All I see is more myth making, with no experimental evidence on either side. Just to play devil's advocate:
* practically every recipe for brewing tea or coffee emphasizes the importance of starting with fresh cold water, not water that has been sitting in a boiler
* in addition to picking up minerals and metals, water that has been boiled in an open container loses oxygen and tastes "flat"
* water is the major constituent of espresso, so even small changes in composition may have an impact on taste

Based on the above, it appears that the burden of proof lies equally on both sides. IMHO, starting with fresh cold brew water is a benefit of the HX design, although I'd be the first to admit there is no hard evidence to support this belief.

I'm currently using a DB machine (Spaz S1) with a large (2.5L) steam boiler and a much smaller (0.45L) brew boiler. One advantage of this design is that the brew boiler water gets replaced rapidly, even under light use. Other manufacturers have chosen much larger brew boilers, sometimes putting steam and brew boilers of identical volume in their DB models. The logic behind this (identically sized boilers) escapes me; perhaps it has to do with parts availability rather than good design.

You can say that the hx machines start with cold, fresh water only if you can persuade me that the group flushing regimen in use completely recharges the hx with the cold water you mention. I don't think that's the usual case.

-Greg

gscace wrote:You can say that the hx machines start with cold, fresh water only if you can persuade me that the group flushing regimen in use completely recharges the hx with the cold water you mention.

The HXs on semi-commercial / prosumer units are tiny (~120 mls). HX "dragons" like the Elektra Semiautomatica, Olympia Maximatic, Gaggia Achille, Expobars, etc. flush the HX completely and then some. The commercial HXs of the Cimbali line and the Nuova Simonelli Aurelia, for example, have much greater volume, require very little flushing, and are therefore an entirely different story.

Good point. Again, no evidence to present. However, home machine HX volume is typically small (about 100ml), so there's a far better chance of a (near) complete flush than with virtually any DB machine.

Oops, sorry for the cross posting.

RapidCoffee wrote:I'm currently using a DB machine (Spaz S1) with a large (2.5L) steam boiler and a much smaller (0.45L) brew boiler. One advantage of this design is that the brew boiler water gets replaced rapidly, even under light use. Other manufacturers have chosen much larger brew boilers, sometimes putting steam and brew boilers of identical volume in their DB models. The logic behind this (identically sized boilers) escapes me; perhaps it has to do with parts availability rather than good design.

Most of the arguments in favour of "fresh" water seem to take it as a given that obviously fresh water is best. The ones that engage with the subject matter a little bit more go on to point out that heating water for a long period of time will drive off dissolved oxygen. Frankly, this strikes me as a case of a little knowledge being a dangerous thing. I can remember reading the dissolved oxygen tidbit in a few places, but I can't remember any sources that actually did taste tests. Perhaps I'm missing some actual research, but it wouldn't surprise me if this is a case of people concluding that obviously it's better to have more oxygen dissolved in the water if you are brewing coffee. Which is quite funny because when coffee is in air instead of water, everyone seems to agree that oxygen is a bad thing. At any rate, if you pick up a copy of Sivetz' Coffee Technology book, it actually states that water should be boiled to drive off all dissolved oxygen before brewing coffee. I thought that it was worthwhile having an informed opinion on this, rather than conjecture, so I did some triangle cuppings using the same coffee, but comparing water brought just to the boil against water boiled vigorously for 25 minutes. FWIW, I couldn't pick the difference. I wouldn't claim that to be a thoroughly researched and statistically valid study that everyone should look upon as gospel and I'll repeat it again when I have time, but, in the meantime, I won't be losing too much sleep over the whole "fresh" water thing.

As for the logic behind having a large brew boiler in a multi boiler machine, I think that one of the reasons is that if you have a large mass of water relative to a small mass of metal, the whole thing is more likely to reach an equilibrium faster, seeing as the water can assist in moving the heat around. If you look at a LM or Synesso saturated group, there is so much brew temperature water there, immediately next to where the portafilter locks in that you have to think that the temperature change as you start to pull shots will be fairly minimal. A while ago, I popped into the cafe where I used to work to check out a new machine that they had been loaned whilst the Synesso was being rebuilt. It was a multi boiler machine with a massive steam boiler (great!) and three small brew boilers, which were bolted on to groups that were not saturated like the LM groups; I like to think of them like "chunk-o-metal" groups. The shots started to taste more astringent as the day went on and this problem was alleviated somewhat by dropping the boiler temperatures to compensate. Our guess was that this was the effect of the group heads heating up as they were used more and more - I think that Abe observed the same phenomenon in his brewtus review. From memory, the S1 has a similar setup. At any rate, I think that everyone was relieved to get the Synesso back.

Another part of the equation that I don't know much about is how the water in the boiler actually behaves. Water doesn't heat uniformly throughout the body; the water nearest the element will heat up more than the water furthest away and then you get currents with the heat moving around. This adds a whole level of complexity to dual boiler and group design, when you consider that just as you can only stabilise the temperature of the water at the point of the probe, you only draw water at one point. Interesting stuff.

Cheers,
Luca

Well, it seems I spoke too soon. This tale of Water on its way to Union with The Bean, is not done. And the myths now look both ways.

It's helpful for me to hear about all the ways the water is changed out, emptied, etc. between the DB and HX, and to consider Luca's outside-the-box question about whether fresh cold water really is the Holy Grail.

I'm wondering if it's worth asking this water question of one or more manufacturers that make both DB and HX machines; or equipment retailers who sell both? If so, would likely manufacturers be Izzo, La Spaziale, or Vibiemme? For retailers, perhaps Chris of Chris' Coffee?

An interesting discussion on many levels.
Ann

P.S. RapidCoffee, I had to laugh and wonder about your "cross posting". I thought of cross-dressing, posting across forum boundaries, and at last it occurred to me you meant cross disposition---am I right?
Ann

lattelover wrote:P.S. RapidCoffee, I had to laugh and wonder about your "cross posting". I thought of cross-dressing, posting across forum boundaries, and at last it occurred to me you meant cross disposition---am I right?

Nothing nearly so kinky. By cross posting (perhaps a misuse of the terminology), I simply meant that Dan and I composed our posts at the same time, and he got there first.

luca wrote:...I did some triangle cuppings using the same coffee, but comparing water brought just to the boil against water boiled vigorously for 25 minutes. FWIW, I couldn't pick the difference. I wouldn't claim that to be a thoroughly researched and statistically valid study that everyone should look upon as gospel and I'll repeat it again when I have time...

Best argument I've heard yet on this thread. Thanks for taking the time to actually try this. Experimentation beats angels dancing on the head of a pin any day.

luca wrote:As for the logic behind having a large brew boiler in a multi boiler machine, I think that one of the reasons is that if you have a large mass of water relative to a small mass of metal, the whole thing is more likely to reach an equilibrium faster, seeing as the water can assist in moving the heat around.

The downside: changing the brew temp (especially reducing it) takes longer with large brew boiler volumes. Dunno if I agree about equilibrium; I'd guess thermal gradients are a bigger problem with larger boilers (but that's more dancing angels).

luca wrote:...The shots started to taste more astringent as the day went on and this problem was alleviated somewhat by dropping the boiler temperatures to compensate. Our guess was that this was the effect of the group heads heating up as they were used more and more - I think that Abe observed the same phenomenon in his brewtus review. From memory, the S1 has a similar setup.

Rising brew temps have been noted on several home DB machines (especially those with E61 groups). I haven't had noticeable problems with the S1, but my usage is far lower than a commercial cafe.