This guide explains how TSM Mining scores work as a geologist's 5-point audit threshold reference. It shows how to verify each of the five performance-area thresholds against the current MAC TSM protocol indicator list and self-assessed evidence before citing or acting on any score.
How TSM Scoring Actually Works
The Mining Association of Canada (MAC) administers the Towards Sustainable Mining (TSM) program, under which participating facilities self-assess against published protocols and submit results to external verification on a recurring cycle. Before citing any TSM score in a procurement or partnership decision, pull the current protocol version directly from MAC's official website — indicator wording shifts between versions, so a score earned against last year's framework may not fully reflect this year's expectations. Confirm the version date printed on the protocol document itself before relying on any indicator list.
TSM evaluates facilities across five distinct performance areas: indigenous and community relationships, safety and health, crisis management, tailings management, and biodiversity conservation. Each area contains multiple indicators, and each indicator receives its own independent level result. A facility's public score is not a single blended number — it is a set of level results spanning all five areas and their constituent indicators. Reading only the summary page without examining each area's individual results is where most misinterpretations begin.
The scoring scale runs from C through B, A, and AAA, with C representing baseline commitment and AAA representing the highest demonstrated performance level. A facility scoring AAA in four areas but B in tailings management carries a fundamentally different risk profile than one scoring B uniformly across all five. Averaging or eyeballing a "strong overall" TSM result obscures exactly the kind of gap that matters most to communities, regulators, and downstream customers who depend on specific safeguards.
This is why per-indicator scrutiny matters. A strong safety and health result does not compensate for a weak indigenous and community relationships score, and a tailings management result sitting below your organization's minimum threshold — regardless of how polished the headline average appears — warrants follow-up before any contractual commitment is finalized.
Before relying on any TSM score, verify two things: first, that the result was assessed against the most current protocol version available on MAC's site, and second, that external verification has been completed rather than self-assessment alone. Unverified self-reported results are preliminary by design and should be flagged as such in any due diligence record.
The Evidence: What You Can Verify
Verification starts with the primary protocol document, not a summary page or a vendor's marketing sheet. Download the current Mining Association of Canada (MAC) TSM protocol PDF for the relevant sector and pull two things for each of the five performance areas: the exact indicator list and the level-definition rubric that maps each indicator to a specific achievement level. Score nothing from memory. If the protocol in front of you does not contain the indicator you are about to cite, you do not have a verifiable claim — you have a recollection.
Next, obtain the facility's published TSM self-assessment. MAC requires member facilities to report their results publicly, which means the self-assessed level for each performance area should be retrievable rather than paraphrased to you. Compare each self-assessed level against the protocol rubric line by line. The check is simple and unforgiving: for every level claimed, ask which indicator in the protocol supports it and what evidence the filing cites. Flag any level asserted without a cited evidence item — an unsupported level is the single most common failure point in a defensible audit trail.
Third, source the external verification report or the verifier's summary of findings. This is the document that tells you whether an independent party reached the same conclusion the facility reached about itself. Where the verification report and the self-assessment diverge, the divergence itself is the finding, and it should be traced back to the specific indicator where the two readings parted ways.
| Verification item | Where to source it | What to check |
|---|---|---|
| Indicator list, all five areas | Current MAC TSM protocol PDF, relevant sector | Exact indicator wording and rubric level definitions |
| Self-assessed levels | Facility's published TSM self-assessment | Each claimed level matched to a cited evidence item |
| Independent findings | External verification report or verifier summary | Agreement or divergence with the self-assessment |
Run the three sources against each other before you cite anything. A level that appears in the self-assessment but not in the protocol rubric is not a score; a level that appears in both but lacks a cited evidence item is not verified; a level where the verification report and the self-assessment disagree is an open question, not a result. Only when all three sources align on the same indicator, the same rubric line, and the same evidence do you have something you can defend in a filing, a board memo, or a public statement.
Keep the sourcing chain intact as you work. Note the protocol version and sector you pulled from, the date of the self-assessment filing, and the identity of the verifier. If any of those three cannot be produced on request, treat the score as unverified regardless of how favorable it looks.
Audit Depth Options Compared
Three audit depths are available to anyone who wants to test a TSM score, and they differ mainly in where you spend your hours. The rubric spot-check is the cheapest: pull the current MAC TSM protocol for the relevant sector, sample two or three indicators per performance area, and score each sampled indicator against the level rubric yourself. A spot-check of that size can be finished in hours, which makes it useful when you only need a rough read before a meeting. Its weakness is structural: because you sample rather than map, a performance area that has been systematically over-claimed — where the self-assessment inflates several indicators at once rather than one — can pass through the sample untouched. Sampling catches isolated errors, not patterns.
