Gypsum Retarder Performance with Dihydrate Contamination

Two loads of stucco arrive from the same supplier in the same week. One behaves as it always has; the other stiffens in the mixer before the trowel is out of the bucket, and the retarder that was correct on Monday is suddenly "wrong". Nothing was changed in the formulation. What changed was the powder, and specifically how much of it had already turned back into dihydrate before it ever met water.

Where dihydrate contamination comes from

Dihydrate reaches a mixer by several routes. Stucco stored in humid conditions rehydrates slowly from the surface of the grains inward, so a bag that has sat through a wet month carries more of it than a fresh one. Board and block plants that grind their own offcuts and return the powder introduce set gypsum deliberately. Under-calcined material from the kiln can carry residual dihydrate, and recycled filter dust carries fines that have already hydrated. Any of these puts finished dihydrate crystal into a system that is trying to grow its own.

What seeds do to an induction period

Setting begins with nucleation, and nucleation is the slow step that a retarder works hardest to delay. A dihydrate particle removes that step entirely: new crystal grows on an existing surface instead of forming from scratch, so the slurry skips the wait and moves straight to growth. The effect is not linear. A few per cent of well-dispersed dihydrate can cut the open time sharply, and because the seeds are already present at mixing, no amount of waiting restores the lost induction period.

Reading a stucco certificate for crystal water

The number to look for is crystal water, sometimes called combined water. Hemihydrate holds about six per cent by mass; dihydrate holds close to twenty-one. A measured value above the theoretical figure for pure hemihydrate indicates that part of the load has rehydrated, and the gap gives an estimate of how much. A thermal method, drying followed by ignition or thermogravimetric analysis, gives the figure directly. A simple plant check is a small mixed sample observed for early stiffening, which flags a problem without quantifying it.

What dihydrate tolerance means in practice

A tolerant retarder keeps the open time close to its clean-powder value when seeds are present. Two mechanisms allow that. One is high enough surface activity to cover the extra crystal area and still leave molecules for the growing faces. The other is a growth-blocking action that works on existing surfaces, so it does not depend on preventing nucleation in the first place. A retarder that only delays nucleation will be partly bypassed by a seeded powder, which is why trial data from clean stucco can flatter a product.

Blending, dose or reject

Three responses are available and the choice is a commercial one. Blending a contaminated load with clean material dilutes the seeds and usually restores behaviour at no chemical cost. Raising the dose works up to a point, but it also lengthens the tail of the set and can soften the finished product. Rejecting the load is correct when the crystal water figure is far out and the schedule cannot absorb the risk. What should not happen is a quiet increase in dose that hides the cause until the stock is finished.

Storage and handling that prevent it

Most contamination is preventable and cheap to avoid. Keep stucco on pallets, off a concrete floor that sweats, and under cover rather than under a tarpaulin that traps condensation. Rotate stock so the oldest bag is used first, and treat a load that has been open for weeks as suspect rather than normal. Where recycled powder is part of the process, meter it rather than tipping it in, so its share stays constant and its effect can be measured instead of guessed.

A seeded bench test you can run

Take clean stucco and add ground set gypsum at one, two and five per cent, then mix each at the standard consistence water and measure setting time with and without the retarder. Plot the four points against the unseeded control. A retarder with real tolerance produces nearly parallel lines, holding the delay as seeds rise. One without tolerance converges on the unseeded time, which is the signature of a product that was only ever delaying nucleation. Repeat with your own water, since hardness changes the picture.

Buying tolerance, not just a name

When a supplier claims dihydrate tolerance, ask for the seed level used in the test, the crystal water figure of the base powder, and the setting time achieved at each level. Ask also how the dose was adjusted, because a product that only works at a raised dose is not tolerant, it is merely dosed around. A specification that states performance at two seed levels, plus the batch-to-batch variation of the active content, is enough to hold a supplier to the claim.

Frequently asked questions

How much dihydrate is acceptable in stucco? Many plants work happily under a few per cent and see trouble as the figure climbs. The acceptable ceiling depends on the product and the process, so set it from your own trials.

Will more retarder fix a contaminated load? Partly. It restores open time up to a point, then adds a soft tail and slower strength gain. Blending is usually kinder to the product.

Why does the same formulation behave differently between deliveries? Crystal water and particle size vary between deliveries. Both change the set, and neither is visible on the bag.

Can I test for dihydrate without a laboratory? A sealed sample mixed at standard water and checked for early stiffening gives a yes-or-no answer. Quantifying it needs a thermal method.

Does dihydrate reduce final strength? Often, because seeded growth produces a different crystal structure and more porosity. Strength testing on the finished product is the honest check.

For retarder grades with seeded-condition data, or to arrange a trial against your own stucco, write to [email protected] or call +86-537-3739818. Product information is listed at yuanlianchem.com, and crystal-growth background at yuanlianchelating.com.