DO sampling techniques

[dick murton]

Agreed. Thanks

In spite of having talked to a number of suppliers, this is more detaile knowledge than I have been given. I suspect some inside knowledge here !!

[TL Services]

Dick - the information is all out there

Incidentally going back to response times, the figure for an optical sensor is circa 30sec. in liquid (beer), which compares to a worst case for EC of 38 sec (membrane for high level analysis, up to 80ppm) to a best of just 7.2sec (using a membrane optimised for low-level analysis).

Both those membranes are capable of measuring across the full range of DO levels normally found in brewing (the low-level membrane has a stock range of zero to 20ppm) compared to an optical sensor that goes over range at 2ppm.

[dick murton]

Well, all I can repeat is that when I had 4 different suppliers optical sensors in the same piece of pipe as an EC sensor - the 4 opticals picked up spikes over two or three samples before dropping to previous levels. The EC sensor barely moved. Hence my general approval of optical sensors.

[TL Services]

Dick - I hope they weren't all there together or the spikes might simply have been the four opticals interacting as they each set off fluorescence in the others...

But seriously, those response times are all taken from manufacturer literature so I can't be accused of bias. I would suggest a change of membrane to improve response time so you would be comparing like with like.

I'm not knocking optical sensors; they have benefits in certain applications. I just don't feel they are the ideal choice for a craft brewer who doesn't have the (financial) option of multiple instruments that are tailored for different stages of production.

[dick murton]

"I hope they weren't all there together or the spikes might simply have been the four opticals interacting as they each set off fluorescence in the others"

- yes they were fairly close together, about a foot apart, but if this was due to interference from one sensor to another, I would not have expected to see a spike, but a continuous high trend which was not reflected by the EC sensor. The EC sensor did pick something up, but it was somewhat after the optical units had all peaked, and didn't reach the same peak due to the "instantaneous" drawing in of air on pump start up. Once the pump mechanical seal and associated spring (on the drive shaft) had been replaced, the spikes went away on all the units.

But I do agree that a single EC sensor is more flexible in use, and therefore better suited to small brewers. Only where you want this sort of response, say just before a packaging line as last ditch monitor on infeed to a filler would you generally want this sort of response.

As it happened, the guys told me there was a problem with high DOs whenever there were regular stops and starts, so in this case it was a really easy match of anecdotal and high speed response evidence.

Anton Paar (and I think Mettler Toledo) were looking to make a single range unit, but I am not sure how far they have got with that one, particularly for a hand held unit. Everyone else was still using two ranges, and I got the impression they intended to stay that way. Perhaps someone who got to Brau (or similar) last year can update me.

[TL Services]

Only where you want this sort of response, say just before a packaging line as last ditch monitor on infeed to a filler would you generally want this sort of response.

That takes me back...I just about went blue in the face trying to convince one brewer that inline DO2 in precisely that location would be a great addition but PARTICULARLY if they took an output from the instrument and used it to control filling.

With some simple SPC the likelihood of high TPO2 would be virtually eliminated.

To the best of my knowledge there isn't yet an optical system that can cover the full range in brewing; I only assume that the technology is such that the trade-off between range and accuracy is too great right now.

[dick murton]

I supect a lot of the big boys will be doing this to reduce manning and improve data collection for fine tuning, customer complaints / excise protection, especially if used with CO2, ABV etc - perhaps even colour / pH.

At 50k bottles / cans an hour it comes pretty cheap per package