Context. Infrared dark clouds (IRDCs) are cold, dense regions of the interstellar medium (ISM) that are likely to represent the initial conditions for massive star and star cluster formation. It is thus important to study the physical and chemical conditions of IRDCs to provide constraints and inputs for theoretical models of these processes. Aims. We aim to determine the astrochemical conditions, especially the cosmic ray ionisation rate (CRIR) and chemical age, in different regions of the massive IRDC G28.37+00.07 by comparing observed abundances of multiple molecules and molecular ions with the predictions of astrochemical models. Methods. We have computed a series of single-zone, time-dependent, astrochemical models with a gas-grain network that systematically explores the parameter space of the density, temperature, CRIR, and visual extinction. We have also investigated the effects of choices of CO ice binding energy and temperatures achieved in the transient heating of grains when struck by cosmic rays. We selected ten positions across the IRDC that are known to have a variety of star formation activity. We utilised mid-infrared extinction maps and sub-millimetre (sub-mm) emission maps to measure the mass surface densities of these regions needed for abundance and volume density estimates. The sub-mm emission maps were also used to measure temperatures. We then used Instituto de Radioas-tromía Milimétrica (IRAM) 30 m observations of various tracers, especially C18O(1-0), H13CO+(1-0), HC18O+(1-0), and N2H+(1-0), to estimate column densities and thus abundances. Finally, we investigated the range of astrochemical conditions that are consistent with the observed abundances. Results. The typical physical conditions of the IRDC regions are nH ~ 3 × 104 to 105 cm−3 and T ≃ 10 to 15 K. Strong emission of H13CO+(1-0) and N2H+(1-0) is detected towards all the positions and these species are used to define relatively narrow velocity ranges of the IRDC regions, which are used for estimates of CO abundances, via C18O(1-0). We would like to note that CO depletion factors are estimated to be in the range fD ~ 3 to 10. Using estimates of the abundances of CO, HCO+, and N2H+, we find consistency with astrochemical models that have relatively low CRIRs of ζ ~ 10−18 to ~10−17 s−1, with no evidence for systematic variation with the level of star formation activity. Astrochemical ages, which are defined with a reference to an initial condition of all H in H2, all C in CO, and all other species in atomic form, are found to be <1 Myr. We also explore the effects of using other detected species, that is HCN, HNC, HNCO, CH3OH, and H2CO, to constrain the models. These generally lead to implied conditions with higher levels of CRIRs and older chemical ages. Considering the observed fD versus nH relation of the ten positions, which we find to have relatively little scatter, we discuss potential ways in which the astrochemical models can match such a relation as a quasi-equilibrium limit valid at ages of at least a few free-fall times, that is ≳0.3 Myr, including the effect of CO envelope contamination, small variations in temperature history near 15 K, CO-ice binding energy uncertainties, and CR-induced desorption. We find general consistency with the data of ~0.5 Myr-old models that have ζ ~ 2-5 × 10−18 s−1 and CO abundances set by a balance of freeze-out with CR-induced desorption. Conclusions. We have constrained the astrochemical conditions in ten regions in a massive IRDC, finding evidence for relatively low values of CRIR compared to diffuse ISM levels. We have not seen clear evidence for variation in the CRIR with the level of star formation activity. We favour models that involve relatively low CRIRs (≲10−17 s−1) and relatively old chemical ages (≳0.3 Myr, i.e. ≳3tff). We discuss potential sources of systematic uncertainties in these results and the overall implications for IRDC evolutionary history and astrochemical models.