More Refractometer Woes

[Clint Yeastwood]

I had a problem with my cheapo refractometer giving conflicting readings while measuring wort taken from my kettle right after the boil. It seemed like stirring the wort made things worse.

Today I used a hydrometer to check a beer that is finishing up fermenting. On brew day, I used the hydrometer at 93 degrees to get a corrected reading of 1.074, and I used a calculator to correct the wort to 1.084 using DME. Today the hydrometer gave me 1.018 (corrected) from both fresh and degassed beer. Beersmith estimated 1.020, and no CO2 is being produced, so this beer is probably done.

I decided to try the refractometer because I want to make sure I'm not the problem. I checked fresh and degassed beer, and I got 11 Brix, or about 1.043. I took 4 readings with different samples, and the figure was always exactly 11 Brix. This puts me at a corrected 1.022 according to Brewer's Friend, with a wild guess of 1 as the correction factor because I don't have 30 wort samples here right now to make a spreadsheet.

The refractometer is clean. It reads 1.000 with tap water and purified bottled water. I don't have any distilled. I used it carefully. I can't find anything physically wrong with it. I think this is as good as it gets. Whatever the actual SG is, the refractometer is totally consistent today.

My question: what the hell? I had huge swings on brew night, and I did not give the wort time to evaporate. Maybe the refractometer's temperature varied a little, but a few drops of wort couldn't affect it a lot, and the temperature of the wort was about the same for all readings anyway.

I think I need a new instrument. The Kegland Saber makes looks great, but I am so frustrated, I am thinking I may go nuclear and get a Smartref or Easydens and get it over with.

Anton-Paar says the Smartref is accurate to within 0.4 Brix (+/-0.2). It says the Easydens has an accuracy of 0.3 Brix, but it doesn't say whether that's the total range (+/-0.1.5) or the possible deviation from the true value, which would give a range of 0.6 Brix. Based on other text from Anton-Paar, it looks like it's +/-, making the Smartref more accurate.

One Brix is around 4 SG points, so whatever the specs mean, the accuracy of both devices is good compared to the junk most of us use.

They say to use both the Easydens and the Smartref to get ABV. I don't really get that. My understanding is that if you really know your OG and FG, you know your ABV, and you only need a Smartref to get these numbers. It sort of looks like you only need the Easydens if you don't have an OG, but what homebrewer would buy a Smartref and not take OG's? If I misunderstand, maybe someone can chime in.

The ABV figures obtained using both devices together (which will never happen in my kitchen) are only accurate to 0.5%, and I don't know whether that's +/- or what. It's not impressive.

Reading up, I get the impression that (FG - OG)x131.25 is about as accurate as the Smartdens/Easydens combo, so I am not sure why I can't just get a Smartref and do the obvious math. I don't plan to go to the grocery and test factory beers.

If I can get within 0.5% using the old formula, it's more than good enough. I don't really need to know my ABV anyway.

 

[DBhomebrew]

a wild guess of 1 as the correction factor because I don't have 30 wort samples here right now to make a spreadsheet.

That would be a great wild guess for a straight sucrose solution. 1.04 would be a better one for beer.

No need to test 30 wort samples. Each wort, wheat vs 100% pale ale vs 10% invert bitter vs cream ale, would have different WCFs. The tool doesn't have a WCF all by itself.

Your getting spot on consistency today leads me to believe your sampling on brewday isn't up to snuff. On that note, do you have a measured original gravity in Brix? If I read you correctly, it seemed like your OG was calculated, but not necessarily measured.

And yes, an accuracy of x typically means a total possibile spread of 2x.

 

[Hoochin'Fool]

My question: what the hell? I had huge swings on brew night, and I did not give the wort time to evaporate. Maybe the refractometer's temperature varied a little, but a few drops of wort couldn't affect it a lot, and the temperature of the wort was about the same for all readings anyway.

If I understood that correctly, you made wort, measured it hot (but "corrected for that temperature") to be 1.074, topped it off DME so that you should have had a starting gravity of 1.084. Did you re-measure the gravity before fermenting started?

Next, you measured the final gravity with a hydrometer to be 1.018, and you want to compare to what your refractometer (corrected for alcohol by OG) says the gravity should be. You're seeing 11 brix, and the calculator says that should be 1.022, and are not happy with the 4 point difference in final gravity measurements, correct?

If I've got that right, allow me to suggest using 1.04 as your "Wort Correction Factor"! I read a bunch of posts on the internet saying that was pretty much THE value to use for beer purposes (and then later other people saying "it's not that simple"). All I know is, every time I've used 1.04 WCF *and* measured with hydrometer, they've been within 1 or 2 points difference.

