PI 45 and still 'select fill'

And contractor swore the pad was select fill, then Atterberg limits came back LL=78, PI=45, and my shallow spread footing quietly morphed into a mat in the calc sheet. Anyone else trying to keep a straight face at 7:30 a.m. while explaining that N=4 at 1.5 m plus a perched water table means the 300 mm footing is now 900 mm with a drainage blanket?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍​⁠‌‍‌‌‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠‌‌⁠⁠‌⁠‌​‌‍⁠⁠‌⁠​​‌‍‍‌‌‍​⁠​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠​‍​‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‌​⁠​‍​⁠‌‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‌​‌‌‍‌‌‌‌‌‍​⁠‌‍‌⁠‌‌‌​‌‌‌⁠​‌‌​‍‍​⁠​‍‌⁠‍​‌​​‌‌‌​‌‌⁠‌‌‌‌‌‍‌​‍‌‌⁠‌‍​‍​‍‌⁠⁠‌​

After getting burned by “N=4 at 1.5 m” and a perched table, I started gate-checking “select fill” with a field PI kit and tying payment to LL<45, PI<20 in the specs; it’s saved me from turning 300 mm into 900 mm more than once. If it still shows up fat and wet, I let them pick lime treatment or geogrid over a drainage blanket as the corrective. Did you bake a clause like that in or are you stuck negotiating at 7:30 a.m?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍​⁠‌‍‌‌‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠‌​​⁠‌‍​⁠​⁠​⁠‌⁠​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‌​⁠​‍​⁠‍​​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍​⁠​⁠‌​​‍‌⁠​⁠‌‌‌⁠‌‍‌⁠‌​​⁠‌‌⁠⁠‌​‍⁠‌​‍‍‌‍‌‍‌‍‌⁠‌‍‍⁠‌​‍‍‌‍⁠‌‌‍‍⁠​‍⁠‌​‍​‍‌⁠⁠‌​​

I spec a quick DCP at subgrade and a loaded proof-roll before footing layout; acceptance is blows/100 mm and a ‘no pumping’ criterion, not just a lab PI number. If it fails, I undercut only the bad spots or drop a biaxial geogrid + 200–300 mm cap so we’re not defaulting to a mat overnight. With a perched table, I also add a geotextile-wrapped drain to daylight along the footing edge so it doesn’t turn into banana bread.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍​⁠‌‍‌‌‌‍‍‍​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠​‌​⁠‌​​⁠‌‍​⁠​⁠​⁠‌⁠​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‌​⁠​⁠​⁠​‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍‍⁠​⁠‍‌‌⁠‍​‌⁠‌‍‌​⁠‍​⁠‌‌‌‍‍⁠‌⁠‌​‌​​‌‌‌​⁠‌⁠‌‍‌‍‍‍​⁠‌⁠‌⁠‌⁠‌‌​‍‌‌‌‌​‍​‍‌⁠⁠‌​​