FlagstoneReport
Does the work support the claim?
Assessment of an NI 43-101 technical report

Côté Gold Project

Near Gogama, Ontario, Canada (Porcupine Mining Division, ~125 km southwest of Timmins)
Study stage
FS · Feasibility Study
Published
November 2018 (feasibility study news release 1 November 2018; report filed 26 November 2018)
Prepared by
Wood Canada Limited (formerly Amec Foster Wheeler Americas Limited) in conjunction with IAMGOLD Corp.; QPs include Peter Oshust, P.Geo. (resources) and Dr. Antonio Peralta Romero, P.Eng. (reserves)
Source
FS_Cote_2018.pdf
Study-level anchor — not scored
FS Feasibility Study

Every judgment below is measured against what a FS is expected to demonstrate — not an absolute bar.

Detected from the filing

Title page and running header: 'NI 43-101 Technical Report on Feasibility Study'; Section 1.1 'results of a feasibility study (the 2018 Feasibility Study)'. Capital estimate 'prepared in accordance with the AACE International Class 3 Estimate with an expected accuracy of +15%/-10%'. Proven and Probable Mineral Reserves of 203 Mt (Base Case) / 233 Mt (total) are declared with a detailed LG pit design, phased production schedule and after-tax DCF (NPV5% $788M, IRR 14.5%). A prior PFS (2017) is referenced, confirming this is the next stage.

Peer Benchmarks

Peer benchmarks for this route are provisional — the peer set is small, so figures are shown as position and context, not firm green/amber/red verdicts.

Provisional peer context
Au Recovery91.8%

Provisional peer context (n=7) — shown as position only, not a firm green/amber/red verdict.

61.6%96%
Peer range 86%–93% · full span 61.6%–96%
Internal consistency
Recovery claim vs. testwork91.8 → 93%

The claimed 91.8% carried into the economics is backed by the 93% testwork result.

Not benchmarked at FS stage
Capex Intensity$97 /t·yr

Reference only — no firm peer band at this stage; shown for context.

Opex/t processed$14.52 /t

Reference only — no firm peer band at this stage; shown for context.

These peer positions are context to investigate, not a verdict. A provisional or reference peer set is limited and shown as context, not a firm band; on their own these positions do not decide the grade.

Flagstone Score

We grade two crucial dimensions, then determine a final score from them.

AExceeds

Methods robustness

Does the disclosed work match what the study stage demands?

Judged against the FS column of the stage-scaled matrix, essentially every required element is present and several are well beyond the FS minimum. (a) Bench leach 'required, extensive': 93 variability composites and 162 tests, with the test condition matrix published (Table 13-4). (b) Leach kinetics 'required': kinetic behaviour disclosed — 'overall extraction of 91% after 30 hours, reaching a plateau average extraction of 93% … at 48 hours', plus the finding that oxygen enhances kinetics and permits NaCN as low as 0.3 g/L. (c) Grind–recovery 'required': 'The positive effect of grind on extraction was recognized early … Each program to date has collected data on this aspect'; SGS 12589-003 compared grind effect for WOL vs gravity-tailings flowsheets and 'regression coefficients between grind size and extraction suggests that grind is indeed the main driver'; additional 2017 grind-vs-extraction work on the five-year composite. (d) GRG 'required if gravity in flowsheet': six Laplante E-GRG tests (61–74%, fine-to-moderate classification) plus explicit circuit-placement modelling (primary / secondary / both) with the honest conclusion that fine GRG and a two-stage mill 'makes gravity recovery more challenging'. (e) Reagent optimization 'required, optimized': optimal in-plant NaCN dosage window 0.3–0.5 g/L established, oxygen shown to reduce cyanide and lime consumption, consumption benchmarked (~100 g/t), with a specific explanation of the anomalously higher consumption in the five-year composite (sample aging; air-sparge volatilization). (f) Variability 'many, spatially representative': 93 composites, spatial locations plotted against the 0.25 g/t grade shell (Figure 13-1), plus a dedicated 2016 six-hole/1,422 m metallurgical drill programme chosen 'on the basis of the prevalent lithology–alteration groupings'. (g) Oxygen/lime optimization and CIL-vs-CIP selection 'expected': oxygen optimization done and CIP specifically modelled with Mintek semi-empirical models to 'check the robustness of the carbon-in-pulp (CIP) option'. (h) Deleterious elements: sulphide mineralogy quantified (1% and 0.06% S-minerals), barren solution metal dissolution low, 'no cyanicides … except for small amounts of S and Fe', ABA studies, and SO2/air detox demonstrated to <1 mg/L CNT — assessed and mitigated, though the statement 'no obvious concerns with deleterious elements' is more qualitative than a full FS quantification. (i) HPGR pilot-plant work plus independent third-party HPGR review exceeds the FS bar on comminution/flowsheet justification. Residual soft spots that keep this from being unqualified: the ~10% high-copper population (Master Composite B) is composited and tested but no separate recovery treatment for it is presented; the gravity recovery number is self-described as 'indicative' with 'a lower confidence level'; and the 91.8% is carried forward unchanged from the PFS rather than re-derived from the FS-stage optimization work. The QP's own Section 13 comments and the Section 1.25 recommendation to 'complete additional HPGR and cyanidation tests' confirm testwork is not fully closed out, and the piston-press result that 'the variability indicates a risk to achieving throughput for the harder ores' is a QP-flagged risk, but these are design-margin issues rather than missing stage-required work. Framework signal (below) is supportive — deportment-grounded, operating-condition-based, staged recovery build-up — and holds the letter at A rather than pulling it to B.

