Monday, May 31, 2010

How hot is that?

We've got a thermometer and we're not afraid to use it. In fact we've got a number of thermometers. Not one of them ever shows the correct temperature - I'm pretty sure - or if one does - it's an amazing coincidence.

The met office tells me that the temperature on Walney Island, which is where I did much of my growing up, is currently 14.3°C. Which seems quite parky. Which is why I've chosen to live a few miles inland (oh, and not in the middle of a windswept airfield). How accurate is that measurement? How precise? And what's the difference?

It's easy to use the terms "accurate" and "precise" as if they were interchangeable. They're not. Accuracy considers how near your measurement is to the true value, whereas precision tells us how close we can expect repeated measures made with the same instrument will be. I think. Unless I've got that the wrong way 'round.

So, if the met office is right (and here we mean right in a very godlike, absolute sense), and I went to Walney and measured 14.8, it would be fair to say that my accuracy was something like 0.5°C. If I made 4 more measurements really quickly (so as to be reasonably sure that the true temperature didn't change) and got 14.7, 14.8, 14.7 and 14.9, then I might say that my precision is something like ± 0.1

This all assumes that I've got a thermometer that reads tenths of a degree, i.e has a resolution of 0.1. If I take one of my crappy old glass thermometers with me ( I can judge the 1/2 degrees, just) I might end up reading 14.0, 14.5, 14.5, 14.0, 14.5. Which would be less precise, but more accurate .

Is any of this important to a brewer? There's one stage in the process where we're quite keen on getting the temperature "right" and that's in the mash tun. You'll be told that 65.5°C is the magic mash temperature (for a simple infusion mash). Of course, if you want drier beer you may aim lower (say 64°C) - for sweeter beers, perhaps higher (67°C ?). Whatever, we need to decide how accurate and how precise our temperature measurements need to be.

One thing we need to remember is that the mash tun is pretty big and has got a lot of stuff in it. For that matter, the mash tun is made out of a lot of stuff itself (which is why we generally preheat it in some way). We cannot make an accurate measurement of the mash temperature by simply sticking a thermometer in the middle and reading it. At the very least we'd need to make multiple measurements and take some kind of average - this is what we do - when shooting for 65.5, we generally get a range from something like 64 to 67, and the average usually works out OK. So we're reasonably confident that we're in the general area. At least at the top of the mash. Right after mashing in.

That said, what we're really after is nice beer. It's not really important that we mash at 66 rather than 65.5, so long as people like the beer and we get paid for it. But controlling the variability removes some of the potential for surprises down the line. This is why we use a thermometer with a resolution of 0.1°C (although 0.5 would probably be good enough, i.e. a good bit smaller than the range we measure), which we calibrate (to ~0°C) every now and then. We may not be very accurate, but at least we're fairly consistent in what we do. We hope that this helps the beer to be consistent also. And nice.


Ed said...

The trick is to keep poking the thermometer into different parts of the mash until you get the temperature you're after ;-)

StringersBeer said...

Absolutely, Ed. And it works for climate change science too...

HardKnott Dave said...

The trouble with calibrating using the constant temperature of phase change of water is that it isn't, constant that is, it depends upon atmospheric pressure and water impurities.

Well done for spotting my deliberate trick question on twitter - I doubt my thermometer is accurate to 0.5 degrees any more than yours is.

StringersBeer said...

I don't know anyone who has got a thermometer that can tell the difference between tap water /ice mixture and de-ionised water /ice. But if you're interested, kids, look up the cryoscopic constant of water.

Big problem with taking a temperature of the mash is the (lack of) homogeneity - unless I stir it so much that I knock all the air out and have it sink and set. My mash tun is about 110cm wide and about the same deep - i.e. just a little guy - but there's no way that I can get a useful number by sticking a 10cm x 3mm probe in at any particular place.

Anyone can try this: just take 3 or 4 reading around the edge and in the middle of your mash, even at the bottom if you can. See what the range is.

I think what I'm saying is that we try not to get hung-up on any particular datum, but try to establish a consistent (validated - kinda) process.

This is why we calibrate thermometers - it won't make them better thermometers, but you'd hope to get at least a consistent error.