Is a hydrometer really better than a refractometer for measuring FG ?

[bobtheUKbrewer2]

Say we have an OG of 1.050 and a FG of 1.010 then the beer will be around 3.8% (for the purposes of this discussion). A 3.8% ethanol by volume solution has a specific gravity of 0.99451 at 20 deg C. Water has a SG of 0.998202 at 20 deg C so the ethanol is dropping the SG of the beer by 0.003692, so we need to add 0.0037 to 1.010 to get a close approximation, =1.0137

Call this 1.014 - now measure the FG with a refractometer and note that (say) BRIX 7.4 corresponds to SG 1.014

Repeat for a few brews of different strengths and you will have a conversion table that is accurate enough for home brewing.

Discuss please.

 

[Pkrd]

Say we have an OG of 1.050 and a FG of 1.010 then the beer will be around 3.8% (for the purposes of this discussion).

1.050 to 1.010 = 5.25% ABV

Arguably for homebrewers OG - 1 = ABV works well enough, unless you have attenuation problems.

 

[ESBrewer]

Say we have an OG of 1.050 and a FG of 1.010 then the beer will be around 3.8% (for the purposes of this discussion). A 3.8% ethanol by volume solution has a specific gravity of 0.99451 at 20 deg C. Water has a SG of 0.998202 at 20 deg C so the ethanol is dropping the SG of the beer by 0.003692, so we need to add 0.0037 to 1.010 to get a close approximation, =1.0137

Call this 1.014 - now measure the FG with a refractometer and note that (say) BRIX 7.4 corresponds to SG 1.014

Repeat for a few brews of different strengths and you will have a conversion table that is accurate enough for home brewing.

Discuss please.

As I undestand, you just compensate for the gravity of ethanol here. But you need to remember that ethanol will change the refraction index of the solution and thus makes it impossible to use the refractometer reading as such. A good program can calculate the expected change in refraction and give you a reasonable estimate of ABV in beer based on the reading. When using refractometer, you are not measuring gravity at all but refraction of light, which is related to the amount of sugar AND alcohol in wort. As mentioned above, 1.050->1.010 in gravity (= measured with hydrometer), is more than 5% ABV.

 

[BrewnWKopperKat]

Is a hydrometer really better than a refractometer for measuring FG ?

For me, a hydrometer works better for measuring FG. Note that I'm not concerned about how much beer is needed to take the measurement (I taste the sample, then dump it).

...

Repeat for a few brews of different strengths and you will have a conversion table that is accurate enough for home brewing.

Discuss please.

Conversion tables seem to have faded away with the wide-spread use of recipe software.

 

[bobtheUKbrewer2]

Pkrd - I was not trying to be precise "for the purposes of this discussion". I wanted the principle to be duscussed. 1.050 minus 1 is 0.050 - that is nowhere near the abv.

ESbrewer - I am asking if, knowing the refractometer and hydrometer final readings, can I just read off the abv in a home made table based on say 5 brews of varying strengths.

BrewnWKopperKat - do you add say 0.004 to your FG hydrometer reading ?

 

[dmtaylor]

Brix is dependent upon both OG and FG, so you'd have to tabulate all 3 side by side, OG, FG, and corresponding final Brix reading, or make different tables for different OGs. Sounds to me like a pain in the rear. All this work has already been done anyway.

If you determine your own Wort Correction Factor (mine is 0.99) and you measure everything right with calibrated gauges, then either of the following calculators by Petr Novotny will give you accurate FG (or even SG *during* fermentation) from a refractometer within +/- 0.002 of a calibrated hydrometer. I've been using this Novotny formula for 2 years now and very satisfied with my results.

https://www.homebrewmap.com/en/tools/calculators/refractometer-correction

https://www.brewersfriend.com/refractometer-calculator/ (this site uses the same Novotny formula)


LOTS more discussion here:

https://www.homebrewersassociation.org/forum/index.php?topic=28544.0

http://diversity-pivo.blogspot.com/2017/01/pocitame-nova-korekce-refraktometru.html


For the utmost accuracy, you'll always need to use a calibrated hydrometer. But if being within 0.002 is close enough, you can get there with a Novotny calculator.

