5.01% off.
Al> However, I rarely care what my alcohol content is, because I usually estimate it as my OG points divided by 10 (1.050 == 50 points == 5.0%... close enough). On the other hand, since I primarily use the hydrometer to calculate the apparent attenuation and because I'm only interested the beer is within the expected range or if it is grossly outside it, even 10% accuracy is enough for me. The precision comes in handy when testing new yeasts and calculating extraction efficiency.
Dave> All of the sucrose (possibly from the addition by the brewer or extract manufacturer or the approx 3-5% from malt) has been converted by the yeast's extracellular enzyme 'invertase' and the sugar concentration is within the range of the test. These inverted sucrose sugars, if any are left, will be detected by the Clinitest. The hydrometer is nearly useless in this region since its broad reading gives indistinguishable results for nearly the same sugar content in the 0-2% range, especially if it is interfered with by any of the many above variables.
Al> What you are missing (although I posted it in my second or third post on the Clinitest, over a year ago) is the fact that yeast eat the sugars in order of complexity. I'm not sure if they release the invertase before or after eating the maltotriose, but they will certainly have eaten all the glucose, fructose, and maltose long before you should be even considering to determine whether or not the fermentation is complete. If, however, yeast eats maltotriose *before* it releases the invertase, then we could have sucrose in the wort at a time when the yeast is slowing down (an eager brewers might be thinking about calling it done and therefore starting to measure the sugar content with the Clinitest), *BUT* still have sucrose in the wort which is very fermentable *AND* does *NOT* react with the Clinitest.
Al (quoted by Dave)>> but I believe that AJ's post clearly pointed out that the Clinitest reacts with some non-fermentable sugars [This is the post prefixed with "AJ>" above.]
Dave> AJ did agree with my past posts that lactose is non-fermentable and is reducible.
Al> I was referring to dextrins, but lactose is imporant too... read on.
Al>> This is what I've *thought* all along yet
Al>> had no proof, so I posted simply that I was "skeptical" of its usefulness.
Dave> You only had to read my past posts both public and private to *clearly* understand this.
Al> No... I wasn't sure if there were unfermentable sugars that react with the Clinitest. That was my biggest question... we still don't know if sucrose is eaten before or after maltotriose and the other bigger fermentables.
Al (quoted by Dave)>> If you always brew the same recipes and always use the Clinitest, you can get a feel for what level of sugar (as reported by the test) is expected for each recipe. However, I've only brewed the same recipe perhaps 6 times. The other 150+ beers were in 30+ styles from 150+ recipes. Given this kind of variation in dextrins from batch to batch, I don't think it's practical to try to use Clinitest for predicting anything, *expecially* how much fermentable sugar remains in a partially fermented batch (i.e. adjusting down priming rates to account for unfermented, but fermentable sugar).
Dave> Whoa!!. You've got that backwards. It's the hydrometer that is totally useless in the above scenario. Clinitest always lets you know when the batch is fermented out *regardless* of the FG. This is the real strength of this method. No guesswork. I repeat again, in my years of experience with using Clinitest, I have never had a Clinitest reading of greater than 1/4% sugar at the end of a fermentation. ( Actually I don't remember one this high).
Al> No, I have it right... I'm not using the hydrometer to determine the end of fermentation (except, as I said before, to see that my 1.100 OG Barleywine should have attenuated better than an FG of 1.040 or that my 1.070 Tripel isn't ready to bottle at 1.035!). Read again the last sentence that I wrote... I still contend that the varibility in the Clinitest due dextrin content is enough to make using it to *adjust priming rates* foolish.
Dave> For special cases like lagering under pressure, I routinely test the beer and put it in the Cornelius keg when its sugar content is below 1%, but it is not completely finished fermenting. Using this method of testing to adjust priming sugar is fine, but as I have commented elsewhere you get a lot more yeast in the bottle, since the beer has not been to the secondary for settling. I have done this only a few times under time pressure, but successfully.
Al> In a keg, you can have a little overcarbonation and you can fix it (over a week of releasing pressure, perhaps). In bottles, it can go from flat to a gusher with a very small difference in sugar. [Furthermore, presumably you will be drawing from the keg periodically so that any excess pressure that builds will simply be used up for dispensing and the CO2 tank will get a break.]
