How do I keep my gravel driveway apron from washing out where it meets the road?

I fixed nearly the same setup at my last house, sloped apron washing out onto the street every storm, and the answer was boring: it's all about what's underneath. The dip at the pavement edge forms because the dirt there is beat to mush from years of water funneling down and cars rolling over it. You can pile all the fresh stone you want on top of mush and it'll pump and wander every time a tire hits it. So yes, dig out the soft stuff first. Excavate that dip and a few feet back up the slope untill you hit ground that doesn't squish, then lay down a geotextile fabric over the subsoil. That keeps the stone from slowly disappearing into the mud, which is half of why you keep losing material. Then build the apron back up with somthing that locks togather, like crusher run or crushed concrete, in a couple of lifts, packing each one with a plate compactor. Loose round stone never compacts no matter how much you tamp it, so if what's there now is rounded river gravel, scrape as much of it off as you can and use it elsewhere. Crown it so water sheds to the sides instead of running straight down the middle toward the pavement, and finish with a coarser top layer, inch and a half minus or similar. The fines in that material knit together and resist rolling way better than clean stone. On the plastic grids, I've used them on a steeper section and they do work, but only as a supplement. They hold stone in place on the surface, they don't fix a soft base. If you put grids over that dip without digging it out you'll be back where you started by spring. If the budget only stretches to one or the other, spend it on excavation, fabric, and good compacted base. A properly built apron with no grid will outlast a grid thrown over bad ground.

Do it now while the weather's dry. Trying to excavate and compact that area after the fall rains start is miserable and the results are worse.

It makes sense that if the ground underneath is soft, nothing we put on top will last. I noticed when I was looking at my driveway that the whole area near the curb was just collecting water and it looked like it was always pooling right there, even when it hadn't rained hard. If I follow the advice about crowning it up, does the angle of the new gravel need to be really steep for the water to move off the side fast enugh? I just want to make sure it doesn't stay sitting right by the edge of the street.

Been fighting that same battle at my old shop, and one thing that helped that hasn't come up yet is putting a header of bigger stone right where the drive meets the hard top. Hand-set a row of large flat rocks across the bottom of the slope years ago, old school style, and it acts like a dam so the loose stuff quits walking off the edge. Before that I was out there with a shovel and a bucket every spring, scooping gravel back up the hill. Also keep in mind that when your tenant's plowing or the road crew scrapes that apron, they'll drag stone onto the pavement anyway, so leave yourself a pile for touch-ups come March.

The geometry of the crowning really matters for effective runoff, because simply having a slope is insufficient if the angle of repose for the top layer isn't optimized for the terrain. If you are trying to force water across a steep grade, you increase the shear stress dramatically, which can actually erode the upper aggregate even if the subbase is perfect. I would suggest that the crown should not just be steep, but that the final gravel layer near the curb should transition gradually enough to manage velocity without creating concentrated scouring where the sheet flow meets the hard pavement. It's a tricky balance between rapid drainage and resisting that lateral drag. I also found that even with a good base, the jointing between the newly placed apron and the existing pavement edge needed supplementary attention. Sometimes, creating a small, slight shoulder built up with finer crushed stone imediately adjacent to the slab helps prevent the water from simply bridging the gap and undermining the transition point. It's less about the overall slope and more about the immediate boundary condition.

I'd push back a bit on writing off the grids as just a bonus. On my old place the slope was bad enough that even a well packed base with good angular rock still crept downhill over a couple winters. What finally stopped it was laying a grid section over the repaired apron. The base repair matters most, sure, but on grades like this the grid is doing real work, not just decorating. My take is that for daily truck traffic the two together are what actually holds, and treating the grid as optional is how you end up raking rock off the road every few weeks.

Quick update - trying out the suggestion about putting a header of bigger stone where the driveway meets the hard pavement. It helped a ton by stopping the material from washing down the slope. But I'm still having trouble with the water pooling up right at the transition zone. Does the type of aggregate underneath make a big difference, or is the problem purely about the surface angle?