Full indicator mapping is the opposite approach. You score every indicator in all five performance areas from primary evidence, working through the protocol document indicator by indicator rather than selecting a subset. This produces the highest confidence available to an outside reader, because no indicator is assumed to be fine. It also costs the most: days of work per facility, plus the discipline to keep every score tied to a document rather than to a summary. For most readers, full mapping is the right method only when the decision at stake justifies that cost — a transaction, a formal filing, or a public claim you will have to defend.
External-verification cross-check is the winner for most users because it concentrates effort where scores are most likely wrong. Instead of mapping everything, you map only the areas where the facility's self-assessment and the external verifier's results diverge, plus any tailings indicator regardless of whether the two sources agree. Divergence between self-assessment and verification is the strongest available signal that a score is unreliable, so spending your hours there buys more accuracy per hour than uniform sampling or uniform mapping. Tailings indicators earn a standing exception because the consequence of an error there is not a reporting problem.
The comparison below states the trade-off directly. Treat the cost column as relative effort, not a quote — the actual hours depend on how many indicators the current protocol lists for your sector, which you confirm from the protocol document itself.
| Audit depth | What you score | Relative cost | Main failure mode |
|---|---|---|---|
| Rubric spot-check | 2–3 sampled indicators per area against the level rubric | Hours | Misses a systematically over-claimed area |
| Full indicator mapping | Every indicator in all five areas, from primary evidence | Days | Highest confidence, highest cost |
| External-verification cross-check | Divergent areas plus any tailings indicator | Between the two | Depends on verifier results being available |
One rule governs all three: whichever depth you choose, verify each of the five performance-area thresholds against the current MAC TSM protocol indicator list and the self-assessed evidence behind it, never against a single headline score. The cross-check earns its place as the default because it applies that rule where it matters most, and reserves full mapping for the cases that genuinely require it.
Thresholds and Numbers That Matter
A headline score is a summary, not a finding. The defensible unit of analysis is the individual indicator, read against the current Mining Association of Canada (MAC) TSM protocol list for the relevant sector. Three numeric decision rules convert that principle into an auditable method. They are thresholds for your own verification work, not statistics published by MAC, and they should be applied before any score is cited or acted on.
Divergence trigger. Score each indicator yourself from the protocol text, then compare your level to the facility's self-assessed level. If your rubric-based read differs by one or more levels, stop and re-audit that indicator against primary evidence before accepting either number. A one-level gap is the trigger, not a rounding error: it usually means the two readers applied different criteria, different evidence, or different protocol versions. Resolve the gap at the indicator level, because averaging a disputed indicator into a performance-area total hides exactly the disagreement a reviewer needs to see.
Currency check. Confirm that the protocol version year and the assessment year match. A score produced against a superseded protocol version is not comparable to current-year scores, even when the level letters look identical. Open the protocol document and read the version date printed on it; then confirm the assessment date falls within that version's active period. If the two do not align, treat the score as historical rather than current, and say so explicitly when you cite it.
Tailings override. Treat any tailings-management indicator below the protocol's defined conformance level as a stop condition for the facility's overall read, regardless of how strong the other four performance areas score. Tailings results carry consequence-of-failure weight that a composite average cannot express, so a high total never offsets a tailings shortfall. Apply the override before you compute or quote any aggregate.
| Check | Trigger | Action |
|---|---|---|
| Divergence | Your read differs from self-assessment by one or more levels | Re-audit that indicator with primary evidence |
| Currency | Protocol version year does not match assessment year | Reclassify the score as historical, not current |
| Tailings override | Any tailings indicator below the defined conformance level | Flag the whole facility read; do not average it away |
Run the three checks in order: currency first, because a stale protocol invalidates the comparison; divergence second, because unresolved gaps undermine every downstream number; tailings override last, as the gate on the aggregate. Document each check with the evidence you used, so a second reviewer can reproduce your conclusion rather than trust it. That record, not the headline score, is what makes a TSM result defensible.
What a TSM Score Does Not Prove
A TSM score is a summary, not a proof. The number or letter you see on a public filing tells you that a facility went through the Mining Association of Canada (MAC) self-assessment and verification cycle — it does not tell you which protocol version was applied, which indicators were scored, or whether the evidence behind any single performance area would survive an independent read. That gap is where most citation errors begin.
The table below is the only place in this guide that lists the edge cases where the audit rule breaks versus still wins. Read it before you cite any score.
| Edge case | Does the rule still win? | What to do |
|---|---|---|
| Self-assessment only, no external verification yet | Yes — your independent rubric read is still the check | Label every conclusion "pre-verification" and lower your confidence accordingly |
| Sector-specific protocol variants (e.g., diamond or steelmaking-coal) | Breaks if you mix rubrics | Confirm the facility's sector protocol first; never score a facility against the wrong sector's indicator list |
| Newly acquired or newly reporting facility with a short disclosure history | Wins only if you treat the record as incomplete | Check whether prior-year filings exist under the former owner; if not, mark the trend as unestablished |
Take the first row seriously. A self-assessed score with no external verification is not worthless — it is unverified. The rule still wins because your own read of the current MAC TSM protocol indicator list against the facility's self-assessed evidence is the check. But the confidence you attach to any conclusion should drop, and every sentence you write from it should carry the "pre-verification" label until a verification report exists.
