This is Sunday Supply Chain Stories, where we revisit the foundations that continue to shape how inventory moves, returns and recovers value.
If you enjoy the newsletter, follow The Supply Loop on LinkedIn and ask your friends to subscribe to receive the next story.
In the spring of 1943, Allied forces were fighting to drive Germany and Italy out of North Africa, one of the central theaters of the Second World War. Both powers had held much of the region since 1941, threatening the Suez Canal and Allied access to Middle Eastern oil. The campaign to dislodge them had run for nearly two years at significant cost in men and equipment. By that spring the end was in sight: Operation Torch had landed Allied forces in Morocco and Algeria in November 1942, opening a western front; Montgomery's Eighth Army was pressing from Egypt in the east; and Rommel's Afrika Korps, caught between both forces, was being squeezed toward a surrender that would yield roughly 275,000 prisoners in May. But the campaign was still running, aircraft were still going down faster than supply lines could replace them, and no wreck on friendly soil could be written off as a loss. It had to be treated as a resource.
Telergma was one of the forward bases sustaining that effort. Seized by American forces in December 1942, weeks after Torch, the airfield ran two operations simultaneously by spring: the 17th Bombardment Group flew B-26 Marauders on interdiction missions against Axis supply lines across Tunisia, and when those aircraft came back damaged, or did not come back at all, a salvage operation took over. Any aircraft that went down within roughly 200 miles was tracked, loaded onto a flatbed, and driven in. The bombing operation and the recovery operation were sharing the same ground.
The ramp at Telergma held multiple aircraft types simultaneously — B-26 Marauders, B-25 Mitchells, P-38 Lightnings — and components between those airframes were not interchangeable in any meaningful way: a Marauder's landing gear assembly, cockpit instruments, and hydraulic fittings were designed for that airframe and served no useful purpose pulled from a B-25. Salvage crews were forced to sort wrecked aircraft by type in order to facilitate downstream decisions.
The extraction of the components itself required knowledge that was specific to aircraft systems and did not transfer from general mechanical work. Cockpit instruments were one example. The luminous dials on altimeters, airspeed indicators, and artificial horizons used radium-226 paint so pilots could read them in darkness at altitude, and a cracked dial face during removal released radon gas and radioactive particulate into a workspace without sealed ventilation. Instruments came out intact and undisassembled, with visibly damaged units segregated for separate handling. Gyroscopic flight instruments required caging before extraction: the internal gimbal locked in a neutral position before the unit was physically disconnected, because a gyro that was not caged first had its precision bearing assembly destroyed by the handling and transport that followed. Hydraulic lines were capped at the point of disconnection to prevent fluid from pooling across a yard that was simultaneously processing aircraft with residual fuel in their systems. From there, parts moved through a formal reclamation process — inspected, approved, then routed into the theater supply system or forwarded directly to the nearest repair shop where they were needed.
What could not be recovered as a usable part was smelted. The aluminum from a destroyed airframe had residual value as raw material, and the system was built for recovery at every level. Nothing corroded in the desert while waiting for a better outcome.
The operational logic behind Telergma was driven by a constraint that shaped nearly every decision in the WWII theater: shipping capacity across the Atlantic was finite and contested. Liberty ships made the crossing in weeks, and the total tonnage available for any given supply category competed with everything else the theater needed — fuel, food, artillery ammunition, replacement troops. A spare part recovered from a wrecked aircraft in-theater was a part that did not have to make that crossing. In practical terms, the recovery value of a serviceable landing gear assembly or an intact instrument cluster could be measured not just in the part's replacement cost but in its shipping weight, multiplied by the scarcity of the bottleneck it was competing through.
Telergma was part of a larger system operating on the same principles across the European and Mediterranean theaters. RAF Burtonwood in Lancashire, transferred to U.S. control in June 1942, grew into the largest American air depot in Europe, processing more than 11,500 aircraft and overhauling over 30,000 engines between 1943 and 1945, with a workforce that reached 18,500 at peak. Aircraft arrived from operational squadrons and were evaluated systematically for what could be returned to service, what could be salvaged for parts, and what could only be reclaimed as raw material. Pilots who survived crash landings were instructed, where possible, to put their aircraft down near friendly airfields. The difference between a crash-landed aircraft and a crashed aircraft, from a supply standpoint, was the difference between a recoverable parts source and a lost asset.
The mechanics at Telergma were operating on an insight that is easy to state but surprisingly hard to hold onto in practice: the unit boundary of an aircraft and the value boundary of its components are not the same thing. An aircraft has a tail number, a maintenance record, and a squadron assignment. It also contains dozens of independent systems, each with its own condition, each with its own recovery potential — and the failure of one system tells you almost nothing about the serviceability of the others.
