It started on a Monday morning when the first production batch of a new waterborne industrial enamel came off the mill looking more like shaving cream than paint. The line had been running the same acrylic system for months without issues, but after switching to a new dispersant grade, foam height in our standard graduated cylinder test shot up from 90 mm to 190 mm after ten minutes of high-shear mixing. Draw-downs were riddled with pinholes, gloss at 60° dropped to 59 units, and the spray booth was already rejecting panels. The mineral oil defoamer we had been using at 0.4 % barely touched it.

I pulled three small 5-litre test batches from the let-down tank and ran a direct comparison, adding each candidate at 0.3 % active in two stages — half during the final grind and half in the letdown. All three were silicone-based, but with different modifications. Here is what the numbers showed after dispersion and again after 24 hours at room temperature:

• Unmodified PDMS silicone emulsion: foam height fell to 28 mm right after mixing but climbed back to 45 mm after 24 hours. Pinholes on draw-downs dropped to 3 per 10 cm² and gloss recovered to 76 units. However, we saw mild pigment flooding on the red shade and a noticeable increase in surface slip that later caused light scuffing during handling.

• Standard polyether-modified silicone: foam height stayed at 19 mm immediately and only rose to 22 mm after 24 hours. Pinholes were essentially gone (less than 1 per 10 cm²), gloss reached 83 units, and storage stability looked clean after seven days. The only downside was a slight haze on the highest-gloss panels when we pushed the dosage to 0.45 %.

• A newer branched polyether-modified silicone designed for high-shear systems: foam height dropped to 14 mm and stayed at 16 mm after 24 hours. Pinholes were zero on both draw-downs and sprayed panels. Gloss hit 86 units with no haze even at 0.4 %. After 30 days at 50 °C there was no viscosity drift and no separation. Recoat adhesion on lightly sanded panels remained excellent.

The branched modified version became our standard for that line. We dropped the dosage to 0.28 % total (split addition) and the reject rate in the spray booth fell from 18 % to under 2 % within two shifts. Filling speed also improved slightly because the paint released air faster. The higher raw-material cost of the new defoamer was offset by lower rework and fewer rejected batches.

What surprised me most was how sensitive the system had become to the dispersant change. The new dispersant was excellent at lowering viscosity and improving colour development, but it created far more stable foam films. The unmodified silicone could knock the foam down initially but lacked the persistence needed once the batch sat or went through pumps. Only the properly modified versions kept working through the whole process.

I have seen the same pattern in other plants. In a high-gloss waterborne wood coating, a conventional silicone gave perfect foam control in the can but caused micro-craters when the customer applied a second coat after light sanding. Switching to a more compatible polyether-modified grade solved the cratering while keeping foam height below 20 mm. In a solventborne alkyd maintenance coating, we actually moved away from silicone altogether because even small amounts caused crawling on poorly prepared steel; a polymer-based product performed better there.

The practical rules I now follow with silicone defoamer are simple but easy to overlook on a busy floor. First, always run a short ladder (0.1 %, 0.2 %, 0.3 %, 0.4 %) on the actual formulation rather than relying on supplier recommendations. Second, split the addition whenever possible — putting everything in the grind can shear the defoamer too hard and reduce its later effectiveness. Third, check compatibility on the real substrate and with the full additive package; a product that looks perfect in the lab can still create fish-eyes once the wetting agent or co-solvent level changes.

Silicone defoamer remain one of the most powerful tools we have for fast, low-dosage foam control in waterborne systems, but they are not automatic. The chemistry has to match the specific foam-stabilising package in the formula, and the addition method has to respect how the batch is actually made and applied. When those pieces line up, as they did on that Monday morning line, the difference is not just fewer pinholes — it is a process that suddenly runs predictably instead of fighting you every shift.