 

[Hoochin'Fool]

After re-reading Clint's first post, I remember the original problem being small samples that are very hot lose significant amounts of water to evaporation while they cool, thus making it difficult to obtain an accurate (or even stable) Brix reading. Hence the resorting to using hot hydrometer (and correction), and being off enough that he needed to add DME to get his OG up where it was supposed to be.

I had better results using an eye-dropper to take a sample, letting it cool, then testing it on the refractometer. But then after rinsing it out, and about to take another sample later on, noticed there was some water that was still inside, luckily I had a small bag of eye droppers! Hardest part was getting the eye-droppers to stay vertical while cooling, so that they didn't leak out.

 

[Clint Yeastwood]

Thanks for the replies.

My notes are not perfect, but I believe I got 1.074 with the hydrometer and then added DME. I don't recall whether I measured the gravity after that. I probably did, but I don't have it written down. I was kind of fed up at that point.

I'm not particularly upset with a 4-point difference between the refractometer and hydrometer, because I have seen people test these instruments, and if what I saw was true, either instrument is doing well when it's only three or 4 points off the true value. What I'm unhappy about is that using the refractometer carefully with hot, unfiltered wort gave me nutty results.

It surprised me to learn that floating hydrometers were inaccurate. Every measuring instrument has some error, but these things can be off by 4 points, which is more than I would have expected. It's enough to make you miss your OG or misjudge your FG pretty badly.

Interesting thing: my floating hydrometer gave me exactly the same result with fresh and degassed beer. I can't find any good information on how much dissolved gas affects specific gravity, but it looks like it's not significant. I put a drop of Fermcap in the cylinder to kill the bubbles. Seemed to work. I have checked, and adding one drop of Fermcap to the cylinder has no measurable effect on the gravity.

I'm confident evaporation is not the problem with the refractometer. The drops go on, and the lens slams shut very quickly. I wonder about heat, though. I don't know how it affects the lens or the body of the instrument. Mainly, I wonder about the stuff that was floating around in the wort. If something heavy and transparent got under the refractometer lens, I would expect a higher index of refraction.

I've seen people suggest using wort that has been cooled. I went to Amazon and bought 100 new plastic droppers so I won't be using a spoon with a high heat capacity any more. It's pretty easy to cool a dropper down a few dozen degrees. I don't know if it will help. I used one of the droppers today, but the beer was at about 75 to start with. I may heat the beer and use a spoon again, to see if the results go crazy.

From where I sit, the Smartref looks like it makes more sense than the Easydens. Better accuracy and easier to use. I haven't figured out why a homebrewer would want an Easydens. Maybe someone can explain.

It's bizarre that a scientific instrument company would say an instrument had an accuracy of 0.3 Brix without saying whether this was a deviation or a range, especially when they were specific elsewhere. Inconsistent. I guess they expect customers to assume.

 

[Hoochin'Fool]

I'm confident evaporation is not the problem with the refractometer. The drops go on, and the lens slams shut very quickly. I wonder about heat, though. I don't know how it affects the lens or the body of the instrument. Mainly, I wonder about the stuff that was floating around in the wort. If something heavy and transparent got under the refractometer lens, I would expect a higher index of refraction.

I'll be brewing next week, a simple recipe I've thoroughly nailed down, so my pre-boil gravity and volume SHOULD be spot-on. Soon as my 60 minute boil happens, I'm going to test a few drops on the refractometer (placing drops and closing the lens cover ASAP), and record. Then, once wort is chilled, will retest again with refractometer. If I wanted to be as scientific as possible, I'm gonna use several tests:

• Drops of hot wart straight onto refractometer
• Cool sample in eye-dropper, then onto refractometer
• Substantial sized sample (I'm thinking a cup or so), covered with a lid and put in fridge to cool, then drops onto refractometer for test
• Stir up the big sample and check more drops on refractometer again, to see if stratification sets in very fast.
• Use rest of the big sample to fill hydrometer test tube.

Anything else I should record?

 

[Clint Yeastwood]

I seemed to get different results before and after stirring. You know how wort gets, with big clumps of stuff floating around.

 

[Hoochin'Fool]

I seemed to get different results before and after stirring. You know how wort gets, with big clumps of stuff floating around.

Not really! I put my hops in a paint straining bag! Only tiny clumps bits floating around! But I've added a good stir on the big sample for another test.

