Transformer high excite at H0H3 and turns ratio at H0H3 deviates by 6% but winding resistances ok? 1

[chao_david]

Hi all. I have a 110-34.5-13.8 kV, ynyn0, NLTC, 100 MVA Toshiba power transformer (40 years old). Long story short, Buchholz relay activated. Here are the summary of results.

Insulation power factor, capacitances = ok
Insulation resistances and PI = ok
Arresters and bushings = ok
Winding resistances = ok
Turns ratio test = H0H3 up by 6% from calculated
Excitation test = unable to excite H0H3 at 10 kV
Acetylene content = from 0 to 78 ppm
Frequency response analysis = shown in image is healthy vs faulted H0H3 SFRA plot

1. Can I assume that there are no shorted turns since my winding resistances are almost the same as the previous test?
2. The only thing I can think of good winding resistance but bad ratio is when the the windings are axially deformed. However, I expect the turns ratio to decrease (which is not the case) since stretching the windings may result in turns loss.
3. SFRA changes mainly points to the area of the core which could indicate core damage? But how common are cores being damaged without damaging the windings?

3 weeks to go before internal inspection so might as well get some insights from here.

Cheers!