The second row is the one that quietly corrupts comparisons. Sector-specific protocols carry different indicator lists, so a facility assessed under a diamond variant and one assessed under a steelmaking-coal variant are not speaking the same language. If you score either against the other's rubric, the rule breaks — not because the facility failed, but because you applied the wrong measuring stick. Confirm the sector protocol before you read a single indicator.
The third row covers newly acquired or newly reporting facilities. A short disclosure history means the trend line is thin, not that performance is poor. Check whether prior-year filings exist under the former owner; if they do not, treat the record as unestablished rather than declining. The rule wins here only when you resist the urge to read absence of history as evidence of anything.
None of this replaces the underlying verification work. The United Nations frames SDG 6 (Clean Water and Sanitation) as an essential end in itself and critical to the 2030 Agenda as a whole — the same logic applies to your reading of a score: the summary is only as good as the evidence you independently confirm behind it.
What to do next
| Step | Action | Why it matters |
|---|---|---|
| 1 | Pull the current MAC TSM protocol and locate its indicator list and level definitions (A–AAA) for all five performance areas: indigenous/community relationships, safety and health, crisis management, tailings management, and biodiversity and water. | Every threshold you cite must trace back to the protocol's own indicators and level definitions — not to a summary or a single headline score. |
| 2 | For the facility's indigenous/community relationships area, match each self-assessed indicator against the protocol's indicator list and confirm the stated level (A–AAA) is evidenced. | This is the first of the five areas that must be verified independently before any TSM score is cited or acted on. |
| 3 | Repeat the same indicator-by-indicator check for safety and health, then for crisis management. | Each area stands on its own threshold; a strong result in one does not carry another. |
| 4 | Repeat the check for tailings management, then for biodiversity and water. | These complete the five performance areas — all five must be evidenced at the stated level. |
| 5 | For any area you cannot evidence at the stated level, downgrade it one level lower before citing or acting on the score. | This is the canonical decision rule: unverifiable claims default downward, not upward. |
| 6 | Record the per-threshold verification results alongside the headline score, and cite only the verified per-area levels. | A single headline TSM score is not sufficient evidence of facility performance; defensibility requires per-threshold verification. |
Frequently Asked Questions
Before using a TSM score in a procurement decision, which protocol document must be pulled and from where?
Pull the current protocol version directly from MAC's official website before citing any TSM score in a procurement or partnership decision.
Why is a score earned under last year's framework potentially misleading this year?
Indicator wording shifts between versions, so a score earned against last year's framework may not fully reflect this year's expectations.
What must be confirmed on the protocol document itself before its indicator list is relied on?
Confirm the version date printed on the protocol document itself before relying on any indicator list.
Are self-assessed TSM results externally checked, and how often are they submitted?
Participating facilities self-assess against published protocols and submit results to external verification on a recurring cycle.
Can a facility's public TSM score be treated as one blended number?
No, it is a set of level results spanning all five performance areas and their constituent indicators rather than a single blended number.
Which performance areas require separate indicator-level checks before any threshold is cited?
Indigenous and community relationships, safety and health, crisis management, tailings management, and biodiversity conservation each contain multiple indicators, with each indicator receiving its own independent level result.
Quick answers
| Who administers the Towards Sustainable Mining (TSM) program? | The Mining Association of Canada (MAC) administers the Towards Sustainable Mining (TSM) program. |
| How do participating facilities produce TSM results? | Participating facilities self-assess against published protocols and submit results to external verification on a recurring cycle. |
| Where should you pull the current protocol version from before citing any TSM score? | Pull the current protocol version directly from MAC's official website. |
| What are the five distinct performance areas TSM evaluates facilities across? | TSM evaluates facilities across five distinct performance areas: indigenous and community relationships, safety and health, crisis management, tailings management, and biodiversity conservation. |
| Is a facility's public TSM score a single blended number? | A facility's public score is not a single blended number — it is a set of level results spanning all five areas and their constituent indicators. |
Also worth reading: Critical Minerals Mining Cutoff Grade: USGS 2025 Bear Lodge Model Raises Threshold From 0.2% to 0.4% TREO: Critical Minerals Mining Cutoff Grade: · Unlocking India's Rare Earths: Assessing the Role of AI in Sustainable Mining: Unlocking India's Rare Earths: Assessing · Sustainable Mining Powers Our Greener Future: Sustainable Mining Powers Our Greener