A second seller that shares one processor still causes downtime when that processor stops. To assess supply chain concentration risk in manufacturing, record the confirmed production site behind each seller of one critical part, and keep the paths separate only when those sites differ.
Record the confirmed production site behind each seller of one critical part, and mark a node as shared only when two records name it. The NIST Manufacturing Extension Partnership page of 16 January 2024, retrieved 18 October 2026, says to collect supplier data, part numbers, and spend for one key product, then identify Tier 2 and Tier 3 suppliers and map routings. That page prints no sample spend. U.S. Geological Survey Open-File Report 2021-1045 states that the geometric mean of the normalized threshold scores 0.20, 0.50, and 0.64 is 0.40. That cutoff is for the report’s critical-minerals list. The part record below uses your documents. The UK Department for Business and Trade framework of 16 November 2022 puts visibility ahead of diversification, partnerships, stock, onshoring, and demand management. Further buying and selling explainers are in the Business section.
| Input or item | Typical value or source | Why it matters |
|---|---|---|
| Product to map | NIST MEP, 16 January 2024: supplier data, part numbers, and spend for one key product | The map starts from a product and a part. A seller count comes later. |
| Upstream names | Same page: Tier 2 and Tier 3 are the suppliers’ suppliers | The name on the purchase order can sit in front of the processor. |
| Routings | Same page: transportation modes. Weather and port congestion are named risks. | A second forwarder can still use one gateway. |
| Mineral-list threshold | USGS Open-File Report 2021-1045, Table 1: normalized scores 0.20, 0.50, and 0.64. Their geometric mean is 0.40. | This cutoff sorts mineral commodities in that report. The part row uses your records. |
| Single domestic producer | Same report: one domestic producer along the raw-material chain is a single point of failure | The test is a count of producers. |
| Sample spend and index weights | NIST page, retrieved 18 October 2026 | The saved page does not print them. Leave the cells blank. |
What does concentration mean for one part?
On this record, concentration means that the purchasing paths for one part depend on the same confirmed site, processor, or transport handoff.
The NIST page, written by Melissa Burant and Marc Schneider and dated 16 January 2024, says a useful map shows where you get materials, where your suppliers get their materials, and where and how you ship. It lists geographic concentrations of suppliers and their suppliers, where the most critical parts come from in number of suppliers and geography, where the most at-risk parts come from, and sourcing for the highest-volume materials. A seller’s office address is a different fact from the site that makes the part.
Pick a part whose absence stops a named operation. Ask the planner which product needs it, and write the part number and revision. NIST says some parts face outside forces, some are hard to source, and some are critical to the product. Map the one in front of you before you open the whole purchasing ledger.
Separate the seller from the manufacturer and from any processor the papers name. A second seller can be a real commercial contact and still ship from the same upstream plant. When that second name is a distributor, how small manufacturers split stock, payment, and after-sales across sales channels shows who owns the goods. For commercial IT and plant computing hardware, the IT hardware distributor evaluation checklist checks OEM franchise authorization, ISO/IEC 20243 serial provenance, and locked BOM revisions before a second distributor is counted as an independent source. The production site stays on this record.
Takeaway: Write the site that makes or processes this part. A second purchase order, by itself, is a second commercial contact.
What belongs on the part record?
Copy one block per seller of the same part, including an alternative you are considering. Keep the part number the same so the blocks can be compared. The last column stays empty until you fill it. An empty cell is an open question.
| Field | What to enter or ask for | Your record |
|---|---|---|
| Critical part | Internal part number, revision, and description | |
| Production consequence | Operation that stops if this part is missing | |
| Direct seller | Legal name, and the site that fills this order | |
| Tier 2 or Tier 3 name | Manufacturer, material source, or processor for this part | |
| Production site | The site that makes or processes this part | |
| Transport handoff | Dispatch point, warehouse, port, or other named node | |
| Spend on this part | The figure on your purchasing record for this seller and part | |
| Spend share | Calculated only after every spend cell for this part is filled | |
| Shared node | The site, processor, or handoff named on another block, with both documents | |
| Alternative approval | The drawing, material, and approval still open | |
| Evidence and date | Document, issuer, scope, and the date on it | |
| Owner and next date | The person who will close one blank cell, and the review date |
Mark a site confirmed only when the document names this part. If a supplier says it uses several factories and does not name the one that makes your revision, the site cell stays blank. File the reply next to the record.
How do you calculate the mineral-list score and your own share?
