What We Were Dealing With
The damaged section was roughly two feet square on an interior wall. Someone had removed what looked like a built-in shelf bracket at some point and repaired the resulting holes with 1/2-inch drywall scrap, without accounting for the fact that the original plaster was closer to 7/8 of an inch thick. The drywall sat about 3/8 of an inch below the surrounding plaster surface and had been feathered over with joint compound that had since cracked in a rough grid pattern — probably from the humidity cycling causing the two materials to move at different rates.
The compound was also mostly delaminated at the edges where it met the plaster. When I pressed on it gently, sections just floated. The previous repair was basically a sticker holding itself on by friction.
Mistake One: Underestimating the Build-Up
My plan was to skim coat over the existing drywall patch with several coats of all-purpose compound until it was flush with the plaster. Simple enough — I've done this kind of build-up before. What I didn't account for was how much a full 3/8 inch of depth actually is when you're applying joint compound.
Joint compound shrinks as it dries. Standard all-purpose runs about 15-25% shrinkage depending on how thick you apply it. A 1/4-inch coat becomes maybe 3/16 of an inch when dry. To build up 3/8 of an inch with standard compound, you realistically need four or five coats with full drying time between each one — not the two I had planned for. In Portland, in February, in a room without heat, full drying time meant more than 24 hours per coat. I ran out of weekend before I ran out of build-up to do.
What I should have used from the start was setting-type compound (the kind that cures chemically rather than drying by evaporation) for the deep fill coats. A 20-minute or 45-minute hot mud sets regardless of humidity and can be built up to fill depth without worrying about ambient conditions. I could have done the fill work in one session, let it set, and then skim-coated with regular compound over it. Derek had to endure a second visit the following weekend.
Mistake Two: Not Matching the Texture
Derek's walls had the kind of surface texture that's hard to describe — not quite orange peel, not quite smooth, somewhere in between with slight variation across the wall. It looked to me like it had been skim-coated at some point and then lightly sanded, leaving a matte surface with slight trowel marks. I thought I could blend into it with a light skim over my repair area.
I was wrong. My skim-coated patch dried noticeably smoother than the surrounding wall, and the edge where new met old was visible even after painting. Not terrible, but definitely noticeable in certain light. Derek's sister noticed. She's an interior designer, which Derek had not mentioned.
Matching existing texture well — especially on walls that have been painted multiple times over 100 years — is genuinely hard. The method that actually worked on the third weekend was to wet-sand the transition zone lightly, apply a thin skim over a wider area (maybe three feet in each direction from the patch) using a wide knife, and then very lightly drag a balled-up piece of plastic sheeting across the wet compound to introduce some texture variation. Not exactly the original texture, but close enough that it read as the same wall under paint. I used two coats of primer before painting, which helped blend the slightly different sheen between new and old compound.
What Portland's Winter Did to the Timeline
I mentioned the humidity, but it's worth being specific. We were working in February with Portland's typical weather: temperatures in the low 40s outside, the room unheated most of the time, relative humidity indoors hovering around 65-70%. Derek had a space heater but was worried about his heating bill and kept it at lower settings than I wanted. We had no fan for air circulation.
Under those conditions, all-purpose joint compound was taking 18-20 hours to dry enough to recoat. The surface would feel dry to touch in 10-12 hours but was still soft and would dent under slight pressure. Recoating too early causes the wet compound underneath to pull through and you get lifting and cracking. I know this. I still did it once because I was optimistic and wanted to finish on Saturday. The coat I applied too early lifted at one corner within a few hours and had to come off.
The lesson is simple but frustrating: in wet Pacific Northwest conditions, budget double or triple your normal drying times for any project running from October through April. If that's not feasible, use setting-type compound for the heavy coats and only use regular premixed for the finish coats where drying time matters less because the layers are thin.
How It Came Out
Third weekend: patch is flush, texture is close, Derek painted over it with the original wall color and it looks fine. His sister apparently said nothing about the wall, which is either a compliment or she was being polite. I'm counting it as a win.
Total cost was around $45 including a small bag of 20-minute hot mud ($9), additional all-purpose compound, primer, and paint (Derek already had the color). Time was the expensive part — three separate weekends spread over a month, mostly because I tried to use cold weather and inappropriate materials for the base coats.
If I were doing this again in February in Portland, I'd heat the room to at least 65 degrees and keep it there throughout the project, use hot mud for any coat thicker than about 1/8 inch, and extend the repair skim out further than feels necessary so the texture blending has more room to work. And I'd ask Derek upfront whether anyone with an opinion on walls would be visiting.