 

[DBhomebrew]

What I'm unhappy about is that using the refractometer carefully with hot, unfiltered wort gave me nutty results.

As mentioned in this batch's brewday thread, there are best practices to be followed for consistent and accurate refractometer readings. I see no reason to be upset about not best results when using not best practices. I definitely see no reason to drop a chunk of change on new equipment based on this experience.

I'm not particularly upset with a 4-point difference between the refractometer and hydrometer,

It's not a 4-pt difference. 1.022 ignores the need to use a wort correction factor. The rather good default of 1.04 brings the hygro/refract difference down to a mere point or two.

 

[Clint Yeastwood]

there are best practices to be followed for consistent and accurate refractometer readings.

What would you say I did wrong on brew day? Stirring the wort? Not cooling it? Using a metal spoon? I can't think of anything else. It's not an evaporation problem, unless two seconds are enough time for that to happen. I was using the same motions and tools all the time, so even if evaporation was the problem, it wouldn't make sense to get wildly different readings. Consistent mistakes should give consistent errors. Also, some readings were low, which isn't what you would get if water evaporated.

It's not a 4-pt difference. 1.022 ignores the need to use a wort correction factor.

I'm guessing you mean "the need to use a better wort correction factor." I did use a factor. Without the correction factor I used, the difference was 25 points, not 4. I looked up the process for getting a correction factor that works, and that's where I got the story about a spreadsheet and 30 wort samples. If all that work isn't needed, it's good news.

My thinking about this problem is colored by a comparison I saw. A brewing professional named David Heath compared various tools, and glass hydrometers and cheap refractometers were way off. Did he use them correctly? I don't know. I'm also wondering what happens if I start using the same correction factor with different recipes.

He got bad results with glass as well as a cheap refractometer, so now I don't trust my glass hydrometer. Using a factor of 1, I got a 4-point difference between the hydrometer and refractometer. With 1.04, I get a 2-point difference. But if I can't trust the glass hydrometer, what does it really mean? I could be correcting to match a spurious figure. Maybe 1.022 is right.

I have a 2002 hydrometer sold by True Brew. It has no error figure on the label. As a metrology instrument, this thing should error information included. I see why these things only cost 9 bucks. I got it on my first trip to a homebrew store. Just took what the guy handed me.

It kind of looks like most homebrewers are relying on bad instruments and bad figures, even if they do things right, because the instruments are not accurate. Are the instruments sufficiently consistent to make using correction numbers realistic? I don't know. I could do some tests and try to find out, but I would need fluids of known gravities, and I can't prepare those without good measurements, which I can't make without tools I know to be reliable. I guess I could make salt solutions using a reloading scale and distilled water. That would work for the hydrometer.

I looked at a bunch of websites to learn about refractometer technique, and I found conflicting information. No surprise there. They all seemed to think they knew all about the subject, but they disagreed, so some of the information was wrong. This ought to be cut and dried in 2023.

Not really! I put my hops in a paint straining bag! Only tiny clumps floating around!

Same here, but "big" may mean something different to me.

I don't even know what that stuff is. It settles in the kettle, and when I stir it, it rises into suspension.

This discussion makes me wonder how many old recipes are accurate, since the guys who wrote them may have used glass hydrometers that were off by 5 points.

 

[Clint Yeastwood]

I'm looking at the Smartref manual. It's called a Smartref because it's a refractometer, while the Easydens is a density meter.

The manual says to keep solids out of the sample

A PDF from the University of Bristol says a refractometer can be used to measure solids (protein) in plasma. Another site says the clumps in hot wort are coagulated protein.

Sort of looks like stirring is the big problem. It puts coagulated protein on the prism.

 

[DBhomebrew]

My process which includes best practices agreed upon by much more scientifically-minded forum members than I:

1) Allow refractometer (ATC even) to come to stable room temperature.

2) Fully stir entire wort.

3) Collect sample. I collect a few teaspoons into a 4oz ramekin. Cover! Set it on edge of plate so the solids collect in the corner.

4) Let cool to room temperature.

5) Take a fraction of a teaspoon of clear, room temp wort and apply to glass.

6) Read in Brix, record for posterity.


Using a WCF of 1.0 is in effect not using a WCF. X*1=X

Our refractometers are made for sucrose, not maltose, maltotriose, dextrins, etc. The WCF (1.04 as a good default) adjusts for the complex wort composition vs sucrose.

 

[DBhomebrew]

Good stuff here...

https://www.homebrewtalk.com/threads/refractometer-wort-correction-factor.703591/

 

[Clint Yeastwood]

I found better hydrometers online. Looks like +/-0.001 accuracy starts at $30.

