Must say I can’t see your problem.
If I clear range B6 to D8 and do a solver run I get this result:
Grade production per year.
x1 x2 x3
R1 1265,4 934,6 0 2200
R2 634,6 0 1665,4 2300
R3 0 1065,4 1334,6 2400
1900 2000 3000
So Solver does produce most of product x1 on line R1, the balance (634,3) is then produced on line R2. Finally there is no production on line R3 of product x1 as this is one of the constraints.
The driving force in this model (object function) is maximizing production of
X1 + x2 + x3 and this is what solver has done.
You got a production of 6900 (sum x1+x2+x3) and this is equal to the max capacity off your line system as given by the constraints 2200 for R1, 2300 for R2 and 2400 for R3.
You then split the yearly production into 3 periods doing a bit of mathematical wizardry. That’s one way of trying to simulate time in the LP model I assume.
Problem with LP models are that the concept of time does not exist so all components are available at the same time. This is seldom so in real life so it’s necessary to juggle numbers a bit in order to create a time effect.
Alf
Bookmarks