The voodoo of rust on steel has always baffled me.
With time, I’ve learned that structural steel from the mill has a protective coat of “mill bloom”, that somewhat gray sheen (distinct from the primer added later) which is a variety of iron oxide formed under the anaerobic heat and pressure of the milling process, and which has an inherent, if slight, rust resistance. To this the steel mill immediately adds a spray of primer (hence the spec “milled and primed” when ordering steel, as opposed to “black” steel).
When rust does develop, we react by deep prepping, meaning we strip away both the mill bloom and factory primer, which were the two greatest barriers against rust formation. We can’t replicate the anaerobic conditions that the mill-bloom was created under, and manual forms of prepping and painting require time, which means the bright steel we’ve exposed in prepping immediately begins rusting from contact with salt air.
Therefore, the time between exposing the steel and applying the first coat of rust preventive solution (RPS), or primer, is of importance. Prepping the steel one day and applying RPS or primer the next day guarantees that rust will have begun in the steel again.
If there is a choice, prepping and priming only as much steel as can be completely done in a day, and then painting ASAP, and only moving onto the next “patch" later, is the best way to avoid the problem.
But rarely is this an option. It adds to the total time of the project, and seems to go against the tide of what most sailors think is “Painting”. Everyone wants to see as much paint brushed on each day as possible, not an anal-retentive few square feet. They want to see a lot of finished product, and if it only lasts a few months, well, that’s life at sea.
For RPS (I never call them “primers”) I use Lifeguard (tannic acid base, I believe), having tried several others, though not all of them. I never use Corroseal (chromic acid), which I have found reacts to some International Paint products.
For small areas, I use grinder/ wire wheel, etc. to remove big scale, then apply the RPS. When it is dry, I remove the rest of the rust with flapper brush, on the theory that the dried solution will remain in the microscopic crevices where the flapper can’t get to. Then I immediately cover the bright “flapped” steel with primer, on the theory that primer makes a better bond with steel than RPS.