Desalination of Porous Stone: Methods and Limitations Explained

You scrape a white, powdery bloom off a limestone sill with a scalpel, and three months later it's back — thicker. That's the moment most people realise they're not dealing with dirt. They're dealing with salt crystals growing just under the surface, and the stone is losing material every time the cycle repeats.

Salts don't sit on porous stone. They move through it. Water carries them in, water leaves, the salt stays and recrystallises. Each cycle pushes crystals harder against the pore walls until the surface gives way. Desalination is the only treatment that removes the salt itself rather than hiding its symptoms, and it is also the treatment most often botched, for reasons that have nothing to do with skill and everything to do with chemistry.

Below: which methods actually pull salt out of masonry, what each one costs you in risk, and where the physics simply refuses to cooperate.

How salts get inside porous stone, and why porosity makes it worse

Open porosity is the whole problem. A dense marble resists salt because there's nowhere for the solution to travel. A sandstone with 15–20% open porosity behaves like a sponge with a plumbing system: water enters through the surface, migrates by capillary action toward the drying face, evaporates, and dumps its dissolved load right where evaporation happens.

Efflorescence versus subflorescence

Salt that crystallises on the surface is efflorescence — ugly, but relatively benign. Salt that crystallises a millimetre or two below the surface is subflorescence, and that's the destructive one. The crystal has nowhere to expand, so it exerts pressure on the surrounding pore walls. Stone in that condition doesn't stain. It powders, flakes and eventually disintegrates from the inside.

Two conditions accelerate the whole thing:

• Rising damp pulling groundwater loaded with nitrates and chlorides up from foundations.
• Modern cement repointing on historic masonry, which traps moisture behind a vapour-tight joint and forces evaporation — and therefore crystallisation — into the stone itself.

Which brings up the question most clients never ask: if the salt is coming from the ground, desalinating the stone without addressing the moisture source is a subscription, not a treatment.

Methods for desalinating porous stone: what actually works

The core principle behind every accepted method is the same. You create a moist, porous medium in contact with the salt-laden stone, let capillary suction draw the saline solution out, then remove the medium before the salt can crystallise back into the stone. The medium has to be more absorbent than the stone, and it has to be replaced before it dries out.

Poultices and compresses

Kaolin, attapulgite, cellulose pulp, sepiolite, and Japanese paper laid over a fine clay layer all belong here. The technique is unglamorous and effective:

1. Pre-wet the stone so the salt is in solution, not in crystal form.
2. Apply the poultice at a thickness of 1–3 cm, pressed into contact.
3. Keep it damp — plastic sheeting or a misting regime — for anywhere from 24 hours to several weeks depending on wall thickness.
4. Remove it while still damp and dispose of it as salt-contaminated waste. Letting it dry in place reverses everything you just did.
5. Repeat. And repeat again. Chloride content in a thick wall typically drops by roughly half with each well-executed cycle.

The limitations bite here. On a 60 cm rubble wall, a poultice only reaches the outer few centimetres. On painted or rendered surfaces the medium can't contact the stone at all, so either the coating comes off or the treatment doesn't happen. And on friable, already-weakened stone, wetting and poulticing can pull away surface material along with the salt.

Baths and immersion

For movable elements — carved capitals, fragments in storage, sculpture — full immersion in demineralised water with regular water changes is the most thorough option available. Conductivity monitoring tells you when the water has stopped loading up with ions. It is also completely impractical for anything still attached to a building, and repeated soaking stresses certain clay-bearing sandstones.

Electrokinetic extraction

Here's where it gets interesting. Apply a low-voltage direct current across the damp masonry with electrodes embedded in poultices on the surface, and the salt ions migrate: chlorides and other anions toward the anode, cations toward the cathode. You're not relying on passive capillary suction any more. You're driving the ions.

Results can be faster and reach deeper than poulticing alone. The catch? The wall must stay saturated for the current to conduct, you need continuous monitoring so you don't dry out one zone, and the technique alters the moisture distribution in ways that can mobilise salts toward untreated areas if the electrode layout is careless. I've seen a badly planned installation push a chloride front 30 cm sideways into a clean plaster panel. That panel had to be redone.

Limitations you need to accept before you start

What are the limitations of desalination on stone? The honest answer is that complete removal is not achievable, and the pursuit of it causes damage. What you can achieve is a reduction below the threshold where crystallisation pressure exceeds the stone's tensile strength.

Method	Reach into the wall	Main risk	Realistic outcome
Clay/cellulose poultice	2–8 cm per cycle	Surface loss on friable stone	Roughly halving salt load per cycle
Immersion bath	Full depth (movable objects only)	Stress on clay-bearing stone	Near-complete removal
Electrokinetic extraction	Up to 20–30 cm	Salt migration into adjacent areas	Fast reduction, needs monitoring
Surface washing alone	Millimetres	Drives salt deeper	Cosmetic only — not desalination

Where does the salt go after desalination?

Into the poultice, mostly — that's the point. The salt-loaded clay, pulp or paper is removed from the wall and disposed of as contaminated waste, which means controlled disposal, not the skip behind the site. Any rinse water used during the process has to be collected and treated the same way. Salt doesn't disappear; it gets relocated from a place where it destroys heritage to a place where it does no harm. Occasionally it relocates somewhere it does harm — into adjacent render, neighbouring stone, or the base of the wall if drainage is ignored.

What desalination removes from water — and from stone

What does desalination remove from water? Dissolved salts: sodium chloride chiefly, plus sulphates, nitrates, calcium and magnesium compounds, and — in the case of seawater — a long tail of trace elements. Reverse osmosis handles this at industrial scale by forcing water through a membrane that rejects the ions.

Poultice desalination on masonry uses the same underlying logic at a much smaller scale: it removes the ionic load, in whatever mixture the wall happens to contain. It does not remove:

• Deep-seated salt beyond the reach of the medium
• Salts bound inside dense or low-porosity inclusions such as flint nodules
• The source of the salt — which is the moisture regime, not the salt itself
• Existing physical damage. Desalination stops deterioration; it does not reconstitute lost stone.

Why desalination is so expensive

Why is desalination so expensive? For water, the answers are energy and membranes. For stone, the answer is labour and time — and the fact that the process cannot be compressed.

A single poultice cycle on a modest façade might take a fortnight of dwell time plus application and removal. A wall needing four cycles is a two-month campaign for one elevation, with scaffolding, containment, waste disposal, moisture monitoring and a conservator on site throughout. Multiply by the number of elevations. Now add the diagnostic work that has to precede it: ion chromatography or simple nitrate/chloride test strips on drilled samples at several depths, so you know what you're removing and how deep it goes.

The uncomfortable part is that without fixing the moisture source, the salt comes back. Repointing in a vapour-permeable lime mortar, correcting drainage, addressing a leaking downpipe — these are cheap by comparison and frequently skipped in favour of a visible, photogenic treatment. I'd argue that's backwards, and I'll defend that position: spend your budget on water management first, desalination second.

The practical verdict

Poulticing remains the default because it is gentle, reversible in its effects and requires no specialised equipment — just discipline about timing and repetition. Electrokinetic methods deserve more use than they get on thick walls, provided somebody is watching the moisture readings. Immersion is unbeatable where objects can be moved.

None of these will give you a salt-free wall. What they give you is a wall where crystallisation pressure has dropped below the failure threshold, which buys decades rather than months. That's the realistic target, and it's worth stating plainly to anyone commissioning the work.

And when the bloom reappears after the third cycle, two years later, thinner each time — that's not failure. That's the process working as designed, in a wall that will keep producing salt for as long as water keeps finding a way in.