The mineral-list score and the part-level share answer different questions. Open-File Report 2021-1045, by Nedal T. Nassar and Steven M. Fortier, calculates supply risk for a mineral commodity on a scale from 0 to 1 as the geometric mean of three normalized components: disruption potential, trade exposure, and economic vulnerability.
Supply risk = (DP × TE × EV) ^ (1/3)
Disruption potential starts as the sum, across producing countries other than the United States, of each country’s share of global production squared, then multiplied by the larger of that country’s ability-to-supply index and willingness-to-supply index. The report normalizes that raw sum to a 0-to-1 scale with the minimum and maximum raw scores in the study. The ability-to-supply index is based on the Fraser Institute’s policy perception index. The willingness-to-supply index uses trade, ideological, and defense ties with the United States. The saved report does not print a sample country’s share or either index on a single arithmetic row. Those cells stay blank, and the raw disruption potential stays blank with them.
Trade exposure is U.S. net import reliance. The report limits it from 0, when the United States was a net exporter, to 1, when consumption came entirely from foreign sources. For the whole window, arsenic, gallium, indium, iridium, manganese, niobium, ruthenium, strontium, tantalum, and yttrium scored 1. Helium, gold, and iron ore scored 0. The study evaluates years 2007 through 2018, and 2018 is the latest year with the data the report required. The report prints those scores of 0 and 1. A filled division for one commodity’s imports, exports, and consumption stays blank on this page.
Economic vulnerability is higher when spending on the commodity is large in industries with low operating profits that account for a large share of U.S. GDP. The report normalizes the raw score to the same 0-to-1 scale. It does not print one industry’s expenditure and profit as a sample product. The economic-vulnerability cell for a factory part stays blank.
Table 1 of the report sets a threshold for each component. Disruption potential: global production outside the United States is concentrated so that one-half comes from a single country less able or less willing to supply the United States than the average country, defined as the 75th percentile of the two indexes, or an equivalent distribution. The normalized score for that case is 0.20. Trade exposure: one-half of U.S. consumption comes from foreign sources. The normalized score is 0.50. Economic vulnerability: annual expenditure equals the median commodity expenditure, in a manufacturing industry with a below-average operating-profit ratio, defined as the 75th percentile. The normalized score is 0.64.
Worked example. Multiply only those three printed scores.
0.20 × 0.50 = 0.10
0.10 × 0.64 = 0.064
0.40 × 0.40 × 0.40 = 0.064
The cube root of 0.064 is 0.40. The report states that this geometric mean is the supply-risk threshold. A commodity whose recency-weighted mean supply risk is greater than or equal to 0.40 was recommended on the quantitative test. The year weights are 40 percent for 2018, 30 percent for 2017, 20 percent for 2016, and 10 percent for 2015. Of 54 commodities with enough data, 36 met the 0.40 test. The report does not print four yearly scores for one commodity in the prose used here, so a weighted-mean cell for a named mineral stays blank until you copy those scores from the report.
China was the leading producer, at one or more stages, for 35 of the 54 commodities, including 26 of the 36 at or above 0.40. The count covers mineral commodities in that report, for production through 2018.
The 24 February 2022 Federal Register notice adopts this methodology for the 2022 final list and says the list is reviewed at least every three years. The notice describes a quantitative supply-risk evaluation, a single-point-of-failure evaluation, and a qualitative evaluation when the other two are unavailable. It points to the open-file report for the calculations.
When this does not apply. The 0.40 cutoff sorts mineral commodities in that study. Put it on a mineral row. A bracket, a motor, or any other purchased part uses the two lines below.
NIST says to collect spend and to show critical parts by number of suppliers and geography. It prints no sample spend, so the share uses your figures only.
Spend share = seller spend / part spend
Worked example. If the seller-spend cell or the part-spend cell is blank, the share stays blank. There is no number to divide.
Confirmed site count = the number of different production-site names filled in for this part.
Rule of thumb: If confirmed production sites < 2, assign an owner to the blank site cell before you treat another seller as a separate source. In the USGS report, a mineral commodity with a single domestic producer along its raw-material chain is a single point of failure. The report names beryllium, nickel, and zirconium on that basis. The test counts producers. Price and lead time stay off this line.
Shared-node count = the number of records that name the same confirmed processor.
Worked example. Two records that name the same processor have a shared-node count of 2. The 2 counts those records. If the second processor cell is blank, the shared-node count stays blank.
Common mistake: Copying 0.40, or a mineral’s trade-exposure score of 0 or 1, onto a component row.
Takeaway: Calculate 0.40 from the three printed threshold scores. Calculate a part’s spend share only from spend cells that your file already shows.