  • Report omissionaffects the gradeNo separate recovery relationship or design recovery is presented for the high-copper material represented by Master Composite B, stated to be ~10% of the deposit, even though copper can affect cyanide consumption and leach behaviour; the single 91.8% is applied deposit-wide.
  • Report omissionaffects the gradeDeleterious-element treatment at FS is qualitative ('metal dissolution … was found to be low, and there are no obvious concerns'), based on barren solution analysis from 'early scoping programs' on Composites 1, 2 and 3, rather than a quantified assay-level assessment across the variability suite.
  • Report omissionaffects the gradeThe gravity recovery entering the recovery build-up (23%) is acknowledged as 'indicative' with 'a lower confidence level' for the secondary-circuit model, and sits below the 32% average gravity recovery measured in the leach tests and well below the 61–74% E-GRG; no reconciliation of these three gravity numbers is given.
  • Report omissionaffects the gradeSilver deportment and recovery are not tested or reported ('The testwork does not report on silver recovery'); silver is consistently <2 g/t, so this is immaterial to economics but is a disclosed omission.
  • Data gapnot in the extracted text — does not affect the gradeDetailed annual production/recovery schedule and the recovery actually applied year-by-year in the cash flow (including ramp-up recoveries referenced in Section 1.14) sit in Sections 16/21/22, which were not provided.
  • Data gapnot in the extracted text — does not affect the gradeNumerical content of the grind-vs-extraction and recovery-vs-head-grade plots (regression coefficients, individual test points) is referenced in text but the underlying figures/tables are not reproducible from the extracted sections.
Metallurgical framework quality
Considers operating conditions beyond head gradeyes
Metallurgical domains defined, each with its own recoverypartial
Range of lab results explained by the frameworkpartial

Explained vs. unexplained The framework explains most of the observed lab range through operating conditions rather than head grade: grind size is identified as 'the main driver' by regression against extraction, with oxygen and NaCN dosage controlling kinetics, and the recovery is built up stage-by-stage (gravity 23% → leach 90.9% at 100 µm → intensive leach 99% → CIP 99% → desorption/refining 99.5%). Mineralogy underwrites the structure: free-milling, 19.8% liberated gold, 54% 'locked' at K80 150 µm (hence the grind sensitivity), 25% unknown of which 75% leachable, no cyanicides. What remains unexplained: the actual spread of the 93 variability results is never shown numerically — only the aggregate statements 'all the variability samples were readily amenable' and 'average … 91% after 30 hours … plateau average 93% at 48 hours' — so the reader cannot see how wide the recovery range is or where the tail sits. The ~10% high-copper population is tested separately but is not given its own recovery relationship. The report is commendably candid that the apparent lithological uniformity 'may be a consequence of the gravity step ahead of leaching, which removes liberated gold to produce a more uniform leach feed sample', i.e. it explicitly names the one mechanism that could be masking residual variability. The five-year composite's better-than-average extraction is explained ('likely related to the higher than average gravity recoverable gold content'), and its higher cyanide consumption is explained (sample aging, air sparging).

BAdequate

Result representativity

Are the reported numbers trustworthy as reported?