With self-made tables................ I'll have my doubts on accuracy until you brew several hundred batches and can thus turn them into empirical formulae with low standard deviation.

 

[myndflyte]

Zymurgy did a good write up on refractometers and tweaked Sean Terrill's calculation. I've been using this along side my hydrometer and the calculation gets me within +/- 0.003 of my hydrometer pretty consistently.

FG = (-0.002349Bxi + 0.006276Bxf) + 1
ABW = 0.67062Bxi - 0.66091Bxf
ABV = (FG * ABW)/0.791

Bxi is brix initial and Bxf is brix final.

 

[dmtaylor]

Zymurgy did a good write up on refractometers and tweaked Sean Terrill's calculation. I've been using this along side my hydrometer and the calculation gets me within +/- 0.003 of my hydrometer pretty consistently.

FG = (-0.002349Bxi + 0.006276Bxf) + 1
ABW = 0.67062Bxi - 0.66091Bxf
ABV = (FG * ABW)/0.791

Bxi is brix initial and Bxf is brix final.

Yup, this is Novotny's linear formula. He basically started with Terrill's formula and improved/corrected it.

 

[myndflyte]

Yup, this is Novotny's linear formula. He basically started with Terrill's formula and improved/corrected it.

Oh, well that's good to know. Now I can stop calculating it by hand.

 

[Pkrd]

Pkrd - I was not trying to be precise "for the purposes of this discussion". I wanted the principle to be duscussed. 1.050 minus 1 is 0.050 - that is nowhere near the abv.

?
0.050 is indeed 5%

 

[BrewnWKopperKat]

BrewnWKopperKat - do you add say 0.004 to your FG hydrometer reading ?

No. I'll periodically calibrate the two hydrometers that I have. Between calibrations, if I get a unexpected FG reading with the finishing hydrometer, I'll measure with the other hydrometer.

Zymurgy did a good write up on refractometers and tweaked Sean Terrill's calculation. I've been using this along side my hydrometer and the calculation gets me within +/- 0.003 of my hydrometer pretty consistently.

It's in the July/August 2017 issue.

 

[TechFanMD]

?
0.050 is indeed 5%

You quick and dirty assumes that every beer ends at 1.010 which is far from the truth. I only like shortcuts and rules of thumb that are true even most of the time.

 

[dmtaylor]

You quick and dirty assumes that every beer ends at 1.010 which is far from the truth. I only like shortcuts and rules of thumb that are true even most of the time.

Not specific to FG of 1.010. The approximation is more closely related to apparent attenuation. It *is* pretty close to accurate for most yeasts of average ~75% attenuation. Try it for a while and see. It's not far off.

 

[TechFanMD]

Not specific to FG of 1.010. The approximation is more closely related to apparent attenuation. It *is* pretty close to accurate for most yeasts of average ~75% attenuation. Try it for a while and see. It's not far off.

I did try it. I found it to be wildly inaccurate because there are so many variables.....lower or higher attenuating yeasts (not uncommon), mashing higher for higher FG in certain styles. Not everyone is Mashing at 148-150 and pitching Notty (which would probably go lower than 1.010 anyway).

 

[balrog]

Last 10 brews, OG only varied between 1.050--1.055, FG varied between 1.006--1.018, the diff of using the simple FG-1 and calculating was within 0.2%ABV-calc half the time. It really is reasonably ballpark value. Good enough for horseshoes and hand grenades as they say.

 

[TechFanMD]

I guess I brew a lot that does not fall in that OG range and doesn’t necessarily fall in that FG range.

