Hydrometer Errors

[ericmark]

Thought I knew how to use a hydrometer be it the meniscus, temperature, or simple spinning it to remove bubbles but today got caught out and realise it may have happened many times. Started with thinking reading rather low with beer ready to bottle so thought I would get one out of shed and measure that bottled on 3rd October 2013 so fermenting would have stopped and I would see what the final figure would get to.
Poured beer into measuring jar and popped in hydrometer spinning it to ensure free and all bubbles off the instrument. Reading 1.010 that seemed rather high spun it again yes that was the reading so went to find the records but recorded how much sugar and start s.g. of 1.044 but not finish s.g. but always measure and unlikely I would bottle it that high. Can't see tea spoon of sugar doing that much to final s.g.
So went back down to look at the hydrometer reading sitting at 1.010 so again spun it and reading dropped to 1.004. After maybe half an hour standing. So yes likely temperature will have rising but 6 - 20 is about 1 point so has to be simply the gas in the beer giving wrong reading.
Or could it be temperature got caught out once before with temperature I have two hydrometers one glass and one plastic both designed for beer and wine, and I found with temperature change one went up the other went down so it's clearly dependent on material the hydrometer is made from.
I had assumed the correction table given in John Palmer's web page was for glass however it does not say glass maybe that page is for plastic or even maybe different glass expands at a different rate?
Air locks and temperature strips it would seem are essential tools and only when both air lock and hydrometer readings both indicate ready to bottle is it a safe.
However 6 demijohns of lager in garage at around 12 degrees C I had been popping in and measuring s.g. been sitting at 1.014 for weeks now I wonder is it really 1.014 only way to measure is to warm it up to 15 degrees C and all other beers will need cooling to 15 degrees C to measure.
I seem to remember with anti-freeze and battery acid you could get an electronic unit and you put a drip on a plate and closed lid. Think it used refraction to measure s.g. can the same be done with beer with a simple scale an beam of light?

 

[rpt]

The correction for temperature is not for the hydrometer expanding but to allow for the density of the liquid changing. You can get refractometers but they aren't really suitable for beer due to the different sugars plus alcohol skews the reading.

 

[ericmark]

If the coefficient of linear expansion for glass and beer is the same (OK I know beer will not expand linear but know what I mean) then what ever the temperature the hydrometer will give correct readings. I am sure by combining plastic and glass a hydrometer could be made which would read correct at all temperatures.

However they do not expand at the same rate so hydrometer only accurate at 15 degrees C. Problem is we don't brew all beer at 15 degrees C lager is colder and other beers are warmer I have assumed with a glass hydrometer the error is not worth worrying about. However the Liverpool firm who made my hydrometer are no longer trading so there is no way to get a correction chart for my hydrometer.

Glass is 8.5 and Glass, borosilicate 3.3 (commonly called heat proof) the figures are not important, what is important is different glass expands at different rates so there is no one size fits all correction chart for glass hydrometers, it has to be supplied by the manufacturer of the hydrometer in question.

Also of course there is lead shot and some epoxy resin inside the hydrometer which may mean the glass is forced to expand or stopped from expending in the normal manor.

The upshot of it all means liquid has to be heated or cooled to 15 degrees C to get an accurate measurement. PVC is 156 and water 207 so in theroy the plastic hydrometer should be better than glass at 25.5 as closer to water in volumetric coefficient yet when I tried using my plastic hydrometer it seemed well out from what I expected in water cooling and heating the glass and plastic went in opposite direction so something wrong with my theroy.

The hydrometer has two roles, One tell us ABV level and two when ready to bottle with today's results in mind in the future I will look at the temperature and air lock activity more than hydrometer reading which I will now regard as approximate only.

Unless some one can direct me as to why the readings altered so much.

 

[rpt]

I always use my hydrometer as near to 20C as I can as that is it's calibration temperature. I still think it is the liquid's density that matters when correcting for temperature. I'd certainly go by the hydrometer rather than the airlock to check when fermentation has finished.

 

[ericmark]

Thank you for that answer since both of mine are calibrated at 60 degs F which is nearly 15 degs C I had assumed that was the standard calibration temperature from what you say you can get hydrometers calibrated at different temperatures so what I need to do is get one of the 20 deg ones like you have.

My hydrometer is for wine and there is a reference on the packet to a beer version with a smaller range so more accurate reading but in both brew shops I tried they looked at me daft when I asked for a beer hydrometer to replace my wine one with seems they have never stocked them.

After your post I find Density Hydrometers are at 20 degs and Specific Gravity Hydrometers are at 60 degs F it would seem Stevenson-Reeves make 4 versions with range 0.980–1.120, 0.990–1.170, and 0.970–1.150 it would seem if I am reading the info correct a GG2563/100 Stevenson-Reeves would measure 1.000-1.100 at 20 degrees and would be far more accurate for beer than what I have being 230 mm long, with 0.002 scale divisions but Stevenson-Reeves do not sell direct and so although they are clearly made can't find a suppler so 0.970–1.150 seems to be the one offered for sale and at less than £5 that does seem the way to go.

 

[Wasabi]

Temperature isn't really critical for hydrometer readings: the density change for +/-2°C is within the typical reading error for a hydrometer.

Digging out my old chemistry data tables reveals that to the nearest thousandth, the density of water at various temperatures is:
0-10°C = 1.000
15°C = 0.999
20°C = 0.998
25°C = 0.997

Given that most hydrometers have divisions at 2/1000 intervals, the density change from even a 5° temperature difference is only half a division and unlikely to be significant, things like bubbles and judging the bottom of the miniscus are probably bigger factors.

The other thing to consider if yuor readings are odd is that some hydrometers have the scale on a bit of paper glued inside the glass and this has been known to slip. If in doubt, just do a reading of plain water to check you get the correct value. (If the hydrometer is calibrated for 20°C, I presume this means it will give a reading of 1000 with plain water at 20°C rather than show the true density of 0.998? I'm not quite sure about that but mine gives 1000 for water at around 20°C anyhow.)

 

[ericmark]

Thank you yes I use other methods like air lock bubbles to judge when to bottle. It started when on transfer from first fermentor to a clean one near end of fermenting to reduce sediment I lost the syphon. As a result the dregs were transferred to a demijohn. When testing I got different readings demijohn to fermentor and so was looking at why with the idea maybe I needed to modify my methods. It would seem I did transfer too early however too things were giving me wrong readings. Temperature and not allowing bubbles to settle out before reading so most of the difference was due to poor reading of hydrometer on my part.

However these mistakes also got me thinking about the lager I am making. Unlike other beers it's first so really don't know what to expect. Now in clear demijohns I can see the fermenting which has for a long time been a steady but small stream of bubbles from the bottom. At 12 degrees everything very slow which was expected so hydrometer reading once a week last time still at 1.014 which was no change from week before.

However week before I had carried demijohns which were tested inside so the temperature tested at would vary between weeks and also I had not noted how long I had waited for any gases to dissipate before testing. Normally 0.002 in the reading would not worry me it would be close enough but with slow lager very different.

So next test was going to be 28th March so on 27th March one demijohn will be brought inside to warm up before testing the next day. What I realise is temperature is double edged as not only will it affect the hydrometer reading due to temperature alone but also the warmer the liquid gets the less gas it will retain so when any liquid with dissolved gasses is warmed those gases will be released. No temperature chart will compensate for dissolved gasses in the liquid. But at next test if I test a pair one at 20 degs and one held at 12 degs I should then have a good idea of how much the combined temperature change and gas release will have.