Criterion 1 (economics-carried recovery): strong. 91.8% is explicitly metallurgical, is decomposed into its component stages in Table 13-5, and corresponds to an average head grade of 0.94 g/t Au — essentially the LOM reserve grade of 0.97–0.98 g/t, not a high-grade first-years grade; the gold-trap of substituting a payable recovery is avoided (refining is separately charged at $1.75/oz and recovery through desorption/refining is itself a line in the build-up). Criterion 6 (internal consistency): strong — 91.8% appears unchanged in the resource pit-shell assumptions, the reserve notes and the optimization input table, with no competing recovery figure anywhere in the provided text; the measured 93% plateau extraction minus CIP/refining losses reconciles to the 91.8% overall. Criterion 4 (mineralogy grounding): strong — recovery is tied to a quantified deportment study (native gold/electrum, 19.8% liberated, 54% locked at K80 150 µm, 25% unknown of which 75% leachable, average grain 12.5 µm), which directly explains why grind, not head grade, drives extraction, and the ore is explicitly characterised as free-milling/non-refractory with no cyanicides and no preg-robbing carbon. Criterion 3 (poor results shown): good — the report discloses unflattering items rather than burying them: Atwal specific wear rate 'classified as high'; piston-press specific energy variability where 'the variability indicates a risk to achieving throughput for the harder ores'; the gravity estimate being only 'indicative' with 'a lower confidence level'; the higher cyanide consumption of the five-year composite; and the caveat that lithological uniformity may be an artefact of the gravity step. Criterion 5 (basis disclosure): adequate — grind, throughput, residence time, reagent regime and each recovery sub-factor are stated, but no operability/scale-up discount is discussed and the report simply says the estimate 'remains as set during the pre-feasibility study', so the reader is not told whether any conservatism was applied or how the FS-stage optimization results were allowed to move (or not move) the number. Criterion 2 (variability behind the averages): the weak link, and the reason this is B not A. Despite 93 composites and 162 tests, the provided Section 13 gives no test-by-test or even min/max recovery table — only averages ('average gravity recovery of 32%', 'overall extraction of 91% after 30 hours', 'plateau average extraction of 93%') and the blanket assurance that 'all the variability samples were readily amenable to gravity concentration and cyanide leach'. At FS the recovery spread across the variability suite should be visible. Combined with the absence of any separate recovery for the ~10% high-copper population and the unreconciled 23% design / 32% measured / 61–74% E-GRG gravity numbers, the presentation is trustworthy and internally consistent but under-exposed on range. Peer position (provisional, limited n): 91.8% sits just above the PFS/FS pooled gold recovery median of 91.0% and below the p75 of 93.0% — an unremarkable, non-outlying position for a free-milling CIL/CIP project.

  • Report omissionaffects the gradeNo test-by-test or ranged presentation of the 93 variability leach results — recovery spread (min/max, distribution by lithology or grade bin) is not tabulated; only averages and a qualitative 'all readily amenable' statement are given, which is thin for an FS.
  • Report omissionaffects the gradeThe three gravity figures — 23% used in the recovery build-up, 32% average measured in the leach test programme, and 61–74% E-GRG — are never reconciled against each other, and the 23% rests on simulations the report itself calls 'indicative' with 'a lower confidence level'.
  • Report omissionaffects the gradeThe 91.8% is stated to be carried over unchanged from the PFS; no statement is made of whether an operability, scale-up or ramp-up discount is embedded, or how the 2017–2018 optimization testwork on the design composite affected (or confirmed) the figure.
  • Report omissionaffects the gradeRecovery for the ~10% high-copper material (Master Composite B) is not separately reported, so the reader cannot see whether the deposit-wide 91.8% is diluted by a lower-recovery copper-bearing sub-population.
  • Data gapnot in the extracted text — does not affect the gradeThe ramp-up recovery profile ('associated recoveries, which steadily increase to reach the design capacity after 10 months') is referenced in the mining section but the actual annual recovery values used in the cash flow live in Sections 16/21/22, which were not provided.
Flagstone Score
B

Work solidly supports the stage; minor reservations.

How the grade was reached

The letter is a deterministic roll-up of the two dimension grades and the peer flags — the same inputs always produce the same grade.