 

[doug293cz]

It kind of looks like most homebrewers are relying on bad instruments and bad figures, even if they do things right, because the instruments are not accurate. Are the instruments sufficiently consistent to make using correction numbers realistic? I don't know. I could do some tests and try to find out, but I would need fluids of known gravities, and I can't prepare those without good measurements, which I can't make without tools I know to be reliable. I guess I could make salt solutions using a reloading scale and distilled water. That would work for the hydrometer.

Making your own calibration solutions is easy. All you need is water, table sugar, and an accurate scale.

Both °Brix and °Plato are defined as the weight percent of sucrose (pure cane or beet sugar) in water:

°Brix or °Plato = 100° * Weight of Sucrose / (Weight of Sucrose + Weight of Water)​

​

So, to make a 10°Brix or °Plato calibration solution, just mix 50 g of sucrose with 450 g of water (weigh the water, don't try to use volume.) For a 20° solution mix 100 g of sucrose with 400 g of water. Your solutions don't have to be any specific concentration, as long as you know the weight of sugar and water used, you can calculate the °Brix/°Plato.

If you need to calibrate a hydrometer, use the following formula to convert °Brix or °Plato to SG:

SG = 1 + (Plato / (258.6 - ((Plato / 258.2) * 227.1)))​

​

The above equation works up to about 28°P or 1.120 SG, but error starts to increase fairly rapidly above that.

Edit: If you want to use an online calculator, check that 20°Brix or °Plato converts to 1.08297 (1.083.) If it doesn't give this result, don't use the calculator.

Brew on

 

[Clint Yeastwood]

I thought salt would be better because its easy to dissolve, but apparently it holds a lot more water than table sugar.

 

[DBhomebrew]

Our refractometers are made for sucrose, not salt.

 

[doug293cz]

I thought salt would be better because its easy to dissolve, but apparently it holds a lot more water than table sugar.

No, you want to use only sucrose because that is how the scales are defined.

Brew on

 

[Clint Yeastwood]

Are you positive? Density is density, after all, regardless of the substance. I mean, specific gravity is the ratio of the density of the surrounding fluid to the density of distilled water. A salt solution of a given density has the same specific gravity as wort with the same density. Are you saying the temperature calibration requires a sugar solution because it expands and contracts differently from other solutions? That seems weird, because we use them to measure finished beer which may have very little sugar and enough ethanol to take them over 15% ABV. I haven't found anything suggesting hydrometer FG's have to be corrected for ethanol.

I'm looking at companies that make hydrometers. I notice they sell specific gravity hydrometers that measure the specific gravity of liquids, without reference to the types of liquids. It sounds like they expect them to work with just about any liquid. There are special salt hydrometers, but it looks like the difference is the units, not the function. Instead of measuring specific gravity, they give the weight or saturation as percentages in order to avoid calculations. If there are specific gravity hydrometers that are intended for general use, it seems natural to guess they could be used for both salt and sugar solutions. Just a guess, though.

It's amazing how many different kinds of hydrometers there are. I never really looked at them before. Different ranges. Different scales and accuracies. You could spend hours looking at all the varieties.

 

[faithie999]

Are you positive? Density is density, after all, regardless of the substance.

a refractometer isn't measuring SG per se, it's measuring the refractive index of the sample. the scale on the display is calibrated in degrees Brix for a sucrose refractometer, SG for a salt refractometer, etc (I have a marine aquarium so have a salt refractometer), etc. Just for fun, make up two calibration solutions--using the same weight percent of sugar and salt in each of the solutions, and see what the difference in readings are with your refractometer.

 

[Clint Yeastwood]

Sorry; I was talking about hydrometers. I must have misread what DBhomebrew wrote. I must need more sleep or glasses.

 

[doug293cz]

Are you positive? Density is density, after all, regardless of the substance. I mean, specific gravity is the ratio of the density of the surrounding fluid to the density of distilled water. A salt solution of a given density has the same specific gravity as wort with the same density. Are you saying the temperature calibration requires a sugar solution because it expands and contracts differently from other solutions? That seems weird, because we use them to measure finished beer which may have very little sugar and enough ethanol to take them over 15% ABV. I haven't found anything suggesting hydrometer FG's have to be corrected for ethanol.

