The values of the partial zeta functions for an abelian extension of number fields at non-positive integers are rational numbers with known bounds on their denominators. David Hayes conjectured that when the associated fields satisfy certain algebraic conditions, the bound at s=0 can be sharpened. I will present a counterexample to Hayes's conjecture. I will then propose a new conjecture sharpening the bounds at arbitrary non-positive integers that implies a weaker version of Hayes conjecture at s=0. I will conclude by proving that the new conjecture is a consequence of the Coates-Sinnott conjecture.