Where each kind of record fails
Each shortcut fails in a different place. The fix is a named document, or a cell left blank.
| Record | Dominant failure mode | Prevention |
|---|---|---|
| Seller count | Two purchase orders stop together because they share one processor | Write the confirmed site on each record before calling the sellers separate |
| Spend share | A low-spend part is the one that finishes the job, so the line stops while the share looks small | Write the operation that stops, and leave the share blank until every spend cell for that part is filled |
| Mineral-list score | The 0.40 mineral cutoff is copied onto a purchased part | Keep 0.40 on the mineral row. Fill the part row from your records |
| Nearby addresses | Two plants on one road are treated as one shared site, or two distant offices hide one plant | Confirm the production site. How manufacturing clusters develop separates a shared street from a documented supplier link |
| Route | A second forwarder is treated as a second port | Name the gateway. What a port-performance index leaves out of a door-to-door move is the check before a port score stands in for your handoff |
Compare replies at the same level of detail. A factory name on one record and a country on the other is an open comparison. Ask for the missing name. NIST says you can ask a supplier for its suppliers and receive only a name, and that routings can show extra movements plus risks from weather or port congestion.
The report also describes a domestic single point of failure in words a buyer can reuse as a question: if one producer along the chain stops, the material has no second domestic producer in that study. For your part, the same question is whether a second confirmed site is already on the record.
Which next step matches the dependency you found?
Choose the action that matches the node you confirmed, or the evidence you still lack. The Department for Business and Trade framework, published 16 November 2022, says data and visibility come first, because a clear picture of the chain shows where the risks are and which actions to consider. The five areas are diversification (an alternate source of supply), international partnerships, stockpiling and surge capacity, onshoring, and demand management (substitutes, alternatives, innovation, and circularity). Test the area that matches the node on the record.
| Finding in your record | Practical next step |
|---|---|
| Production site or processor blank | Send a part-specific request and name the person who follows it |
| Same processor on two records | Look for a second processing path and the approvals it still needs |
| Same transport handoff on two records | Ask logistics whether another handoff is already arranged |
| Alternative named, checks open | List each approval before anyone treats the alternative as ready |
| Node confirmed, and no second source is available now | Ask planning for the operation that stops. Keep the exposure on the record |
The hidden cost of a shared node is the production time that stops, the approval work still open, and any stock decision you have not quantified. Wage rates, part prices, and lead times come from the plant’s own purchasing records.
Blocked hours = hours the planner writes for the operation that stops
Worked example. If the planner’s hours cell is blank, blocked hours stay blank. With that cell empty, there is no hours figure to carry forward.
| Cost line | What to write | Document | Amount |
|---|---|---|---|
| Time the shared node blocks | The operation that stops | Planner’s note | |
| Stock or surge capacity | Whether holding stock or a contract surge is under review | The clause, if a contract already has one | |
| Alternate-source approvals | Each check still open | Quality record | |
| Charges on an equipment offer | Quoted, extra, and unknown lines | The offer |
Leave the amount column blank until the document prints the figure. Stock quantity and lead time stay blank on this page. Money on a machine offer belongs on the industrial equipment import-cost worksheet, which keeps each charge included, separate, or unknown.
Finish with one request, one recipient, and one owner. Set the review date to the open question. NIST says the mapping program then moves into a regular risk review. The same page says the CIRAS program includes a weekly one-hour session on one topic. That hour is the length of the program’s class. Put a replenishment time on the planner’s note when the planner has written one.
Takeaway: Close the record with a second confirmed site, or name an owner for the blank site cell.
How this page was checked
The mapping steps, the Tier 2 and Tier 3 definition, the instruction to collect part numbers and spend, the routing and port-congestion risks, and the weekly one-hour CIRAS session come from NIST’s “Mapping Your Supply Chains Helps Prioritize Risks, Actions,” dated 16 January 2024 and retrieved 18 October 2026. The geometric mean, the Table 1 scores 0.20, 0.50, and 0.64, the 0.40 threshold, the year weights, the 54 and 36 commodity counts, the China counts of 35 and 26, the trade-exposure scores of 0 and 1, and the single-domestic-producer test for beryllium, nickel, and zirconium come from U.S. Geological Survey Open-File Report 2021-1045, retrieved the same day. The 24 February 2022 Federal Register notice is the adoption of that methodology for the 2022 final list, including the review at least every three years. The five areas and the visibility statement come from the Department for Business and Trade framework published 16 November 2022, retrieved 18 October 2026.
The blank part record, the spend-share line, and the blocked-hours line are this page’s organization of those sources. A cell stays blank when the saved page, or your own document, does not print the figure.
