2) IR value decreases by nearly half for every 10C rise in temperature.
But PI will not change with
temperature. IR value in air will be 10-20 times that under oil. 3) IR value of today's transformers is exceptionally high compared to those of units made 15-30years's back. This is because of the high resistivity of transformer oil manufactured nowadays. Today oil is made by hydrocracking (using catalysts with hydrogen at high pressure of couple of hundred bars and temperature around 800 C) the petroleum crude which removes all polar conducting elements in oil. This change over from acid treatment or solvent extraction happened nearly 15-20 years back. The result was extremely high resistivity for oil. (color changed from light yellow to water white) This is giving high IR to transformers and not because of inhibited oil. Of course these super refined oils are always inhibited type as all natural oxidation inhibitors will be removed during hydro-cracking and synthetic inhibitors are to be added to oil to get adequate oxidation stability. This oil resistivity sometimes masks the actual insulation condition resulting in moderate or high IR for even for wet transformers. So, Tan delta of windings may give a better evaluation of insulation than IR. 4) Transformer IR depends on the weakest link in circuit, usually neutral bushing/lead or tap changer IR to earth. Many times SRBP sheets will have high moisture content decreasing the IR use of FRP is a good move in right direction.IR may also get affected by the insulation level of the connecting leads used (esp. at 5kV) or the lead length. 5) Many of the minimum values of IR, polarisation index or absorption ratios given in old Megger guides or IEEE standards are no longer valid for new transformers because of the major change in the resistivity of oil. 6) Strictly speaking polarization index is relevant only for dry insulation and not for oil immersed equipment. This is mainly because of the ionic movements in oil under electric field. Absorption coefficient is not recommended due to the small time of voltage application involved.( There was a paper from China in 90's establishing this point) 7) With high IR values, as noticed in today's transformers, getting high PI values get more and more difficult for new transformers as the leakage current involved is extremely low. Rotating machinery engineers are aware of it. In IEEE std 43-2000(R2006) Testing Insulation resistance of Rotating Machinery, under clause 12.2, it is mentioned "When IR value is more than 5000 Mohms, PI may or may not be an indication of insulation condition and is therefore not recommended as an assessment tool" This standard is a good guide for understanding the theory of PI. 8) IEEE standard 62(R2005) is now under revision (D7) and may be issued as C57.152 in future. Even though PI value of 1.25-2.00 is retained as fair value, a word of caution has been added- The polarization index for insulation liquid is always close to 1.Therefore the polarization index for transformers with low conductivity liquids (egg new mineral oil) may be low, in spite of good insulation condition.