Fanning the coals
Clearing the stem.
18+Reference resource. Nicotine is addictive. Tobacco use harms your health.
Clearing the stem.
Flavor fades before the smoke does: aromatic compounds break down at a lower temperature than glycerin and propylene glycol, which last the session.
The working temperature of shisha tobacco (moassel) under the coal is around 150 °C, far below a cigarette's combustion temperature (about 950 °C). At that temperature, humectants pass into the smoke almost entirely rather than burning. Industry sources describe top heating (coal or a heat management device) as "evaporation, not combustion": the goal is to bring the mix to its evaporation temperature without pushing the leaf into pyrolysis, which produces bitterness (see the article on why shisha turns bitter and acrid). This is a qualitative description, not a measurement, and it comes from industry observation rather than a study.
This is the core mechanism of the phenomenon. Humectants can make up to 70% of the mass of flavored tobacco and form up to 23% of the mass of the smoke's solid particles; reducing sugars make up around 34% of the mass of some of the studied blends, but they behave differently — they don't evaporate whole, they thermally break down into carbonyls, aldehydes, and furans. It comes down to volatility: the boiling point of pure propylene glycol is around 188.6 °C, of pure glycerin around 290 °C, and their mixtures boil in the 188.6–292 °C range depending on the ratio and traces of water and nicotine. A general principle from flavor chemistry (the food and perfume industries, not a measurement specific to shisha tobacco): substances with a boiling point below 100 °C are especially volatile and leave first when a mix is heated in the presence of less volatile "holding" components like propylene glycol.
The minute-by-minute dynamics confirm this. In the first 10 minutes, the smoke is dominated by ultrafine particles of 4–100 nm from sugar derivatives, including levoglucosan; by the 15th minute their number drops by roughly 64% — from 9.92×10¹⁰ to 3.55×10¹⁰ particles per puff. The same study detected flavor markers in the smoke (for example, hexyl acetate) and showed that differences between blends in mass spectra are explained precisely by differences in flavorings — aromatics behave as a separate, short-lived component of the smoke. Glycerin behaves differently: the signal of volatile compounds dominated by glycerin and its breakdown products — acetaldehyde, acrolein, hydroxyacetone — doesn't drop in sync with the aromatic particles, but instead rises with puff count and plateaus.
It's important not to confuse two different measurements here. One study recorded temperature with a sensor inside the bowl head, near the coal's surface: over a 30-minute session it stayed in the 265–318 °C range, dipping about 50 °C with each puff and rising again in the pauses between. Another study placed a thermocouple deep inside the tobacco mass itself: the average tobacco temperature was 118.5 °C (116.5–122.4 °C) while the coal itself was at 429.7 °C. This isn't a contradiction, just two different measurement points — temperature drops sharply the further you move from the coal into the pack. Collapsing both numbers into a single "bowl temperature of 118–318 degrees" would be a mistake: measurement methodology shapes the result as much as the session itself does.
The coal-heated bowl warmed up to maximum slowly — about 15 minutes — while a commercial electric heater reached maximum in 2 minutes. Tobacco mass consumption depends directly on heater temperature: at target temperatures of 350/370/380/400/430/470 °C, the share of tobacco consumed averaged 26.14/31.57/39.10/40.18/41.22/55.74% — the higher the temperature, the faster the whole tobacco resource, including its aromatics, burns through. In that same experiment (171 puffs of 2.6 s with a 17 s interval, a 57-minute session), coal was topped up by roughly 5 g on top of the initial 10 g at puff 105 — a standard practice of adding coal over the course of a long session (more on this in the article on how much coal to use and how to rotate it).
A separate, disputed thread is coal versus electric heat. A presentation at an American Chemical Society conference (2017, University of Cincinnati) showed that both sources initially heat the tobacco to around 300 °C, but the smoldering coal's temperature "sags" with each puff while electric heat holds steady higher — the authors link this to a more active release of volatiles. This is conference data, not a peer-reviewed paper, and it's about toxicity, not flavor. Manufacturers of metal heat management systems claim they deliver a more stable flavor than foil on coal, which cools down and has to be moved (more on how such systems are built in the article on heat management devices and foil), but no direct measurements of "how many minutes the flavor lasts" by heat source have been found.
According to industry sources, a loose pack ("fluff") lets more air through and heats the tobacco faster, speeding up early flavor loss; a dense pack restricts airflow and slows heating, but excessive density causes uneven heating and bitterness. No direct temperature measurements of different pack densities have been found — this is descriptive blog practice, not a thermometer experiment (more on the logic of density in the articles on pack density and on tobacco packing methods).
We won't go deep into the bowl's own shape and material: there are no peer-reviewed measurements on this parameter, only descriptive logic from manufacturers and bloggers (see the article on hookah bowl types for an overview of shapes). A separate matter is the base's water volume — not the same thing as the bowl. A study of hookah sizes (22–55 cm tall, 300–1250 mL base volume) found that as base volume grows, smoke particle concentration drops (from about 10¹⁴ to 10¹³), particle mass drops (from about 10,000 to 2,800 mg/m³), and average particle diameter grows (from 125 to 170 nm). But this study isn't about flavor — it's about toxicity, which turned out comparable across all sizes, so thinner smoke with a larger base doesn't mean the flavor lasts longer.
According to tobacco leaf suppliers, light flue-cured Virginia leaf contains more sugar, absorbs glycerin and molasses well, and produces sweeter smoke; dark air-cured Burley leaf contains less sugar, holds nicotine better, and produces a more pronounced tobacco note. Leaf moisture is usually 14–15% for Virginia-based blends and 16–17% for Virginia + Burley + Oriental blends — the higher the moisture, the slower the heat-up (more on this in the article on leaf cutting and tobacco moisture). This is an industry source, not a peer-reviewed study.
Aromatics are the most volatile part of the mix, so their concentration in the smoke drops fastest (about –64% in particle count over 15 minutes), while glycerin and propylene glycol keep boiling off almost until the end of the session, keeping the smoke from thinning out. Smoke and flavor, in this sense, are two independent processes running at different speeds, not the same indicator.
Why does shisha flavor fade before the smoke does?
Aromatics are more volatile and thermally less stable than glycerin and propylene glycol: they evaporate and break down within the first minutes of heating, while the humectants keep boiling off for nearly the whole session.
How long does the flavor actually last?
There's no measured constant — sources cite different ranges, and the often-quoted "45 minutes" figure traces back, in one source, to a study on toxic load rather than flavor duration.
Is it true that glycerin runs out first?
No: the available data suggest the opposite — propylene glycol is more volatile than glycerin, and the glycerin signal in the smoke rises and plateaus rather than dropping in the first minutes.
Is it worth changing the foil mid-session for the sake of flavor?
There are no measurements of this effect; industry sources more often warn that intervening with the coals mid-session can throw off the temperature regime.
Does cold water in the base help preserve flavor longer?
No confirmation has been found — it's physically unlikely, since aromatics evaporate or break down in the bowl above the coal before the smoke ever reaches the water.
Related topics: pack density — overpack, underpack, and everything in between, tobacco packing methods, how much coal to use and how to rotate it, heat management devices and methods, why shisha turns bitter and acrid, leaf cutting and tobacco moisture, hookah bowl types, and how heat management devices and foil are built and differ.