One Site, Two Sumps: Dosing the Closed Cell Differently From the Active Cell

Two sumps, four hundred metres apart, one polymer store, two doses that differ by forty percent. That is the ordinary state of affairs at a site that capped cell 3 in 2019 and is still tipping into cell 5. The north sump runs black and oily at pH 8.2 with a smell of ammonia across the whole compound; the south sump runs pale brown at pH 6.0 with a scum ring visible from the access road. An operator who carries one number across both will spend the winter re-working batches.

The cost of getting this wrong is not subtle. At 26 kg per cubic metre applied to both sumps, the closed-cell batches came out loose and glossy, and the active-cell batches came out dry enough to crumble under the boot and throw dust at the discharge end. Both failed. One failed the free-liquid check, the other failed the receiving site’s handling check and the haulier’s patience.

What the Two Liquors Actually Contain

Read that table the way a polymer reads it and the puzzle appears at once. The closed-cell liquor is a strong brine — high conductivity means a lot of dissolved ions shielding the charged groups on the resin, which normally pushes absorption down. The active-cell liquor looks weak by comparison. And yet it was the active cell that needed 31 kg per cubic metre against 22 for the old cell.

Why the Weaker-Looking Liquor Took More Polymer

Three things in fresh leachate work against a gel. The first is suspended solids at 1,800 mg/L: fine refuse particles coat the granule surface and slow wetting, so the outside of the particle turns to a skin while the core is still dry. The second is surface-active material — fatty acids and detergents from the tipping face — which drops surface tension and produces the foam you see in the south sump after rain. Foam on the surface is polymer that has gone to the air instead of into the liquid. The third is the scum ring itself, which is a different material with a different dose and should never be counted as part of the batch volume.

So the counter-intuitive result holds: the pale, dilute, low-ammonia liquor is the harder one to gel, and the black, salty, ammonia-heavy liquor is the easier one, provided you accept that its ceiling is lower. At 22 kg per cubic metre the north sump gives a firm, stackable mass in eleven minutes. Push it to 30 and you get a stiff paste that will not blend, plus a mixing paddle under load and a batch that has to be diluted back.

Setting Two Setpoints Without Two Laboratories

You do not need separate bench capacity. Run one cup test routine — 2 g of production polymer into 200 mL of liquor at the temperature it actually sits in the sump, inverted ten times, read at ten and at thirty minutes — and run it on both sumps on the same morning, once a week. The two curves will separate within a fortnight and stay separated.

Then fix two working numbers and write them on the board by each sump, not on a shared sheet in the office. Ours were 21-23 kg per cubic metre for the capped cell and 29-32 for the active cell, adjusted down when the active sump had been still for a week and the solids had settled. Mixing time went the other way: nine to eleven minutes north, twelve to fifteen south, because the solids need time to wet through before the gel sets around them.

Dose the active-cell batches at 60 percent fill, let the gel take up the first volume, then top up. Adding all the liquid first and all the powder second is how you get the fish-eye lumps that never hydrate, no matter how long the paddle runs. The leachate solidification method described for full-scale work follows the same order for the same reason.

The Switch-Over Trap

When a cell closes, its leachate does not change character on the day the cap goes on. It drifts for a year and a half. Ammonia climbs as the methanogenic phase settles in, conductivity climbs with it, and the dose falls in steps of roughly 8 to 12 percent every few months. Sites that re-test only at the annual review end up over-dosing a stream for eleven months, which shows up as a transport bill rather than a compliance problem.

Set a re-test trigger instead of a calendar: any change in which sump is being drawn from, any cap or liner work, or a conductivity shift of more than 15 percent against the four-week rolling average. Three lines in the batch log, checked by the shift supervisor, catches almost all of it.

If both sumps feed one tanker, blend before you dose, not after. Two gelled batches made to different recipes will not merge in the tanker, and the receiving site sees a load with a wet top and a dry bottom and rejects it on appearance alone.