Part One: Products and Measurement Bases
Chapter 1 The Five Product Categories of Wire and Cable and Their Boundaries
On the industry basis, wire and cable products are divided into five major categories. According to industry reporting by a Jiangsu media outlet, wire and cable nationwide "are divided into five major categories of products, nearly 2000 varieties and more than 150,000 specifications" (industry report). The five categories are power cable, electrical equipment cable, bare wire, communication cable and magnet wire (winding wire) (industry report).
The two figures, "nearly 2000 varieties" and "more than 150,000 specifications", show how highly dispersed the products are. Varieties are divided by use and structure; specifications are combinations of conductor cross-section, number of cores, voltage, insulation material and sheath material, and the number of combinations far exceeds the number of varieties. Buyers place orders by specification, factories schedule production by specification, and every change of specification requires a change of die, an adjustment of process parameters and a first-article inspection. Economies of scale are thereby weakened, which is one structural reason why industry concentration has long been low; Chapter 34 returns to this point when it discusses the shape of competition.
The basis for this division is use and structure; material and voltage are not on this classification axis. A copper-core polyvinyl chloride (PVC) insulated wire belongs, by use, to electrical equipment cable; by voltage, to 450/750V and below; and by material, to a copper conductor with PVC insulation. The three coordinates each hold independently and cannot be derived from one another. When procurement treats "copper core", "450/750V" and "building wire" as synonyms, line items on the quotation will fail to match.
Power cable carries the transmission and distribution of electrical energy; it is characterised by a large conductor cross-section, an insulation and screen structure, and a division of the standard into parts by voltage. The national standard GB/T 12706 series covers extruded-insulation power cable from 1kV to 35kV, with Part 1 corresponding to 1kV and 3kV (National Standards Full-Text Public System), Part 2 to 6kV to 30kV, and Part 3 to 35kV (CCGF 708.1—2015). The product standard for 66kV and 110kV is GB/T 11017.2-2024, Power cables with cross-linked polyethylene insulation for rated voltage 66kV (Um=72.5kV) and 110kV (Um=126kV) and their accessories — Part 2: Cables; the test methods and requirements are given in Part 1 of the same series, GB/T 11017.1-2024; and 220kV is governed by the GB/T 18890 series, as detailed in Chapter 2. Cable accessories fall outside the five product categories yet are published under the same standard as the cable: the title of GB/T 11017.2-2024 includes "power cables and their accessories", and accessories include terminations and joints, which are often priced together with the cable at procurement. Rolling the accessory amount into the cable unit price destroys the comparability of quotations from different suppliers.
Electrical equipment cable is the connecting wire inside equipment and between equipment, mostly at 450/750V and below. PVC-insulated products follow the GB/T 5023 series, and rubber-insulated products follow GB/T 5013.1-2008 (National Standards Full-Text Public System). Building wire for home renovation belongs to this category. It is not included in the statistics for power cable because the voltage ranges covered by the standards do not overlap: the upper limit for building wire is 450/750V, and the lower limit for power cable is 1kV.
Bare wire has no insulation layer or sheath; it is delivered as stranded wire, profile wire or busbar. Its boundary with power cable is the presence or absence of insulation, not the voltage. Both are used on overhead lines, yet bare steel-reinforced aluminium stranded wire is classed as bare wire, while overhead insulated cable is listed separately. The production licence treats "overhead stranded wire" and "overhead insulated cable" as two different product units for certification (Fangchenggang Municipal Administration for Market Regulation), a division consistent with the direction of the product classification.
Magnet wire (winding wire) is the conductor wound into the coils of motors, transformers and reactors; its common forms are enamelled round wire and enamelled flat wire. Its position in the value chain differs from that of the other four categories: downstream buyers purchase it as a manufacturing material, and once wound into a winding it no longer exists independently in wire form. Comparing magnet wire with cabled products in the same unit-price table is meaningless.
Communication cable refers to copper-core cable that carries voice and low-speed data. The current standard GB/T 13849-2025 covers communication cable with polyolefin insulation, a moisture barrier and a polyolefin sheath at 1MHz and below; it is effective 1 May 2026 and replaces GB/T 13849.1-2013 in full (National Standards Full-Text Public System). Optical cable transmits through optical fibre and has an independent standards system; GB/T 7424.3-2025 is the sectional specification for outdoor optical cable, so it is discussed separately outside the five categories, in Chapter 3.
On the 2024 basis of a white paper jointly issued by a consulting firm and an association, the market shares of the five categories are power cable 39.5%, electrical equipment cable 27.5%, bare wire 14.8%, magnet wire 9.9% and communication cable 8.4% (white paper report). The five add up to 100.1%, the difference arising from rounding in the original. The same material gives a 2024 market size of RMB 1.35 trillion (1.35 万亿元); converted by share, power cable is about RMB 533.3 billion and electrical equipment cable about RMB 371.3 billion. These two amounts are estimates obtained by the Institute, multiplying the two figures in the original; the original does not state them directly.
These shares can serve only as a structural reference for a single year. Another edition of the white paper gives a 2023 power cable share of 39.9% (seen only in a search abstract; the original text was not obtained), 0.4 percentage points from the 39.5% of 2024. Two years, two figures: the breakdown statistics themselves fluctuate, and the decimal places do not constitute a stable structure.
The first use of product classification is to translate a quotation into a comparable basis. The same question, "how much is copper cable per metre", may refer to 450/750V building wire or to 1kV four-core power cable, and the two differ greatly in conductor content and insulation structure. The second use is to determine regulatory status: building wire is inside the China Compulsory Certification (CCC) catalogue, control cable is inside the five product units of the industrial production licence, and power cable and optical cable each have their own standards and access rules. Lumping them together means missing certifications that should be obtained and writing tests into a contract that should not be there. The third use is to determine test items: the five categories have different test-method standards, and a single test report cannot be used across categories.
