Lesson A8 proved the resource: the practice plant's dryer condensate alone flashes 465 kg/h of perfectly good low-pressure steam, worth around ₹20 lakh a year — currently whistling from a vent. This lesson is the engineering that captures it: the vessel, its sizing, the matching of supply to a consumer, and the honest checklist that decides whether a flash project pays or disappoints.
The system in one sentence
Route hot condensate into a flash vessel held at a chosen lower pressure; the A8 fraction of it re-evaporates and leaves the top as clean LP steam to a consumer, while the remaining liquid leaves the bottom through a float trap to the condensate system (lesson D6). Three design decisions follow: the flash pressure, the consumer, and the vessel size.
Decision one: the flash pressure
The vessel's pressure is a lever with two ends. Flash low and you harvest the most steam (the A8 formula rewards a big hf drop) — but LP steam is fat steam (vg!) and must find a consumer content with low temperature. Flash high and the steam is more useful but scarcer — and the condensate leaves hotter, still carrying flashable heat. The practical answer: flash to just above the pressure your chosen consumer needs. The classic sweet spot is 0.5 kg/cm²g for a feed-tank/deaerator consumer — low enough to harvest well, high enough to push itself through a sparge (lesson B5).
Decision two: the consumer — the make-or-break
Flash steam has no storage patience: it is made the moment condensate arrives and must condense somewhere the same moment. A flash project succeeds or fails on this matching, so audit like this: does the consumer accept low-pressure steam (feed tank sparge, deaerator, hot-water generation, LP coils, air-heater preheat rows)? Is it alive whenever the flash supply is (a feed tank is — it is why it is the default consumer; a day-shift-only duty against a continuous flash source wastes half the harvest)? Is it big enough (a consumer smaller than the supply vents the excess — fit a small relief-to-vent and accept it, or find a second consumer)? The practice plant's 465 kg/h finds a natural home in its feed tank, which lesson B5 already runs at 85–95 °C on exactly this kind of free heat — completing a loop three lessons in the making.
Decision three: the vessel — sized for calm, not volume
A flash vessel is a separator, and its diameter is set by vapour rise velocity: slow enough that droplets fall back rather than ride out with the steam — the working band is around 1 m/s superficial. The practice plant's 465 kg/h at 0.5 kg/cm²g (vg ≈ 1.16 m³/kg) is ~540 m³/h = 0.15 m³/s of vapour: at 1 m/s that needs ~0.15 m² of cross-section — a vessel of roughly DN450–500, about a metre tall with tangential or baffled inlet, level-trapped at the bottom. Note what did not size it: condensate flow, pressure, or enthusiasm. Undersized flash vessels deliver wet steam and re-import lesson A9's problems into the very system meant to be the improvement.
Does it pay? The honest screen
Three lines answer it. Harvest: the A8 fraction × condensate flow × running hours × B2's ₹-per-kg — the practice plant's ~₹20 lakh/yr. Cost: vessel + pipework + trap + controls; flash projects are civil-works-light and typically land inside a one-year payback when a consumer already exists. The spoiler check: if there is no honest consumer — nothing that can drink LP steam whenever it flows — stop, and spend the same money on lesson F5's alternatives instead (recompressing the flash to a useful pressure is the thermocompressor's whole career — lesson F4). A flash vessel venting its harvest is the same waste, one flange later.
- Stand at the condensate receiver vent at full production: a hard, continuous plume is the site survey. Ten minutes, no instruments.
- List LP-steam consumers with their pressure and hours honestly — the feed tank first, always. Matching beats maximising.
- Check trap types upstream: flash recovery raises the back pressure traps discharge against — lesson D1's stall chart may need redrawing for equipment that was marginal.
- Size the vessel by vapour velocity (~1 m/s), never by "the drum we have lying in the yard" — the yard drum is how flash projects earn bad reputations.
- Flash to just above the consumer's pressure; 0.5 kg/cm²g to the feed tank is the classic pairing.
- The consumer decides the project: LP-tolerant, co-scheduled, big enough. No consumer, no project — recompress instead (F4/F5).
- Vessel diameter comes from ~1 m/s vapour rise — it is a separator, not a tank.
- The practice plant's number: 465 kg/h ≈ ₹20 lakh/yr, home to its own feed tank.
Paper mills were flash recovery's proving ground: a paper machine's dryer section is dozens of steam-heated cylinders in cascading pressure groups, each group's condensate flashing into the next lower one, the lowest feeding a condenser or thermocompressor — an entire orchestra of A8 played at machine scale since the mid-twentieth century. Walk any modern Indian paper machine and its "steam and condensate system" drawing is mostly flash vessels; the industry simply refuses to vent what it can cascade. Every smaller plant's single flash vessel is that philosophy, bought one drum at a time.