On: “Effect of surface polarization on resistivity modeling” by D. Guptasarma (GEOPHYSICS, v. 48, January 1983, p. 98–106).

Abstract

I am glad to see that this paper confirms the resistive behavior of metallic models when immersed in electrolyte and subject to low‐frequency currents, which my coauthor and I reported in 1978 (Saydam and Duckworth, 1978). It is also gratifying that this paper confirms the transition from resistive to conductive behavior with increase of frequency which we reported in that same paper. However, I must call into question the validity of the conclusions reached by the author of this paper. I do so for several reasons, the first of which is that the author is not entitled to display complex resistivity spectra derived in a model tank environment unless evidence is provided that those spectra are invariant for a range of primary currents extending over at least two decades from an upper limit of 1 mA. I say this because a current of 1 mA, as apparently used by the author, causes severe nonlinearity in tank modeling as my coauthor and I showed in 1978 (Saydam and Duckworth, ibid). In tests of stainless steel which we performed and in tests on pyrite performed by Anderson and Keller (1964) the maximum permissible current density for linear conditions to exist was [Formula: see text]. Current densities at the surface of the aluminum cylinder of 3 cm diameter used by the authors would have been as high as [Formula: see text] when it was located at a depth of 7 mm (as quoted) directly under one of the current electrodes if the delivered current was 1 mA. It seems unlikely that aluminum would behave linearly at current densities 32 times greater than the maximum permissible for either stainless steel or pyrite, but we have no way of knowing because this paper fails to provide any experimental evidence in this regard. However, in the case of the results for a vertical stainless steel cylinder shown by the author in Figure 6 we can readily compute that if a 1 mA input current was used, the current density at the surface of the cylinder closest to the [Formula: see text] electrode would have been [Formula: see text]. Thus, in this case, linearity could only have been ensured by using an input current at least an order of magnitude lower than the 1 mA quoted by the author.