We have Projects where we need to deliver a subset of the main data as an EDIF file. To achieve this we created a new SITE/PROJECT and reimported the desired subset. 

However there is a major issue with this due to the Lab_Reports table. As the Lab_Report_Number is not unique to any one PID - If we need to do a lab chemistry reissue - we can see ESdat wants to delete the Lab_Report first. 

This also actually deletes the duplicated samples in the Export copy of the Project. However the reimported reissued chemistry goes back into the Master Project, but not the duplicate. 

I can see whats happening. But im not quite sure why Lab_Report_Number  is independent of PID? If we have any duplicate Projects sitting in ESdat - is there any way to delete these Projects and NOT have it strip out chemistry from the Master project for instance ?

Its because, for a variety of reasons to do with client confidentiality and reporting requirements. We need to have a deliverable and non deliverable subset. 

In that instance I could have possibly achieved the same results with a seperate PID for the deliverable data. However i had to ensure it covered all data - not merely chemistry - eg Geology Bore_depth_Descriptions and so on. Creating an Export deliverable data set got us around the issue -but it entangled with the Lab_Reports a bit. 

The second case was more tricky. Here - a separate client handed us data we had generated for another client ( With formal approval for this) But we needed to incorporate this into other information -and it needed to be the formal dated snapshot... So to Sites - with the same chem data on the server.

These are rare situations - but seem to happen on the big projects.