Industries / Industrial equipment and services / Industrial machinery / Wire and cable
Insulating
Insulating covers each conductor in a layer that keeps the current in, today mostly by extrusion, pushing melted plastic through a die around the moving wire, and for most power cable by curing that plastic so it holds its shape when hot.
7
Milestones achieved, from 1846 to 1999.
3
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1846
Gum pressed on without seams
Werner Siemens built a screw press that covered wire with gutta-percha, a rubber-like gum softened by heat, in one layer with no seam. His obituaries of 1892 and 1893, reproduced by Grace's Guide, date it to 1846 and say its principle was then in use "in most cable works"; Siemens's own history dates the press to 1847.
Credit
Werner Siemens
1890
Paper insulation at 10,000 volts
Sebastian Ziani de Ferranti's Deptford power station sent 10,000 volts into London through underground mains. Cables "made according to the best practice of the day were continually breaking down", until his design of concentric copper tubes separated by paper soaked in ozokerite wax worked. The Science Museum holds a sample as "the first paper-insulated 10,000-volt electricity supply cable", in use from 1890 to 1933.
Lessonat a new voltage, the insulation is the first thing to fail.
Credit
Sebastian Ziani de Ferranti, for the London Electric Supply Corporation
1924
Oil in the paper
Luigi Emanueli at Pirelli developed cable whose paper insulation is filled with fluid oil. Pirelli's foundation says it was tested at the Brugherio plant in 1924 at 132,000 volts and lit New York and Chicago from 1927; an obituary in The Engineer has the first oil-filled cable in service shortly after Emanueli became chief engineer in 1917.
Credit
Luigi Emanueli, Pirelli
1939
Polyethylene goes into cable
Reginald Gibson and Eric Fawcett rediscovered polyethylene at ICI in England in 1933, and almost all of the 4,000 tonnes made between 1939 and 1945 went to insulate radar cables, The Chemical Engineer reports. By 1953 the Bell System was extruding it onto multi-pair telephone cable in field trials, starting with about four miles of 26-pair cable at Cooperstown, New York.
Credit
Reginald Gibson and Eric Fawcett, ICI; Bell System
1963
Polyethylene that holds its shape
Frank Precopio and Alfred Gilbert of General Electric patented polyethylene cross-linked with a peroxide, filed in 1955 and granted in 1963. Cross-linking ties the long molecules together so the plastic does not soften and flow when the cable runs hot. Cross-linked polyethylene, XLPE, became the usual insulation: most US medium-voltage cable uses it, an EPRI review said in 2000.
Credit
Frank M. Precopio and Alfred R. Gilbert, General Electric
1974
Water trees in buried cable
Moisture grew branching damage, called water trees, in the insulation of buried cable: high-molecular-weight polyethylene cable of 1971 had a very poor service record, and early XLPE cable without a jacket failed early too. EPRI's review credits Bahder and co-workers with bringing the problem to North American attention in 1974. The fixes included cleaner compounds, putting the insulation and both its shield layers on together ("true triple extrusion"), tree-resistant XLPE and jackets.
Lessoninsulation that passes its tests when new can still fail after years in wet ground.
Credit
Bahder and co-workers; review by B. Bernstein for EPRI
1999
Plastic cable for direct current
ABB put the first commercial extruded-plastic DC cable link into service on the Swedish island of Gotland in 1999, two 70 km cables at 80 kV. Earlier plastic DC cables had broken down because electric charge built up inside the insulation. By 2019 more than thirty such projects, over 10,000 km, were installed on four continents, Borealis engineers counted. Sources differ on the date: operation in June 1999 (Hitachi Energy) or an inauguration on 19 November 1999 (Borealis).
Lessondirect current needs insulation designed against trapped charge.
Credit
ABB, now Hitachi Energy
Watch for
Hotter-running DC cable
NKT says it has qualified a 525 kV extruded XLPE subsea DC cable system for continuous operation with the conductor at up to 90 °C, using a new semiconducting shield compound from Borouge. A 2025 review by Giovanni Mazzanti of the University of Bologna counts the 400 kV Nemo Link as the highest-voltage extruded DC cable in service.
Achieved whena link runs at 525 kV and 90 °C in service, or a third party publishes witnessed test results.
Claimed by
NKT, August 2026
Compounds rated for 90 °C
Dow says its ENDURANCE insulation and shield compounds are qualified for 525 kV DC and 90 °C continuous operation, against the usual 70 °C, in tests on full-size subsea cable.
Achieved whena named cable maker's system using these compounds is in service on a commissioned link.
Claimed by
Dow, March 2026, as summarized by ChemXplore
XLPE scrap back to oil
SCGC and the Trade Association of Thai Cable Manufacturers say they are piloting a chemical process that turns used XLPE insulation, which cannot be remelted, into naphtha, a feedstock for new plastics. They say more than 2,000 tonnes of XLPE wire go to landfill in Thailand each year.
Achieved whenthe process runs at commercial scale with tonnages reported by someone outside the project.
Claimed by
SCGC and the Trade Association of Thai Cable Manufacturers, October 2026
Open
01
One table of DC qualifications
A public table of every qualification of an extruded DC cable system, one row each: maker, voltage, conductor temperature, land or subsea, insulation and shield compounds, test standard, laboratory and witness, date, and whether a link now runs on it, each row linked to its source.
The briefCredit · unclaimed02
An open curing-tube model
Open-source software that takes a cable design and the settings of a continuous vulcanization line, the long heated tube where XLPE is cured after extrusion, and predicts how far the insulation has cured and how much methane and other by-product gas is left in it.
The briefCredit · unclaimed
Open 01 · the brief
One table of DC qualifications
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public table of every qualification of an extruded DC cable system, one row each: maker, voltage, conductor temperature, land or subsea, insulation and shield compounds, test standard, laboratory and witness, date, and whether a link now runs on it, each row linked to its source.
Why it is still open
Qualifications are announced one maker at a time. Prysmian describes systems "fully qualified up to 525 kV", NKT announced its 90 °C system in August 2026, and Mazzanti's 2025 review notes that systems at 525, 600 and even 640 kV have been qualified without naming them.
Built with what exists
The press archives of cable makers and compound suppliers, CIGRE and Jicable conference papers, transmission operators' project pages, and a small model for pulling fields out of documents (Note "The jobs agents repeat").
First step
List the qualifications announced by NKT, Prysmian, Nexans and LS Cable since 2017 and agree the columns.
Done looks like
Every announcement found since 1999 in one table, each row linked to its source, every "in service" row matched to the operator's own page, and a published count of rows confirmed by a source outside the maker.
Open 02 · the brief
An open curing-tube model
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
Open-source software that takes a cable design and the settings of a continuous vulcanization line, the long heated tube where XLPE is cured after extrusion, and predicts how far the insulation has cured and how much methane and other by-product gas is left in it.
Why it is still open
Published models exist only as papers. A 2021 model of a whole line by KAUST and Dow, for one, gives no code and does not compare its results with measurements from a working line.
Built with what exists
Python with an open finite-element library such as FEniCS, the heat and diffusion equations and constants given in published papers, and a coding agent to write and test it.
First step
Re-implement the 2021 paper's diffusion model and reproduce its figures.
Done looks like
Code under an open licence that reproduces the 2021 paper's results and predicts the methane content measured on at least five cables from a real line to within 20%.