  1. Recovery: claimed 91.8% (carried into economics) vs measured 93% -> gap -1.2 pts
  2. Gap not material -> no representativity penalty
  3. Weakest link: methods=A, representativity=B -> take the lower = B
  4. Weighted penalty = -0 notches, capped at -3 and floored to whole notches -> -0
  5. Final grade (floored at D) = B

Study level is unambiguously FS (title, AACE Class 3 ±15%/-10% estimate, Proven and Probable Reserves, detailed LG pit design and after-tax DCF), so both dimensions are judged against the FS column. METHODS — A (Exceeds). Every FS-required testwork element in the gold CIL/CIP matrix is present and several are well beyond the FS minimum: extensive bench leaching (93 variability composites, 162 tests, test-condition matrix published), leach kinetics with a disclosed 30 h / 48 h plateau, a grind–recovery relationship established by regression across multiple programmes and identified as the dominant driver, six Laplante E-GRG tests plus explicit gravity-circuit placement modelling, reagent optimization establishing a 0.3–0.5 g/L NaCN window with oxygen benefit quantified, Mintek CIP modelling that actively checks the robustness of the CIP choice, SO2/air cyanide destruction to <1 mg/L CNT, thickening/rheology work, and an HPGR pilot plant with independent third-party review at each stage. The metallurgical-framework signal supports this: the recovery model is not a head-grade curve but a deportment-grounded, operating-condition-anchored staged build-up (gravity 23% → leach 90.9% at 100 µm → CIP 99% → refining 99.5%), with a quantified gold deportment study explaining why grind rather than grade governs extraction, and with head grade, lithology and alteration explicitly tested and rejected as determinants. Formal metallurgical domains are not carried into the recovery model, which would normally be a deduction at FS, but the deposit is demonstrably free-milling, non-refractory and mineralogically uniform, and the report supplies the data supporting that — so the undomained treatment is justified rather than naive, and the framework holds the letter at A. The residual methods soft spots (qualitative rather than quantified deleterious-element closure, no distinct treatment of the ~10% high-copper population, and the QP's own recommendation to 'complete additional HPGR and cyanidation tests') are design-margin issues, not missing stage-required work. REPRESENTATIVITY — B (Adequate). The 91.8% is a metallurgical (not payable) recovery, tied to a 0.94 g/t head grade that matches the 0.97–0.98 g/t LOM reserve grade rather than high-grade early years, decomposed into auditable sub-factors, and repeated without conflict across the resource pit-shell assumptions, the Whittle optimization inputs and the reserve notes — internal consistency is a genuine strength here. Mineralogy grounding is strong and unflattering results are disclosed rather than suppressed (high Atwal wear rate, hard-ore throughput risk from piston-press variability, 'indicative' low-confidence gravity estimate, the candid caveat that lithological uniformity may be an artefact of the gravity step). What holds it at B is criterion 2: with 93 composites and 162 tests available, the report shows only averages and a blanket 'all the variability samples were readily amenable' — no recovery spread, no min/max, no per-sample table — which is thin disclosure at FS; this is compounded by the unreconciled 23% / 32% / 61–74% gravity figures and the statement that the recovery simply 'remains as set during the pre-feasibility study' with no stated operability or scale-up basis. Per instruction 7, no deduction has been taken here for any claimed-versus-measured recovery gap; on the extracted numbers the 91.8% economics recovery is in fact below the 93% measured plateau extraction once CIP and refining losses are applied, so the two reconcile. PEER POSITION (context only, not a verdict). Gold recovery 91.8% sits just above the PFS/FS pooled gold recovery median of 91.0% and below p75 of 93.0% — a mid-pack, non-outlying position for a free-milling CIL/CIP operation; this peer set is provisional with limited n and all band-eligible CIL rows are Canadian, which Côté also is, so the comparison carries a jurisdictional-concentration caveat and should not be read as a firm band test. Capex intensity of roughly $98 per annual tonne of ore ($1,281 M over ~13.1 Mt/a) is reference-only: it sits between the p75 ($75.6) and max ($156.6) of the PFS/FS open-pit capex-intensity spread, and below the pooled PFS/FS median of $152. Total opex of $14.52/t processed (processing alone $6.32/t) sits near the p25 of the pooled PFS/FS opex distribution ($13.91), consistent with the report's own finding of low cyanide and lime consumption in a clean, cyanicide-free ore. No flag is assigned on any cost metric. The final letter is computed in code from the two dimension letters and the peer flags.