A B-17 with a destroyed wing and a burned-out engine still had intact hydraulic actuators, serviceable radios, functional instruments, and a usable landing gear assembly. As a unit, it was a total loss. As a source of recoverable components, it was a supply event with a defined value that could be assessed, graded, and routed through the system.
This distinction changed the decision rule. Evaluated as a unit, a grounded aircraft sorted into flyable or unflyable, and the unflyable ones became scrap. Evaluated as a bundle of separable components, you introduced a middle category: aircraft that were genuinely beyond repair but still had parts worth recovering, in specific quantities, for specific downstream uses. The mechanics at Telergma worked in that middle category every day. The aircraft was not recoverable, but many of its contents were. The practice had a name that described exactly what it was: cannibalization — stripping one aircraft to keep others flying.
The system also required that mechanics develop a clear model of what was worth pulling — a practical classification accounting for the condition of each component, its compatibility with types still operational in the theater, and the downstream demand that made recovery worthwhile rather than merely possible. A landing gear recovered from a type that had been phased out of the theater was not a supply asset; it was storage overhead. The value recovered was always conditional on knowing what the receiving end of the system actually needed.
Throughput discipline mattered as much as component knowledge. A grounded aircraft waiting to be processed was not a neutral event. Storage was scarce, personnel were occupied, and holding cost was real and accumulating. Moving material through the recovery process quickly was not a secondary concern; it was part of what made the operation viable at all.
Over the decades, the language evolved: "cannibalization" became "parts harvesting," which in the ITAD industry became "component recovery."
Today, the electronics refurbishment industry faces a version of the constraint that shaped Telergma, expressed through a different scarcity. AI infrastructure buildout has driven component shortages now working through the secondary market in specific and measurable ways. Passive component suppliers including KEMET, Panasonic, and AVX announced multiple rounds of price increases in late 2025 and early 2026 on tantalum capacitors, citing stronger orders from high-end computing and networking customers, with increases in the 15–30% range for selected server-grade lines. AI servers require roughly 10 to 15 times more multi-layer ceramic capacitors than general-purpose servers, which has created sustained pressure on MLCC supply at the high-performance end of the product mix. The boards being retired from data centers and enterprise networks today — equipment built between 2020 and 2024, when these components were cheap and plentiful — carry exactly the component densities that the repair and broker market now needs.
The consumer side of electronics recovery has known this for years, even if it hasn't always been systematized. An iPhone screen resells for $40 or more — shredded, it is worth essentially nothing. A DDR4 memory stick in working condition can be resold for five times its scrap value. A 1TB hard drive in good condition commands more than forty times its scrap equivalent. That gap exists right now, across millions of devices processed annually.
What the Telergma system had that many modern electronics operations lack is accountability and traceability. Parts pulled from wrecked aircraft went through a formal reclamation process — inspection, approval, tracked routing from source aircraft to repair shop. That traceability existed as operational necessity: the crew chief installing a salvaged landing gear assembly needed to know it had been inspected and cleared, not simply pulled and stacked. Modern component harvesting operations vary enormously in how rigorously they manage this chain. The ones that treat traceability as incidental to throughput tend to discover that ungraded, untracked harvested components create quality problems downstream that erode the margin the harvesting was supposed to generate. The ones that build grading and traceability infrastructure first — before scaling volume — recover more value per unit processed and do so without the rework that follows from skipping that step.
There is a second constraint that the Telergma system did not face but modern component recovery operations increasingly do: the availability of workers who can actually execute the work accurately. The mechanic removing a gyroscopic instrument from a wrecked aircraft in 1943 had been trained on that system. He understood its function, knew how to handle it safely, and could make a reliable judgment about whether it was serviceable. That knowledge existed because the manufacturing base that produced those instruments and maintained those aircraft was domestic and largely intact. The workers were embedded in the same industrial system they were recovering from. Decades of manufacturing outsourcing have changed that. Accurately grading a circuit board, testing a capacitor bank, or assessing the integrity of a display assembly requires engineers and electronics technicians with hands-on systems knowledge, and that workforce is considerably thinner than most recovery operations anticipate when they design their processes. The capacity to recover value at the component level, rather than routing everything to bulk scrap, depends on having people who can evaluate what they are handling without destroying it. In 1943, that workforce was embedded in a domestic manufacturing base. Today it is not, and operations that treat this as a routine staffing gap rather than a structural constraint tend to underestimate how much margin it is quietly consuming.
