A.     If both an assembly and one or more of it’s subassemblies are in a forecast, then both the forecasted subassembly and the subassembly for the forecasted assembly will have demands.  This is how the system is intended to function. This is so that users can forecast requirements for spare parts if needed. I don’t think there is any other way to have spares built except to use a sales order instead of a forecast. Using a firm-planned work order for spares will cause the spares to be consumed by the assembly order. 

So, if you have a Top level Assembly A and a Sub level Assembly B that reports to A; and have a forecast at Assembly B but final Sales Order is only at Assembly A  

Perhaps the way to deal with this is to create a firm-planned work order for the subassembly (instead of a forecast) due on the date they placed the forecast for the subassembly. Then when the order is placed for the assembly (due after the work order for the subassembly), MRP will recognize the availability of the subassemblies and will not “double-drive” the requirements.