Does "A has B", "A has C", "B has D", "C has D" imply "A has two Ds"?

[0UmfHxcvx5J7JoaOhFSs5mncnisTJJ6q]

Question for @OpenEnergyPlatform/oeo-general-expert-formal-ontology: While looking at #1362, I noticed that

CCGT has some gas power unit → which has some power generating unit → which has part some generator
CCGT has some steam power unit → which has some power generating unit → which has part some generator

Does that imply that CCGTs have two generators, one in the "gas power unit" and one in the "steam power unit"? Because in reality they might have just one [↑].

[areleu]

I think this is easily resolved by looking at instances instead of universals.

A is an instance of CCGT
B is an instance of gas power unit
C is an instance of steam power unit
D is an instance of power generating unit
E is an instance of power generator
C is part of A, B is part of A
D is part of B, D is part of C
E is part of D
This means that B and C overlap at D (namely at E)
This means that either B is C or there is a F entity of which B and C are part.
from the definition we can say that F is A

This would be also congruent if there was indeed a power generating unit G different from D.

Edit: And I think this deduction is only valid if one considers Minimal Extensional Mereology. The definitive axiom would be SA6 from Simons, 1987:

overlaps(x, y) → ∃z∀w [is part(z,w) = is part(x,w)∧is part(y,w)]

I think the confusion comes from the fact that we do not really have axioms to differentiate proper parts from parts. I think it is because of owl limitations. AFAIK the is part relation we use in the OEO is not is proper part(former allows something to be a part of itself and the latter does not.), from what I see in the annotations (imported from RO).