disaggregation of data in proportional series when base series has null values

we have a series A that is proportional with the proportion series being B .when we enter data in series A at a higher level say product line (PL)we expect it break down to lower level (item) in the proportion in the series B but if  B has null values for all the items under a certain product line ,how does demantra break down the value entered at the higher level?as per the document , any item that has a null value in series B is ignored in the disaggregation but in our case we notice that if for a certain PL say PL1 having items ITM1,ITM2,ITM3 under it and all having null values in series B , the top level value still gets disaggregated into the three items and the total adding up to the PL level value.