The European Space Agency Rosetta mission escorted comet 67P for a 2-yr section of its six and a half-year orbit around the Sun. By perihelion in 2015 August, the neutral and plasma data obtained by the spacecraft instruments showed the comet had transitioned to a dynamic object with large-scale plasma structures and a rich ion environment. One such plasma structure is the diamagnetic cavity: a magnetic field-free region formed by interaction between the unmagnetized cometary plasma and the impinging solar wind. Within this region, unexpectedly high ion bulk velocities have been observed, thought to have been accelerated by an ambipolar electric field. We have developed a 1D numerical model of the cometary ionosphere to constrain the impact of various electric field profiles on the ionospheric density profile and ion composition. In the model, we include three ion species: H₂O⁺, H₃O⁺, and $\mathrm{NH_4^+}$. The latter, not previously considered in ionospheric models including acceleration, is produced through the protonation of NH₃ and only lost through ion-electron dissociative recombination, and thus particularly sensitive to the time-scale of plasma loss through transport. We also assess the importance of including momentum transfer when assessing ion composition and densities in the presence of an electric field. By comparing simulated electron densities to Rosetta Plasma Consortium data sets, we find that to recreate the plasma densities measured inside the diamagnetic cavity near perihelion, the model requires an electric field proportional to r^-1 of around 0.5-2 mV m^-1 surface strength, leading to bulk ion speeds at Rosetta of 1.2-3.0 km s^-1.