I'm looking at companies that make hydrometers. I notice they sell specific gravity hydrometers that measure the specific gravity of liquids, without reference to the types of liquids. It sounds like they expect them to work with just about any liquid. There are special salt hydrometers, but it looks like the difference is the units, not the function. Instead of measuring specific gravity, they give the weight or saturation as percentages in order to avoid calculations. If there are specific gravity hydrometers that are intended for general use, it seems natural to guess they could be used for both salt and sugar solutions. Just a guess, though.

It's amazing how many different kinds of hydrometers there are. I never really looked at them before. Different ranges. Different scales and accuracies. You could spend hours looking at all the varieties.

Yes, you can calibrate a hydrometer with a salt solution, if you have information about weight percent of salt vs. SG. You can't calibrate a Brix refractometer with salt solutions. You can calibrate both hydrometers and Brix refractometers with sucrose solutions.

Brew on

 

[Clint Yeastwood]

I decided to get a Smartref. It sounds like it will get me pretty accurate gravities during brewing with a lot less hassle, and it should make less of a mess. It still uses a correction factor for fermented beer, but using the hydrometer to test beer is pretty painless.

 

[DBhomebrew]

Does it assure quality sampling?

 

[Clint Yeastwood]

I believe I'm still on the hook for that.

The PDF says: "For samples containing suspended solids it is recommended to filtrate [sic] the sample." I would guess I just need to avoid sampling right after rousing the crud on the bottom, but I don't know.

 

[DBhomebrew]

Post #12

 

[Clint Yeastwood]

I think stirring the wort is a big mistake. See above.

 

[DBhomebrew]

Oy

 

[Clint Yeastwood]

I don't want to get into an emotional exchange. Let's see what happens.

 

[IDMaynard]

Nice discussion on the limitations of a refratometer.
https://beersmith.com/blog/2010/11/02/how-to-use-a-refractometer-brix-and-beer-brewing/

 

[Hoochin'Fool]

I decided to get a Smartref. It sounds like it will get me pretty accurate gravities during brewing with a lot less hassle, and it should make less of a mess. It still uses a correction factor for fermented beer, but using the hydrometer to test beer is pretty painless.

I wish you luck with that. But that smartref device looks like it's going to be even more susceptible to hot-sample evaporation than your cheapo refractometer. Just my $0.02 tho

 

[Clint Yeastwood]

The company that makes it doesn't seem worried. Also, I now have a whole bunch of disposable droppers which cool quickly and don't permit much in the way of evaporation.

I don't think evaporation had anything to do with the problem. I think stirring up all that sediment was the real issue.

 

[DBhomebrew]

It might have been, but not stirring at all will cause another major issue. You must overcome stratification.

 

[Clint Yeastwood]

I think that is completely true.

 

[doug293cz]

It might have been, but not stirring at all will cause another major issue. You must overcome stratification.

I think that is completely true.

Stratification post-boil does not occur. Boiling effectively homogenizes the wort, and spontaneous stratification is not a thing in solutions.

However, if you add top-up water post boil, then you need to do something to homogenize the wort in order to get a valid SG reading.

Pre-boil, stratification is an issue if you sparge post mash.

Brew on

 

[Hoochin'Fool]

I'll be brewing next week, a simple recipe I've thoroughly nailed down, so my pre-boil gravity and volume SHOULD be spot-on. Soon as my 60 minute boil happens, I'm going to test a few drops on the refractometer (placing drops and closing the lens cover ASAP), and record. Then, once wort is chilled, will retest again with refractometer. If I wanted to be as scientific as possible, I'm gonna use several tests:

• Drops of hot wart straight onto refractometer
• Cool sample in eye-dropper, then onto refractometer
• Substantial sized sample (I'm thinking a cup or so), covered with a lid and put in fridge to cool, then drops onto refractometer for test
• Stir up the big sample and check more drops on refractometer again, to see if stratification sets in very fast.
• Use rest of the big sample to fill hydrometer test tube.

Anything else I should record?

So some test results from this morning's imperial stout...

Worst case: boiling wort, open spoon sample. If tested ASAP, got brix reading about 1.5 degrees higher than actual. After open spoon sample cooled, brix reading was 2.0 degrees higher than actual.

The eye-dropper was better: tested immediate (still hot), got brix reading about 0.5 degrees higher than actual. After cooling, brix was still 0.5 degrees higher than actual.

So "actual" was measured after the boil-kettle wort had cooled all the way down to 70F, and given a good stir. And it was 25 brix (as recipe software predicted).

Not sure how repeatable this is, my stove hits a VERY gentle boil, and lower gravity worts quite possibly would either show more or less effect from evaporation.

Each "brix degree" is about 4 gravity points

 

[Clint Yeastwood]

I hate metrology.