A pyramid sea salt crystal is hollow. Look at one through a hand lens and you see a small, thin-walled pyramid that has built itself, layer by layer, on the surface tension of still brine. The crystal is not solid the way a granule of table salt is solid; it is a four-walled shell, sometimes nested inside another. This is the structural fact that makes the salt work the way it does on the plate.

I think about the crystal a lot, more than is reasonable for someone who makes it for a living. The shape is not invented; the geometry is what sodium chloride does when it crystallises out of a slowly-concentrating brine on still water. Pyramid salt is not a brand decision. It is a chemistry.

What the shape does.

The hollow pyramid snaps cleanly between the fingers, which is the first thing a chef notices. Granular salt rolls. Pyramid salt sits where it is put, then collapses on warmth. The flake makes contact with food the way granular salt cannot; it stays on the surface long enough to register as texture, then dissolves into the fat or moisture of the dish.

It is also more efficient. A pinch of pyramid flake covers a dish with less salt by weight than a pinch of granular salt would, because the flake spreads. This is not a small effect. A restaurant we supply in London tracked their salt usage for six months after switching and found they were using about a third less by weight, with no perceptible change in seasoning.

The pyramid is not a marketing flourish. It is the shape salt wants to take when nobody hurries it.

From a 2018 monograph on solar evaporation, Aix-en-Provence

What the shape costs.

Pyramid crystals form slowly. Speed up the evaporation, by heat or by mechanical agitation, and you get granules instead. The pyramid wants still water and a long, undisturbed concentration. This is why the salt is expensive: because the production rate is slow and dependent on weather.

A good summer day evaporates about two centimetres from a shallow pan; a poor one evaporates none. Twenty good days, in sequence, gives a harvestable pan. Fewer good days means a thinner harvest. The economics of pyramid salt are weather-dependent in a way that mass-produced salt is not, and there is no fix for that without breaking the shape.

The numbers.

600 kg per year, in a good summer.

Twelve pans, three weeks per harvest cycle, two or three cycles per summer. The retail and trade orders are filled in that order: trade first, retail next, the gift box last. Anything left over at year-end goes into the next year’s tasting kits.

• Spring: pans cleaned, repairs done, water tested.
• Early summer: first harvest cycle begins on a high spring tide.
• Mid summer: peak harvest, three pans working at once.
• Late summer: final cycle, pans drained for winter.
• Winter: packing and shipping, recipe development, accounts.

What chefs do with it.

I keep a list of the dishes Bannow Bay salt has finished on the plates of restaurants I have visited. The list is selective and not exhaustive; many of the kitchens we supply, I have not eaten in. Among the entries I keep returning to:

1. A piece of cured trout at Variety Jones, with three flakes laid on the surface.
2. A salted-butter ice cream at Liath, dusted at the table.
3. A slice of dark chocolate cake at L’Enclume, with a pinch of flake at the side of the plate.

Each of these uses the pyramid as an accent rather than a seasoning. The salt is the last thing the dish meets, and the texture is part of what the diner registers.

Reading.

For anyone interested in the technical detail, two references that helped us when we were building the pans: Salt by Mark Kurlansky for the long history; Salt of the Earth: A Practical Guide for the geometry of pan-design. Neither is published by us; we have no commercial interest in either.

If you want the chemistry of the pyramid specifically, the 2018 monograph quoted above (now out of print) goes into more detail than is necessary for most readers. We have a photocopy at the studio and will lend it on request.

Sample code, in case it is useful.

For the technically curious, this is roughly the shape of the harvest schedule we run, expressed as the small Python loop that drives the pan-monitoring sensor on each tank. The sensor reads brine concentration twice daily and writes a CSV file that I open every Sunday morning to see which pans are ready. Inline values like BRINE_TARGET = 0.26 are the practical thresholds we have learned by trial.

BRINE_TARGET = 0.26  # 26% by weight, the pyramid window
INTERVAL_HOURS = 12

for pan in pans:
    reading = pan.read_brine()
    if reading >= BRINE_TARGET:
        notify("Pan {} ready to rake".format(pan.id))
    log_to_csv(pan, reading)

The pop element below pulls a single line into a designed moment. We use it sparingly in the journal, mostly for things worth slowing down to read.

If pyramid salt has a season, it is the last week in August.

Newsletter.

If you would like the journal in your inbox four or five times a year, leave an email below. We do not run promotional offers; the newsletter is the journal, posted as it is written.

Subscribe to the journal

Your name

Email

A note (optional)

Subscribe Four or five letters per year. No promotional content.

Filed under: salt, harvest, journal · 14 August 2025