Can zinc-rich primer go over a St3 surface?
In repair yards and site touch-up work the sentence comes up almost daily: the blasting machine cannot reach this area, or the owner will not allow abrasive blasting, so can we go St3 and put zinc-rich straight on?
Technically it can be done. Treating it as one grade lower with roughly the same performance, however, is how you plant a problem for yourself. The gap between St3 and Sa 2.5 is not about looking clean; it sits in two things you cannot see: profile and consistency of cleanliness.
1. What St3 actually delivers
ISO 8501-1 defines St3 as thorough hand and power tool cleaning: the surface should be free of visible oil, grease, mill scale, rust, previous coatings and foreign matter, and show a uniform metallic sheen. That reads strict, but it differs from blasting in one fundamental way: St3 says nothing about roughness.
Sa 2.5 requires a defined profile, usually expressed as an Rz or Ra range, in addition to cleanliness. St3 describes clean, not rough. Those two properties do different jobs: cleanliness governs chemical and electrochemical contact, roughness governs mechanical keying.
Worse, St3 cleanliness is judged by eye. A wire wheel brightens the surface until it looks compliant while rust and salts remain in the pits. Blasting removes the surface layer by impact; a wire brush flattens and polishes the rust.
2. Why zinc-rich is so particular about the substrate
Zinc-rich primers protect by cathodic action, and that requires an electrical path: zinc particles touching each other and touching the steel. A layer of rust or mill scale in between breaks the path. The zinc is still there; it simply cannot reach the steel.
Adhesion is the second problem. Without a profile the coating holds only by intermolecular forces. Nothing shows at first, but under immersion or thermal cycling the interface opens first. That is why zinc-rich fails so quickly in immersed service.
So the substrate requirement for zinc-rich is conductivity plus grip, not visible cleanliness. St3 satisfies both only partly.
3. What the failure looks like
It does not come off in sheets. The typical pattern is spot blistering, or isolated rust spots showing through the film. Open a blister and the base is black rust, with the zinc layer lifting off intact above it. This is the hardest kind of failure, because at acceptance the appearance is normal.
On the time axis, atmospheric areas may take three to five years to show, while immersed areas show within a year. So "we did it last time and nothing happened" is not evidence of feasibility. Ask which area, and for how long.
4. When it is acceptable and when it is not
Four preconditions, and losing one means reconsidering: atmospheric rather than immersed exposure, corrosivity C3 or below, a relatively small area forming local repair rather than the main protection, and an intermediate plus topcoat completing the system above.
Conversely, do not do it in ballast or potable water tanks, in any immersed area, on main structure in C4 or C5 industrial and marine atmosphere, or wherever the specification calls for Sa 2.5. If the design says Sa 2.5 and the site substitutes St3, that is an unauthorised change and the contractor carries it.
If it is done anyway, record it: which area, which tool, what time, who inspected, what thickness. That is not paperwork for its own sake; it is the only thing available when the argument arrives five years later.
5. Two routes better than forcing zinc-rich
First, change the product rather than lowering the standard. Where blasting cannot reach, use a surface-tolerant epoxy primer that accepts tight rust and damp, instead of insisting on zinc-rich. These products film by penetration and wetting and do not need a cathodic path, so they tolerate St3 or even St2. The trade-off is no cathodic protection, which limits them to moderate corrosivity.
Second, zone the work and overlap. Areas that can be blasted go to Sa 2.5 as specified, areas that cannot go to St3, and the boundary becomes an overlap band: coat the blasted side first, then carry it across the joint. Overall performance is then set by the blasted area and the weak point is confined. This needs a zoning drawing agreed in advance, not an improvised decision.
The test was never whether the paint goes on. It is what you find when you open it five years later. Put that on the table and most of the argument disappears.