The datasets were combined for each structural zone in order to identify gaps in the data, primarily for DTR and concentrate analyses. These gaps were the result of unavailable data i.e. sample analysis that is still in progress (HIO), original non-sampling, mainly by CAP, of the oxide/transition zone, and insufficient material for analysis due to low recovered magnetic fraction in the samples (CAP/HIO). There were also gaps in the downhole geophysical data due to hole collapse preventing the probe passage to the bottom of the hole (for both CAP and HIO) and some probe results are still being processed (HIO).

Very minor gaps in the handheld magnetic susceptibility data were noted generally due to measurements not having been made for whatever reason (CAP/HIO). Downhole magnetic susceptibility data from both slimline surveys and handheld instruments were used to supplement the unavailable DTR composite data via the use of regression equations for the different structural domains, different companies and different oxidation levels. A total of 8,918 5m composites were generated from the H&SC drillhole database and subsequent data processing, which were modelled for DTR, and the concentrate elements of Fe, Al2O3, P, S, SiO2, TiO2 and LOI, of which 4,315 composites were in fresh rock.

Iron head grade was not modelled at this time due to significant amounts of unavailable data and no way of using regression equations to estimate the unavailable head grades. The unavailable head iron assays are currently being measured at the laboratory. Grade interpolation of the composite data was completed with Ordinary Kriging using the H&SC in-house GS3M modelling software.

Estimation within each structural domain (acting as hard boundaries) was unconstrained by any lithologic wireframes as the nature of the mineralisation boundaries is considered gradational throughout. A block size of 50m by 25m by 10m was considered appropriate based on the locally 100m spaced drilling at Core West. Within these domains separate search sub-domains were delineated to cater for changes in dip and strike of the sedimentary package.

A total of 10 search sub-domains were used. A four pass search strategy was applied with an circular search rotated to the dip and strike of the host sediments within each of the search domains. Initial search radii of 150m (X) by 150m (Y) by 25m (Z) were applied and expanded in two increments to 450m by 450m by 75m.

The maximum number of data for all passes was 24 with the minimum number of data for Passes 1 and 2 being 12, with a minimum of 4 octants, decreasing to a minimum of 6 data and 2 octants for Passes 3 and 4. The search radii and data requirements are in keeping with the strong stratigraphic control to mineralisation. The estimated grades were loaded into a Surpac block model for further processing and resource reporting. Any estimated grades within the cover were removed as the accuracy of these estimates was more questionable considering the nature of the hosting material.

A total of 8,918 5m density composites were generated from the short-spaced density measurements from the downhole geophysics. This data was supplemented by the use of oxide/transition and fresh rock regression equations from the head iron assays generated for the 2017 Mineral Resource update. Due to unavailable Fe head data, it was also necessary to insert default density values appropriate to the level of oxidation.

Validation work for the downhole density measurements comprised density measurements on a combined 350 core samples (weight in air/weight in water on 10-15cm long samples) indicated a correction of +5.2% was required for the CAP downhole data and +4.94% was required for the HIO data. The density values were interpolated in the same way as for the DTR grade interpolation. On the whole, no significant change in overall density was observed with the inclusion of the new drilling data, although some further checking on the 2021/2022 downhole calibration for some holes is required.

The estimation search pass categories were then viewed in conjunction with other aspects important for Mineral Resource classification and assigned a resource category. Features that were considered included geological understanding, the nature and controls to mineralisation, the data spacing i.e. drillhole spacing, impact of unavailable data and the minimisation methods employed, the density data, sample recoveries, sampling techniques and assay method, and the QAQC program and outcomes. Pass 1 was equated to Measured Mineral Resource, Pass 2 became Indicated and Passes 3 and 4 were Inferred.

A review of the resource categories showed `spotted dog' issues (artifacts of the grade interpolation) with the Measured Resource.