 

[dmtaylor]

I just went ahead and calculated the standard deviation of the quick & dirty ABV estimation method for my last 82 batches. Quick & dirty is good within +/- 0.7%, with average difference of 0.0% compared to brewing software. Data:

OG-----FG--Actual-Quick-Difference
1.054 1.011 5.6% 5.4% -0.2%
1.055 1.015 5.2% 5.5% 0.3%
1.082 1.026 7.4% 8.2% 0.8%
1.055 1.010 5.9% 5.5% -0.4%
1.064 1.022 5.5% 6.4% 0.9%
1.049 1.011 4.9% 4.9% 0.0%
1.073 1.015 7.7% 7.3% -0.4%
1.059 1.013 6.0% 5.9% -0.1%
1.043 1.010 4.3% 4.3% 0.0%
1.049 1.011 4.9% 4.9% 0.0%
1.082 1.026 7.4% 8.2% 0.8%
1.057 1.018 5.1% 5.7% 0.6%
1.055 1.020 4.6% 5.5% 0.9%
1.060 1.019 5.4% 6.0% 0.6%
1.060 1.017 5.7% 6.0% 0.3%
1.050 1.010 5.2% 5.0% -0.2%
1.048 1.009 5.1% 4.8% -0.3%
1.051 1.012 5.1% 5.1% 0.0%
1.061 1.014 6.2% 6.1% -0.1%
1.070 1.016 7.1% 7.0% -0.1%
1.052 1.013 5.1% 5.2% 0.1%
1.067 1.015 6.9% 6.7% -0.2%
1.065 1.019 6.1% 6.5% 0.4%
1.040 1.007 4.3% 4.0% -0.3%
1.051 1.020 4.1% 5.1% 1.0%
1.061 1.018 5.7% 6.1% 0.4%
1.069 1.018 6.7% 6.9% 0.2%
1.051 1.010 5.4% 5.1% -0.3%
1.069 1.012 7.5% 6.9% -0.6%
1.045 1.012 4.3% 4.5% 0.2%
1.061 1.012 6.5% 6.1% -0.4%
1.114 1.020 12.6% 11.4% -1.2%
1.058 1.013 5.9% 5.8% -0.1%
1.058 1.012 6.0% 5.8% -0.2%
1.048 1.011 4.8% 4.8% 0.0%
1.067 1.011 7.4% 6.7% -0.7%
1.093 1.011 11.0% 9.3% -1.7%
1.050 1.016 4.5% 5.0% 0.5%
1.051 1.011 5.2% 5.1% -0.1%
1.052 1.014 5.0% 5.2% 0.2%
1.048 1.017 4.1% 4.8% 0.7%
1.058 1.008 6.6% 5.8% -0.8%
1.073 1.012 8.1% 7.3% -0.8%
1.065 1.009 7.4% 6.5% -0.9%
1.057 1.013 5.8% 5.7% -0.1%
1.055 1.008 6.2% 5.5% -0.7%
1.072 1.019 7.0% 7.2% 0.2%
1.050 1.018 4.1% 5.0% 0.9%
1.056 1.009 6.2% 5.6% -0.6%
1.059 1.013 6.0% 5.9% -0.1%
1.080 1.021 7.8% 8.0% 0.2%
1.037 1.015 2.8% 3.7% 0.9%
1.087 1.021 8.8% 8.7% -0.1%
1.047 1.013 4.4% 4.7% 0.3%
1.047 1.012 4.5% 4.7% 0.2%
1.054 1.012 5.5% 5.4% -0.1%
1.054 1.021 4.3% 5.4% 1.1%
1.061 1.008 7.0% 6.1% -0.9%
1.075 1.018 7.6% 7.5% -0.1%
1.047 1.010 4.8% 4.7% -0.1%
1.057 1.033 3.1% 5.7% 2.6%
1.047 1.009 5.0% 4.7% -0.3%
1.059 1.010 6.4% 5.9% -0.5%
1.057 1.009 6.3% 5.7% -0.6%
1.051 1.011 5.2% 5.1% -0.1%
1.077 1.009 9.0% 7.7% -1.3%
1.113 1.045 9.1% 11.3% 2.2%
1.049 1.015 4.4% 4.9% 0.5%
1.053 1.012 5.4% 5.3% -0.1%
1.055 1.016 5.1% 5.5% 0.4%
1.054 1.021 4.3% 5.4% 1.1%
1.054 1.013 5.4% 5.4% 0.0%
1.074 1.014 7.9% 7.4% -0.5%
1.055 1.013 5.5% 5.5% 0.0%
1.055 1.014 5.4% 5.5% 0.1%
1.098 1.023 10.1% 9.8% -0.3%
1.046 1.008 4.9% 4.6% -0.3%
1.046 1.008 4.9% 4.6% -0.3%
1.038 1.016 2.8% 3.8% 1.0%
1.059 1.026 4.3% 5.9% 1.6%
1.062 1.013 6.5% 6.2% -0.3%
1.071 1.010 8.1% 7.1% -1.0%
average diff 0.0%
max diff 2.6%
min diff -1.7%
stdev 0.7%

