Long post - if you respond please trim most of it.

This topic comes up from time to time. I've only had one old oolong
that had been re-cooked annually. Rich color, interesting taste, very
little aroma. Good, but not nearly as appealing to me as a high-ferment,
relatively high-roast Anxi oolong. I might have liked it better had each
treatment been a bit lighter. Most descriptions I've heard of this
process indicate that it's about aging, with the re-roast necessary to
prevent staling. I suspect, though, that it would better be seen as
multiple re-roastings with "hold" periods in between. Might be an
analogy with the seasonal rotation alleged to help mature great old Puerh.

I've re-roasted a number of oolongs (and, less often, sheng Puerh and
other teas) when they grew stale or just started boring. Works for
bread, after all. These most notably include some very expensive
"special" dan cong and other oolongs from a famous West Coast dealer,
all of which were disappointingly pallid and distinctly stale on arrival.

I've re-roasted these and others in a few ways, including:

- microwaving, in an open dish with or w/o a cup of water to protect the
magnetron;

- stirring in a small frying pan over a moderately hot electric stove
(wok over flame would be better, but no gas here);

- in the tray of a toaster-oven; and

- in an ash shovel right in the flames of a wood stove.

The first one has had a mild but noticeable and distinctly positive
effect, perhaps surprising given the limited peak temperature. (Once the
water's all gone, energy absorption drops right off.)

Frying pan and toaster-oven were about the same in just a few tries;
stove-top frying gives much better access for turning and frequent
visual/smell inspection.

Limited experiments with right-over-the-fire gave the best results,
possibly because of the slight wood-smoke addition.

Some people use specialized vessels, heat sources, techniques, etc. From
the chemistry perspective, method follows the desired effect. (-Though
the history of science is punctuated by discoveries based on available
equipment and techniques, rather than any special intention.) So I'd
like to float the question: what parameters seems to work best for
restoring old/stale teas, perhaps of different kinds; and for creating
substantially new flavor profiles?

To a chemist, this is an interesting question. Allow me to introduce the
notion of reciprocity: if some temperature for some time produces some
effect, what greater temperature for what lesser time (or vice versa)
will produce the same effect? Or can it even do so? Serious
photographers know about photochemical reciprocity failure in films. And
serious tea-drinkers know that while steeping time and temperature can
be traded off to achieve a given nominal brew strength from a given pot
of leaf, the resulting complex flavor profile will change dramatically.
(Think overnight refrigerator brewing vs. gong-fu.)

Superficially, heat has four kinds of effects: it helps to split stuff,
join stuff, change stuff, and drive stuff off.

Fragmentation usually accelerates very rapidly with temperature due to a
high positive entropy of activation (for those who care and didn't know).

Joining small bits into bigger bits often has a negative entropy of
activation, and does not accelerate as much with temperature. (Chemical
change rarely slows down overall with heat; something else happens
instead.) Relevant example: the Maillard reactions, responsible for so
much of cooking's best effects, happen when amino acids and sugars react
on heating. Or autoxidation, where atmospheric oxygen reacts (most
typically) with olefins, producing species that are already funky and
rapidly decompose into things with very different flavor/odor. This is
probably a key staling cause - and sometimes a chain reaction that can
run away, as when oily rags self-combust.

"Change" covers a multitude of effects, from cis-trans isomerization
(hence "trans fats") and other thermal rearrangements to internal
condensations and many, many more.

What gets driven off, beyond water, is anything of low-to-moderate
molecular weight, including many taste and almost all aroma components.
I'm guessing that the only reason multiply roasted oolongs have any
smell at all beyond char is that some desirable elements are generated
freshly in each cycle.

That was either more or less than you needed to know. My over-arching
point is that the chemical kinetics of something with as many
simultaneous and interlinked processes occurring simultaneously as
tea roasting is not only immensely complex, but almost by definition
non-linear, hence difficult to analyze. Why experience usually beats
theory... At the same time, I'd like to offer a close and a remote
analogy. Fine coffee is roasted at temperatures above the autoignition
point, with split-minute timing and consequences of error ranging from
unpalatability to conflagration. Since the process is so fast, there
would be (at least on the boutique scale) little consequence to dropping
the oven setting a few degrees and adding a few relaxing minutes to the
cycle. But reciprocity fails, and (unlike with roast beef and chili)
long, slow cooking doesn't work very well.

Further in the same direction is the chemical technique of flash
pyrolysis. One can run fragile little molecules through a quartz tube
under inert atmosphere or vacuum at at 2000 degrees F, cracking or
rearranging them into reasonable yields of even more fragile molecules.
It works because contact times are about a millisecond, and the effluent
is collected fresh on a cold finger full of liquid nitrogen. There are
even faster transient-heating techniques.

Slow-roast is easy. Faster methods, like wok-frying, are in common
use in the tea industry and reputed to be done by seasoned experts. I'm
wondering how far the fast-heat approach has been pushed, and what
beneficial results might be found. I plan to test a few more variables,
though perhaps not until autumn. Anyone else care to report experience
in this area?

-DM