The useful range of the Kelvin equation (and its derivative methods e.g. BJH, HK) is limited at the narrow pore end by the question of its applicability and at the wide pore end measurements are limited by the rapid (logarithmic) change of the core radius with relative pressure. Modem methods of calculation from gas sorption data (e.g. OFT) confidently extend accurate pore size determination well into the micropore region. Pore diameters in excess of 500 nm are rarely reported from gas sorption data.

Mercury porosimetry has parallel constraints at the narrow pore end of its range, in that questions arise regarding the constancy of surface tension and wetting angle for mercury, not to mention the practical difficulties associated with very high pressure generation. Consequently, mercury porosimetry has a lower limit that does not approach the upper limit for micropores (2nm). However, at the large pore end, mercury porosimetry does not have the limitations of the Kelvin equation and, for example, at 0.0069 MPa pore volumes can be measured in pores of approximately 107 micrometer radius or 1.07 x 105 nm.

Comparisons have been made in the range where the two methods overlap and they generally show reasonable agreement. Zweitering obtained distribution curves from mercury porosimetry and the nitrogen isotherms on chromium oxide-iron oxide catalysts. Each curve showed a narrow distribution with the peak near ISnm, the nitrogen curve being slightly narrower and higher. Using nitrogen, Joyner, Barrett and Skjold obtained good agreement between the two methods. By adjusting the wetting angle for mercury of charcoals to values between 1300 to 1400 , they were able to closely match the curve produced from the nitrogen isotherm. Cochran and Cosgrove using n-butane, reported total pore volume of 0.4S8 cm3 and 0.373cm3 with adsorption and porosimetry, respectively. The maximum pore radius was under 50 nm and they attributed the difference to n-butane entering pores narrower than 3nm. Dubinin et al found that nitrogen and benzene both gave type I isotherms with hysteresis on several activated carbons. The pore distributions based on the Kelvin equation agreed with porosimetry measurements.

Comparisons have been made between surface areas measured by porosimetry and gas adsorption as well as by permeametry with results ranging from excellent to poor.

The most definitive surface area measurements are probably those made by nitrogen adsorption using the BET theory. Neither the Brunauer, Emmett and Teller (BET) theory, nor equation used to calculate surface area from mercury intrusion data, makes any assumptions regarding pore shape for surface area determinations. When these two methods are compared, there is often surprisingly good agreement. When the two methods do not agree, it does not imply the theoretical failure of either one. Indeed the differences between results from the two methods can be used to deduce meaningful information that neither alone can supply. For example, when the BET value is large compared to the area measured by porosimetry, the implication is that there is substantial volume of pores smaller than that penetrated by mercury at the maximum pressure.

If the porosimetry can generate 414MPa (60,000 psi) of hydraulic pressure, the minimum diameter into which the intrusion can occur will be about 3.5nm. Assuming that pores centered about 1.5nm are present and have a volume of 0.02cm3, an approximation of their surface area can be made by assuming cylindrical geometry. Thus, 

Therefore, the area measured by porosimetry would be approximately 53.3 m2 less than that measured by the BET method.

Another factor that can lead to BET areas slightly higher than those from porosimetry is pore wall roughness. Slight surface roughness will not alter the porosimetry surface area since it is calculated from the pore volume while the same roughness will be measured by gas adsorption.

Cases that lead to porosimetry-measured surface areas exceed those from nitrogen adsorption can result from inkbottle shaped pores having a narrow entrance with a wide inner body. Intrusion into the wide inner body will not occur until sufficient pressure is applied to force the mercury into the narrow entrance. It will appear, therefore, as if a large volume intruded into narrow pores, generating an excessively high, calculated surface area.