 

[TechFanMD]

I just went ahead and calculated the standard deviation of the quick & dirty ABV estimation method for my last 82 batches. Quick & dirty is good within +/- 0.7%, with average difference of 0.0% compared to brewing software. Data:

OG-----FG--Actual-Quick-Difference
1.054 1.011 5.6% 5.4% -0.2%
1.055 1.015 5.2% 5.5% 0.3%
1.082 1.026 7.4% 8.2% 0.8%
1.055 1.010 5.9% 5.5% -0.4%
1.064 1.022 5.5% 6.4% 0.9%
1.049 1.011 4.9% 4.9% 0.0%
1.073 1.015 7.7% 7.3% -0.4%
1.059 1.013 6.0% 5.9% -0.1%
1.043 1.010 4.3% 4.3% 0.0%
1.049 1.011 4.9% 4.9% 0.0%
1.082 1.026 7.4% 8.2% 0.8%
1.057 1.018 5.1% 5.7% 0.6%
1.055 1.020 4.6% 5.5% 0.9%
1.060 1.019 5.4% 6.0% 0.6%
1.060 1.017 5.7% 6.0% 0.3%
1.050 1.010 5.2% 5.0% -0.2%
1.048 1.009 5.1% 4.8% -0.3%
1.051 1.012 5.1% 5.1% 0.0%
1.061 1.014 6.2% 6.1% -0.1%
1.070 1.016 7.1% 7.0% -0.1%
1.052 1.013 5.1% 5.2% 0.1%
1.067 1.015 6.9% 6.7% -0.2%
1.065 1.019 6.1% 6.5% 0.4%
1.040 1.007 4.3% 4.0% -0.3%
1.051 1.020 4.1% 5.1% 1.0%
1.061 1.018 5.7% 6.1% 0.4%
1.069 1.018 6.7% 6.9% 0.2%
1.051 1.010 5.4% 5.1% -0.3%
1.069 1.012 7.5% 6.9% -0.6%
1.045 1.012 4.3% 4.5% 0.2%
1.061 1.012 6.5% 6.1% -0.4%
1.114 1.020 12.6% 11.4% -1.2%
1.058 1.013 5.9% 5.8% -0.1%
1.058 1.012 6.0% 5.8% -0.2%
1.048 1.011 4.8% 4.8% 0.0%
1.067 1.011 7.4% 6.7% -0.7%
1.093 1.011 11.0% 9.3% -1.7%
1.050 1.016 4.5% 5.0% 0.5%
1.051 1.011 5.2% 5.1% -0.1%
1.052 1.014 5.0% 5.2% 0.2%
1.048 1.017 4.1% 4.8% 0.7%
1.058 1.008 6.6% 5.8% -0.8%
1.073 1.012 8.1% 7.3% -0.8%
1.065 1.009 7.4% 6.5% -0.9%
1.057 1.013 5.8% 5.7% -0.1%
1.055 1.008 6.2% 5.5% -0.7%
1.072 1.019 7.0% 7.2% 0.2%
1.050 1.018 4.1% 5.0% 0.9%
1.056 1.009 6.2% 5.6% -0.6%
1.059 1.013 6.0% 5.9% -0.1%
1.080 1.021 7.8% 8.0% 0.2%
1.037 1.015 2.8% 3.7% 0.9%
1.087 1.021 8.8% 8.7% -0.1%
1.047 1.013 4.4% 4.7% 0.3%
1.047 1.012 4.5% 4.7% 0.2%
1.054 1.012 5.5% 5.4% -0.1%
1.054 1.021 4.3% 5.4% 1.1%
1.061 1.008 7.0% 6.1% -0.9%
1.075 1.018 7.6% 7.5% -0.1%
1.047 1.010 4.8% 4.7% -0.1%
1.057 1.033 3.1% 5.7% 2.6%
1.047 1.009 5.0% 4.7% -0.3%
1.059 1.010 6.4% 5.9% -0.5%
1.057 1.009 6.3% 5.7% -0.6%
1.051 1.011 5.2% 5.1% -0.1%
1.077 1.009 9.0% 7.7% -1.3%
1.113 1.045 9.1% 11.3% 2.2%
1.049 1.015 4.4% 4.9% 0.5%
1.053 1.012 5.4% 5.3% -0.1%
1.055 1.016 5.1% 5.5% 0.4%
1.054 1.021 4.3% 5.4% 1.1%
1.054 1.013 5.4% 5.4% 0.0%
1.074 1.014 7.9% 7.4% -0.5%
1.055 1.013 5.5% 5.5% 0.0%
1.055 1.014 5.4% 5.5% 0.1%
1.098 1.023 10.1% 9.8% -0.3%
1.046 1.008 4.9% 4.6% -0.3%
1.046 1.008 4.9% 4.6% -0.3%
1.038 1.016 2.8% 3.8% 1.0%
1.059 1.026 4.3% 5.9% 1.6%
1.062 1.013 6.5% 6.2% -0.3%
1.071 1.010 8.1% 7.1% -1.0%
average diff 0.0%
max diff 2.6%
min diff -1.7%
stdev 0.7%

