Mudman is really asking how do you determine the Qu based on SPT? Are you using Meyerhoff's direct capacity based on N? Or are you using N to determine phi - where you then go to charts of Nq, Ngamma and Nc? If one looks at all the different estimations of Nq based on phi, there is a very wide range - Berentzev, Meyerhoff, Terzaghi, Vesic among others. Your correlation Qupile = m*Qu-SPT is hugely dependent on which Nq value you use - and if you are using the direct N to Qu equations of Meyerhoff, they will be off similarly. It's all a crap-shoot and estimation.