For well-guarded plane-parallel ionization chambers, international dosimetry protocols recommend a value of unity for electron perturbation factors in water. However, recent data published by various groups have challenged this. Specifically for the NACP-02 chamber, non-unity electron perturbation factors have already been published by Verhaegen et al (2006 Phys. Med. Biol. 51 1221-35) and Buckley and Rogers (2006 Med. Phys. 33 1788-96). Recently it was found that the mass thickness of the front chamber window can be 35% greater than is listed in the IAEA's TRS-398 absorbed dose protocol (Chin et al 2008 Phys. Med. Biol. 53 N119-26). This study therefore recalculated NACP-02 electron perturbation correction factors for energies 4-18 MeV at depths z(ref) and R(50) to determine the effect of the chamber model change. Results showed that perturbation factors at z(ref) are fairly stable for similar chamber models but become highly sensitive to small changes at deeper depths. The results also showed some dependence on using 1 keV versus 10 keV for the transport cut-off. Additional investigations revealed that the wall perturbation factor, p(wall), is strongly influenced by the chamber back wall at z(ref) and at larger depths small changes in the positioning of the effective point of measurement cause large fluctuations in the final value. Finally, the cavity perturbation factor, p(cav), was found to be primarily influenced by electron backscatter.