So it works well for beers that fall in that ‘normal’ range and not so well for beers on the fringes which for some reason mine tend to be. Your outliers are my normal.

 

[dmtaylor]

So it works well for beers that fall in that ‘normal’ range and not so well for beers on the fringes

Exactly.

 

[Pkrd]

3.8% is more precise than 5%

You also should factor in the inaccuracies associated with reading the gravities, and serving as well. I’m probably not going to get a fine or lawsuit if my pale ale is 5.6% instead of 4.9%, just a bigger headache. A commercial brewer might find themselves in real trouble.

 

[balrog]

3.8% is more precise than 5%

You also should factor in the inaccuracies associated with reading the gravities, and serving as well. I’m probably not going to get a fine or lawsuit if my pale ale is 5.6% instead of 4.9%, just a bigger headache. A commercial brewer might find themselves in real trouble.

'Xactly.

 

[bobtheUKbrewer2]

I started this thread to get thoughts on the principle. Sadly a lot of responses went straight into detail, eg "Say we have an OG of 1.050 and a FG of 1.010 then the beer will be around 3.8% (for the purposes of this discussion)" and somebody comes up with "1.050 to 1.010 = 5.25% ABV "

Anyways, I am going to purchase a high quality narrow range hydrometer and measure future OG & FG with that and a refractometer. If I can find a correlation that is within 0.2% abv I shall just use the refractometer.

thanks for all posts !

 

[dmtaylor]

My first answer above provides the correlation. No need to reinvent the wheel, that's my answer.

 

[balrog]

I started this thread to get thoughts on the principle. Sadly a lot of responses went straight into detail...thanks for all posts !

...but...this is the internet!

 

[Epos7]

Refrac reading plugged into the correct formula is very accurate, within 0.002 of my hydrometer for FG measurements. Check all your measuring equipment, always. I went through a few hydrometers before I found one that was accurate. The cheap ones weren't good.