Dave> If the brewer added lactose before fermentation ( why?) then this will have to be taken into account. Were this the case, it would represent the same problem for the hydrometer method in finishing "high". In either case, successive readings would show that there was no change and the Clinitest method is no worse that the hydrometer in this highly unusual and concocted case. The Clinitest still uses only a small sample and is easier to use without all that cleaning up.
Al> When did I say that one should take successive readings? Never. I've never advocated that method. Others have suggested this in books, so you presume I'm using this (useless, to me) method? Oh, and regarding adding lactose pre-fermentation... this *is* what I recommend because if you have some bacteria or wild yeast in your ferment and you add the lactose (which these nasties *can* eat) at bottling time, I guarantee gushers and would put money on exploding bottles.
Dave> For all those struggling all grainers who have been taking 100 ml or larger samples at the end of your sparge, cooling them while continuing the sparge, measuring the specific gravity to find a value less than 1.010,and finding you are reading a result that does not reflect current results of the sparge stream, try the Clinitest as an *indication* of the sugar content at the end of the sparge. It's not as perfect (bearing in mind the above comments)
Al> It may be useful in this capacity, but I don't even use a hydrometer to determine the end of the sparge. I stop sparging when I've taken enough runnings (usually 17 gallons, if I'm not doing parti-gyle).
Dave> Al, I respect nearly all of your opinions and read them carefully, and I cannot believe you have never tried Clinitest, yet you have all these dark, unfounded suspicions about it.
Al> I read your posts and emails carefully and value your opinions also, but I must comment when I feel there is information that can lead people into trouble. You'll note that I simply mentioned that I was skeptical when you initially brought it up. I didn't make a big fuss like this post until you started saying that you can reliably *adjust priming sugar for partially fermented beers* using the Clinitest. I feel this method can get a brewer in the exact same trouble as your suggestion to prime with a randomly partially-fermented kraeusen wort.
Al> I think you have a lot of good ideas and you certainly know a large portion of the science behind it all, but sometimes you leave out a few little pieces (which I'm sure you consider trivial) which I believe are important. There are a handful of things you continue to post that I disagree with (like the infection problems from blowoff tubes or or your seat-of-the-pants kraeusening technique) and I'll keep posting rebuttals.
Dave> I suggest the only way you will be satisfied is to actually go out, buy one and try it in parallel to your current method and draw your own conclusions instead of writing about something of which you admit to knowing nothing.
Al> As I've noted above... even if the Clinitest was 100% reliable, I would not use it because I rely on yeast sedimentation and not SG drop or sugar content to decide when to bottle/keg. Even if it was possible to use it reliably to reduce priming sugar (and it may be... I don't have the time to do the research), what would this buy us? We could bottle/keg three days sooner, but we would have to condition it three days longer till the green beer flavours were removed by the yeast. You can't rush the yeast.
In a private discussion, Spencer Thomas and I discussed the usefulness of Clinitest... I'll just quote my response to Spencer which highlights some of my biggest reasons for being skeptical of Clinitest:
Al> I still contend that the use of Clinitest is unproven and I'll continue to remain skeptical until someone with far more time to do experiments and far more accurate scales/flasks/etc. than I will verify the usefullness and consistency of these tests. For example, is a drop an accurate measure of volume? Doesn't it vary significantly with SG? How does alcohol affect this test? All unanswered questions. I'm not saying it will not work... I'm saying I have yet to be convinced that we know enough about it to be basing procedures upon it, ***especially*** using it to bottle while there is still some residual fermentable sugar in an effort to use this sugar for carbonation.
Most recently, in Homebrew Digest #2787, Dave writes:
Dave> On a different subject, without offering a solution, AlK said:
"As for when you should rack out from under this pancake of yeast,
the question is: "is the beer done?" Better put: "Have I gotten
enough attenuation?" Part of this is determined by knowing what
your expected apparent attenuation is for this yeast and calculating
it. No fermentation is perfectly clean and if you leave too much
fermentable sugar in the beer, you'll have gushers eventually."