The broader dynamics reinforcing this shift are not short-term. Tantalum is the metal that gives capacitors in servers and telecommunications equipment their density and thermal stability, and it is extracted primarily from coltan ore mined in central Africa. The Democratic Republic of Congo alone accounts for roughly 40% of global production, and the Great Lakes region as a whole accounts for well over half. That concentration creates a supply chain that is fragile in ways that go beyond normal commodity risk. In April 2024, M23 rebel forces took control of Rubaya, a mining area in North Kivu that the United Nations estimates accounts for more than 15% of global tantalum supply. In January 2026, a mine collapse at Rubaya killed more than 200 people and caused an immediate spike in prices across the region. Conflict minerals concerns and the documented smuggling of DRC coltan through Rwanda have put the integrity of the broader African supply chain under growing scrutiny, at precisely the moment when demand from AI infrastructure is rising fastest. MLCC manufacturing capacity takes years to expand regardless of raw material availability, and the AI infrastructure cycle is early enough that its appetite for components is likely to grow before it plateaus. The hardware being retired from data centers today carries the components the market is short on right now, sourced and installed when none of this constraint existed. The window for recovery that is genuinely valuable is not permanent.
The mechanics at Telergma were responding to a clear-eyed read of where the bottleneck was and what assets they had on hand to address it — without theory, without abstraction. That read is available again, for operations that choose to make it.
For practitioners:
Where in your current processing workflow is the disposition decision made at the device level when it should be made at the component level, and what would a structured component-grade assessment reveal about the parts surviving in your current scrap stream? For component categories now under supply pressure — tantalum capacitors, MLCCs, high-layer boards, functioning display assemblies — do you have the incoming hardware data to know which product families in your inbound flow are likely to contain them? When your operation harvests components, is there a traceability chain from the source device through inspection, grading, and routing to the receiving end, or does the part leave the source device and enter the supply chain without that record? And how does your cost-to-serve for component-level recovery compare to the margin differential between current scrap pricing and secondary market pricing for those specific components?
Sources
Joris Nieuwint. "Scrapping and Salvaging World War II Aircraft in North Africa." War History Online. October 15, 2015. https://www.warhistoryonline.com/featured/scrapping-and-salvaging-world-war-ii-aircraft-in-north-africa.html
"17th Bombardment Group, Martin B-26 Marauder." B26.com. Accessed June 2026. https://www.b26.com/page/17th_bombardment_group.htm
"B-26 Marauders in Action over North Africa." Key.Aero. Accessed June 2026. https://www.key.aero/article/b-26-marauders-action-over-north-africa
"Radioactive Materials in Flight Instruments." The Aviationist. Accessed June 2026. https://theaviationist.com/special-reports/radioactive-materials-in-flight-instruments/
"Burtonwood." American Air Museum in Britain, Imperial War Museum. Accessed June 2026. https://www.americanairmuseum.com/place/69
The Army Air Forces in World War II, Volume VI: Men and Planes, Chapter 11. Historical Division, Department of the Air Force. https://www.ibiblio.org/hyperwar/AAF/VI/AAF-VI-11.html
"What Were the Challenges and Techniques in Maintaining World War 2 Airplanes?" World War 2 Planes. October 2024. https://world-war-2-planes.com/what-were-the-challenges-and-techniques-in-maintaining-world-war-2-airplanes/
David Daoud. "Hardware demand puts new focus on parts harvesting." Resource Recycling / E-Scrap News. June 5, 2026. https://resource-recycling.com/e-scrap/2026/06/05/hardware-demand-puts-new-focus-on-parts-harvesting/
Arun Karottu and Alex Cummings. "In Our Opinion: Maximizing value with component harvesting." Resource Recycling. November 7, 2019. https://resource-recycling.com/analysis/opinion/2019/11/07/in-our-opinion-maximizing-value-with-component-harvesting/
Dimension Market Research. "Refurbished Smartphone Market Size to Reach USD 154.6 Bn by 2034." 2024. https://dimensionmarketresearch.com/report/refurbished-smartphone-market/
Argus Media. "How the Rubaya Mine Collapse Impacts Global Tantalum Supply." 2026. https://www.argusmedia.com/en/news-and-insights/market-opinion-and-analysis-blog/rubaya-mine-collapse-tantalum-supply-chain
African Ores. "Risks and Volatility in the Tantalum Market: Supply, Pricing, and Trade Disruptions." 2025. https://africanores.com/blog/risks-and-volatility-in-the-tantalum-market-supply-pricing-and-trade-disruptions/
Discovery Alert. "M23-Controlled Tantalum Deposit: Strategic Mining Risks." 2026. https://discoveryalert.com.au/strategic-metal-vulnerabilities-tantalum-conflict-2026/
Thank you for reading Sunday Supply Chain Stories!