Hydrometer vs Refractometer for OG

[Carnot]

Here is my take on refractometer vs hydrometer which was posted on the Mead topic.

Essentially the correlation of SG: RI is dependent on the calibrating sugar. For accuracy then it would be necessary to create a scale based upon the wort being processed i.e. take hydrometer readings and RI readings to create a scale as the fermentation proceeds. Remember that the calibration was made for various sugar (sucrose) : water concentrations, and not for sugar, water and alcohol mixes. As the alcohol content builds the RI will change. I think that this is where the greatest source of error arises. The type of beer, mead, cider, wine will also have an effect.

There is a belief among bee keepers that the water content of heather honey is higher, due to the refractive index being higher. As I stated I am not convinced about this as using the refractive index as a means to determine water as it depends on a close correlation of the RI:water scale to heather honey. All of the data I have seen involves the RI:water scale being based on sucrose loadings in water . i.e sucrose was used as the calibration substance. This might be applicable to the food industry for estimating sugar or water content in foodstuffs but for honey it is at best indicative as honey does not contain a great deal of sucrose; in other words stretching reality. In order for it to be accurate a correlation of RI:water would have to be made for each type of honey which is well beyond the scope of bee keepers, since you would have to have an accurate means of moisture determination . That means an expensive Karl Fischer analyser. RI measures optical density which can be correlated to water in a pure substance like sucrose. With honey I am not so sure. The thixotropic nature of heather honey is not indicative of higher water content but due to gel properties. Gels are formed by dilute cross linking, possibly due to proteins in the honey.

Likewise I do not put much faith in refractometers over hydrometers in must sugar content estimation. Depending on what you are brewing, cider, mead,beer, etc the sugars in the must are far away from the sucrose that was used to produce the scale to estimate water (or sugar). I had very poor correlations on correlating the Brix: SG with an ale must (Courage Directors) such that it was next to useless. In my opinion the hydrometer is probably best.

[MTW]

^^^Yes, that. What problem are we solving by using a refractometer over a hydrometer for OG? Hydrometers/saccharometers work well. As far as I know (please correct me if I'm wrong!) HMRC would not (knowingly) allow ABV calculations based on refractometer readings in a commercial setting, for good reason. I'm happy to spare 100ml of a 20L+ brew for an accurate OG figure and to save me loads of preparatory work making sure I'm adjusting the refractometer reading correctly.

Occasionally, I have used the refractometer to see if fermentation is continuing rather than drawing off several jar samples. I totally disregard the reading displayed at that point, other than if it moves the next day, it is still fermenting. That, and as a rough indication for runnings and pre-boil SG... at least for beer brewing.

[Aleman]

Actually HMCE do allow ABV Calculations based on refractometers. (the phrase is "a calibrated measuring device" )

If they are that bad, why do people who use both the refractometer, AND HIGHLY ACCURATE HMCE NARROW RANGE HYDROMETERS, actually report accurate calibration between the wort brix value and the hydrometer, especially when using an appropriate correction tool?

I've said it before, the vast majority of home brew refractometers are about as accurate as allowing a monkey to take a gravity reading with a blunt stick! Things are different when using say a 0-1050 hydrometer with a 0.5 degrees scale or a 0.990-1.020 with a 0.2 degree scale, but a .0990 - 1.200 with a 2 degree scale?? guess work, especially considering the vast majority won't actually use it correctly either.

[Carnot]

In short a refractometer will probably work fine as long as your RI: gravity table has been constructed from the same product that you are examining. If you make a beer/cider/wine from consistent ingredients to the same recipe I would expect that a table could be constructed that shows near perfect correlation. It is when you use a table calibrated for brew A for Brew B. This would be especially relevant when you change the beer type/ wine type/ cider type. The comment about hydrometers is especially relevant.

[Carnot]

In short a refractometer will probably work fine as long as your RI: gravity table has been constructed from the same product that you are examining. If you make a beer/cider/wine from consistent ingredients to the same recipe I would expect that a table could be constructed that shows near perfect correlation. It is when you use a table calibrated for brew A for Brew B. This would be especially relevant when you change the beer type/ wine type/ cider type. The comment about hydrometers is especially relevant.

[Kev888]

For me, the sample size is the key reason for considering a refractometer - it is conveniently small and therefore far quicker to cool when measuring otherwise hot wort so i use it extensively throughout brew-days. Yes it requires a wort correction factor for greater accuracy but that is hardly some insurmountable mystery; the results then correlate with hydrometer readings to any degree of accuracy I'm ever likely to need. I could therefore use either, but may question what problem the hydrometer's less convenient sample size is solving

But to a degree I'm playing devils advocate there; I use the refractometer mostly for convenience and speed - in cases where the picture is changing, slow results or improperly cooled samples can bring inaccuracy. My main issue is that the refractometers most of us homebrewers use are fairly cheapo things, which even if calibrated tend to be less repeatable. Mine also struggles to show the results of dark wort crisply enough. How good many of the cheap hydrometers are is also debatable of course, but I think it is easier to make a decent hydrometer cheaply than a decent refractometer. For any important readings I find it is worth taking a few separate samples for validation, or using a trusted hydrometer.

HMRC calculations are a little different; these relate to attenuation and so there is alcohol present. Both hydrometers and refractometers are skewed by alcohol, neither give accurate readings of actual/real attenuation and need correcting to measure this. The difference is that hydrometer's Apparent Attenuation is historically common enough that many ABV calculations (I believe including the HMRC's standard ones) are designed to work directly with it rather than with true attenuation. But there is no reason why a refractometer can't be re/un-corrected to correlate with expected hydrometer inaccuracies; if you take many samples during fermentation this can save quite a lot of wort, or reduce contamination risk of returning samples.

With all the above going on, and that ABV calculations are themselves just estimates, it may be worth keeping in context what we really expect or need when talking about accuracy of our measurements.

[MTW]

Fair enough point from Kev regarding both instruments being subject to some correction for ABV, not showing actual attenuation as such.

Actually HMCE do allow ABV Calculations based on refractometers. (the phrase is "a calibrated measuring device" )

All the micros I've been in use saccharometers; how many breweries do you know who use a refractometer for OG? Sure, the sample size saving is less relevant, but if it were that easy and accurate...

If they are that bad, why do people who use both the refractometer, AND HIGHLY ACCURATE HMCE NARROW RANGE HYDROMETERS, actually report accurate calibration between the wort brix value and the hydrometer, especially when using an appropriate correction tool?

Because it is possible, just unnecessary, but take your choice.

I've said it before, the vast majority of home brew refractometers are about as accurate as allowing a monkey to take a gravity reading with a blunt stick!

My point entirely: My OG hyrometer cost a few quid. I've had it in the same jar alongside £80 brewery instruments and it is surprisingly good, but then it would be; the science of floating/sinking in a certain density fluid is pretty simple to this end. I don't want to spend out on an expensive refractometer, then build an index of known worts, only to get the same result.

Things are different when using say a 0-1050 hydrometer with a 0.5 degrees scale or a 0.990-1.020 with a 0.2 degree scale, but a .0990 - 1.200 with a 2 degree scale?? guess work, especially considering the vast majority won't actually use it correctly either.

Yes, spend out on a decent finishing hydrometer. Seeing the last few points or part of a point stop is clearer than on a refractometer and my 1-point 0-20 is good enough for that.