Dave> Forced fermentations may offer a clue to when the fermentation will be furnished, but are subject to indeterminate errors. Remember that this same forced fermentation method is also used to determine the stability of the beer, as a check for unwanted bacterial fermentation. Bacteria function better at elevated temperatures, so likely any even slightly infected beer will give an FG dependent on the bacterial content as contrasted to a fermentation that was carried out at cooler temperatures.
Al> Good point. However, I wasn't suggesting using a forced fermentation... I believe it was George that suggested it. I do believe that fermeting the wort with a very large dose of yeast at say 75 or 80F would not give a radically different FG than could be expected in the main batch. I don't believe that forced fermentations are by any of the Industrial Brewers to determine stability, because they all pasteurise their beer.
Dave> AlK also brings up the other issue that certain yeasts like Ringwood and Samuel Smith's require rousing to bring the beer to completion. This implies of course that the fermentation is vessel and agitation dependent as well as dissolved CO2 dependent and therefore temperature dependent. Ergo, the smaller test bottle at a higher temperature may give a different result from the larger vessel. Finally, there is a need to bring the beer back to near the temperature of the brew, so that a highly accurate FG can be determined. As we have often pointed out in the past, accurate determination of the FG of a beer full of CO2 is not hydrometry at its finest. In fact, it is full of errors as a method of determining the FG because of the bubbles of gas that cling to the hydrometer. Worst of all these errors are in the wrong direction, since the readings of the hydrometer will be higher and lead the brewer to conclude that the main batch is finished when it is not.
Al> If the beer is still and the yeast is already settling (that's the only time I suggest using a hydrometer... I *never* suggested using it on actively fermenting beer) there should be no bubbles forming on the body of the hydrometer unless it is dirty. Dirt on the hydrometer is not a problem that is inherent to hydrometers in general and should not be used as a wholesale condemnation of hydrometers.
Dave> AlK also suggests that we, who make many different kinds of beer, actually know what the FG is ahead of time. If we were Budweiser or used only extracts, this could be an acceptable method, perhaps, but we aren't and don't. The whole point of this hobby for most homebrewers is to make all kinds of beers with various ( intentional or accidental) mash temperature profiles and yeast that behave differently. As a result, we almost never know what the final FG will be, since we have no experience. Given this, how do we know what the expected FG is supposed to be?
Dave> Actually you don't care what the FG is supposed to be, but as AlK points out to know "Is the beer done?" This is another way of saying "Are there any more fermentable sugars in the beer?"
Dave> Forced fermentations are probably very useful to determine if your beer will have a shelf life (most often we don't care) and to evaluate your basic cleanliness in your preparation. However, I believe they are a cumbersome, 1800s, time consuming and full of error method of determining a fermentation endpoint. Worst of all these errors are indeterminate and vary from batch to batch.
Dave> Clinitest is superior to this forced fermentation/hydrometer method for this purpose of determining "Is the beer done?", since it directly determines the state of sugars in the beer and not the specific gravity which is only marginally related to the (total) sugar content. One minute to run this simple test and you know exactly the state of the fermentable sugar content of your beer at that point in time. Visit your local pharmacy for a Clinitest Kit and actually know the current status of the fermentation near the endpoint without having to wait days for a suspect result. If you use Clinitest you will never have to worry about gushers
Al> I find it bizarre that Dave would continue to suggest that Clinitest is superior, *years* after AJ's post [quoted at the top of this file]. Both hydrometers and Clinitest can be used to determine grossly underattenuated batches (i.e. "stuck" fermentations), but for any finer measurement, for an idea of whether this particular batch finished a few SG points high, to more accurately calculate your % alcohol or to determine your apparent attenuation, a hydrometer infinately superior to the Clinitest. Determing the expected apparent attenuation is not without some potential for error (so I'm not about to say that the hydrometer is not without faults)... it involves finding the expected range of apparent attenuations for the yeast you used (available from the yeast manufacturers and in some homebrew books) and compensating for high levels of unfermentables (from lactose, malto-dextrin, crystal malt (which adds *some* unfermentables), etc.).
Al> Clinitest perhaps has its place in determining if you have a grossly underattenuated batch without needing a large sample as you would with a hydrometer, but it is far from the panacea Dave wants us to believe it is.