I'm not interested in any decaf except water process - would be glad to hear suggestions.

Any list of recommended Swiss Water Process decaf blends for espresso should include Water Avenue Coffee's Decaf El Toro. WAC (Portland, OR) is, in my opinion, fully deserving of the great reviews they're getting; the regular El Toro is getting most of the press, but the decaf version, while it's focus is a little different, absolutely stands on its own. It gives a terrific toffee layer (as advertised), but it's also balanced fairly brightly, so you'll likely find your favorite brewing temperature at 200F and above.

Coffee snobs usually turn up their noses at any large seller, but Peet's makes an SWP Decaf Mocca-Java that's extremely consistent and very, very good, hitting all the right notes you'd expect from this blend, and in plenty of volume, and comparing well to the regular Arabian Mocca-Java that Peet's also offers.

When differentiating espresso technique using standard versus decaf beans, it's instructive to compare matched pairs from the same purveyor. My experience where the same or similar blend is offered in both regular and decaf standard versus decaf (Metropolis Red Line, Water Avenue El Toro, Peet's Major Dickason and the Peet's Mocca-Java mentioned above) has revealed some standard patterns:

First, the decaffeinating process does reduce the amount of protective carbon dioxide in the bean, so it's logical that the roasted coffee loses its quality faster. Where most baristas advise that regular beans are at peak (for espresso) between, say, four and ten days from roasting, that range typically shrinks for decaf, starting instead in just one or two days and lasting up to a week (or even less) for decaf beans.

Second, and for the same reason, a decaf version almost always provides less - and less persistent - crema than will the regular version of the same product. All other things remaining equal, the crema-producing carbon dioxide is simply in shorter supply.

Third, if you use the same grind and dose, a decaf puck will act more porous and will blonde faster than its regular counterpart. The standard reaction would be to tighten the grind to slow down the pour, but remember that the decaf essential oils and flavor elements have already been removed from bean's porous structure and then put back, so ground decaf seems to release them that much faster the second time around. This is why up-dosing with the same or coarser grind may prove to be a more successful compensation strategy, slowing the extraction and the pour back to normal.

Finally, I find that decaf blends, especially from smaller roasting companies, are seldom as consistent, batch-to-batch, as the caffeinated versions of the same blend (and by consistent I mean "different" as opposed to varying quality or "goodness"). There is just not as much green coffee available to roasters in decaffeinated form, and availability of a particular bean/element can be iffy at best. Making matters worse, some maintain that, even in its green state, coffee loses quality and changes character faster after the decaffeination process. So with fewer building blocks, further complicated by relatively short-term warehousing capability for the green coffee, maintaining the profile of a decaf blend poses an especially severe challenge. Perhaps this explains the growing number of "seasonal" blends where variance is assumed (and even advertised). A bigger operation (e.g. Peet's) keeps larger stocks with more variety and has greater access to green coffee - likely explaining the unusual consistency of their decaf blends.

Anvan wrote:Any list of recommended Swiss Water Process decaf blends for espresso should include Water Avenue Coffee's Decaf El Toro.

Full disclosure - our Decaf Toro is Water processed, but not SWP. Thanks Anvan for the kind words about our decaf. One of those coffees that need to be treated with a little finesse. -b

FWIW, the water process does more damage to the coffee than a simple process based on a strong volatile solvent like dichloromethane. The Swiss process rips all the flavor constituents out of the bean, removes the caffeine (among other flavor constituents), then stuffs what's left back into the bean like a kishke.

I've seen the Fischer-Kummer patent (and the last time I was in a serious discussion of this - many years ago on a.c - Coffex still had a detailed, illustrated technical description on their corporate website).

Here's Ken David's description of the process:

In layman's terms, the "Swiss Water Process" (copyrighted name of course) involves soaking the beans in hot water until they've given up everything but the woody structure, running the waters through charcoal filters to remove the caffeine, then recombining the beans and the water. The beans soak up the good stuff from the water again, leaving the caffeine behind in the filters.

It is true that by starting with constituent-saturated water the newer methods are said to remove fewer of the non-caffeine constituents from the bean than the original Swiss process, but the process is still more disruptive to the bean than the traditional methylene chloride "European" decaffeination process.

Incidentally, Ian Bersten has a nice chapter on the history of decaffeination (and anti-caffeine hysteria) in Coffee, Sex, & Health.