Chapter 2 The Voltage-Class System for Power Cable
In the national supervision sampling inspection of wire and cable, the product scope for the power cable category is stated concretely: "extruded-insulation power cable with rated voltages from 1kV to 35kV", and this scope is divided by voltage into three tiers: rated voltages 1kV and 3kV, rated voltages 6kV to 30kV, and rated voltage 35kV (CCGF 708.1—2015, an older system, cited for reference; the current basis is the 2022 edition of the Rules for the Implementation of National Supervision Sampling Inspection of Wire and Cable Product Quality (《电线电缆产品质量国家监督抽查实施细则》), issued by the State Administration for Market Regulation (SAMR)). This is the tiering text that can be found and is stated in writing by the regulator, and it also delimits the product scope covered by that specification.
The three tiers correspond to three sets of product standards. 1kV and 3kV follow GB/T 12706.1-2020 (National Standards Full-Text Public System); 6kV to 30kV follow GB/T 12706.2-2020; and 35kV follows GB/T 12706.3 (CCGF 708.1—2015, Table 3). The three parts belong to one series, split by voltage, with the standard-number suffix corresponding one-to-one to the voltage tier.
Beneath the product standards lies a layer of test-method standards. The inspection-item table of CCGF 708.1—2015 cites method standards including the GB/T 2951 series (mechanical and physical properties of insulation and sheath), GB/T 3048.12 (partial discharge), GB/T 3048.4 (DC resistance of conductor) and GB/T 3956 (conductors of cables) (CCGF 708.1—2015, Table 3). The three-layer structure is: the product standard sets the requirements, the method standard sets how to measure, and the conductor standard sets the material and resistance limits. When a procurement contract states only the product standard number, the method standards are brought in automatically by the normative references of the product standard and need not be listed separately.
Standard titles state two voltages: the rated voltage and the highest voltage for equipment, Um. 1kV cable corresponds to Um=1.2kV (draft standard for 1kV to 35kV aluminium-alloy cable), and 220kV cable corresponds to Um=252kV (draft 220kV standard). The rated voltage marks the nominal value of the system, while Um is the highest voltage at which the equipment is permitted to operate continuously; both figures must be checked in selection, and neither substitutes for the other.
Standards at 66kV and above stand separately. GB/T 11017.2-2024, Power cables with cross-linked polyethylene insulation for rated voltage 66kV (Um=72.5kV) and 110kV (Um=126kV) and their accessories — Part 2: Cables, is effective 1 April 2025; test methods follow Part 1 of the same series, GB/T 11017.1-2024 (National Standards Full-Text Public System). The current version for 220kV is still the 2015 edition of GB/T 18890.2; revision has entered the drafting stage, and the January 2026 draft carries a title stating rated voltage 220kV and Um=252kV (draft standard). Domestic capacity in the 110kV and 220kV tiers is concentrated in a small number of enterprises, for reasons of equipment and test capability, as discussed in Chapter 29.
Voltage class also determines the source of materials. For extra-high-voltage cross-linked polyethylene insulation compound and semi-conductive screen compound at 220kV and above, the domestic industry still has to use imported materials (industry report). With each step up in voltage, the requirements on the cleanliness and uniformity of the insulation rise accordingly, and shortcomings in materials and process are exposed most sharply at this tier.
Industry convention and the regulatory basis are not wholly consistent. The industry usually calls 1kV and below low voltage, 6kV to 35kV medium voltage, 66kV to 220kV high voltage and 330kV and above extra-high voltage. Divergence arises in two places: for the starting point of high voltage there are two competing accounts, 35kV and 110kV; and for the starting point of extra-high voltage, two further accounts, 220kV and 500kV. No national standard gives a unified definition of these boundaries; this article adopts the first set, and readers comparing different sources must first establish which set the other source uses.
Voltage tiering has a further use in actual business: tenders are divided into lots by voltage. The qualification and track-record requirements for 110kV and 220kV differ, and bids across tiers are rare, whereas low and medium voltage are itemised by specification within the same lot. This division makes "how high a voltage a company can make" the first label in classifying enterprises, a better indicator of a factory's position in the industry than capacity figures.
Voltage class determines, first, the insulation thickness and screen structure. Low-voltage cable has only one layer of insulation; medium-voltage cable adds a conductor screen, an insulation screen and a metallic screen outside the conductor; as voltage rises further, the insulation thickens markedly and water-blocking, armour and oversheath structures are added. For the same conductor cross-section at different voltage classes, the outer diameter, weight and price per unit length differ widely, which is why a quotation must state the voltage.
Voltage class determines, second, the test items. The inspection-item table of the older national supervision sampling inspection lists "bending test and the subsequent partial discharge test", "dielectric loss tangent", "4h voltage test" and "conductor resistance" as Class A serious items (CCGF 708.1—2015, Table 3, an older system, cited for reference). The standards cited in the table for partial discharge and dielectric loss are GB/T 12706.2 and GB/T 12706.3, that is, products at 6kV and above; low-voltage cable is not subjected to these two items. Different test items mean different testing costs and delivery cycles.
Voltage class determines, finally, the access boundary. The management catalogue for industrial product production licences confirms that wire and cable is one of 14 categories (Guofa No. 11 of 2024), and the licensing implementation rules limit the scope to five product units, among which power cable comprises the two items "extruded-insulation low-voltage power cable" and "extruded-insulation medium-voltage power cable" (Fangchenggang Municipal Administration for Market Regulation); 110kV and above fall outside this scope. High-voltage and extra-high-voltage products are constrained by standards, type tests and tender qualification review; the specific access rules are given in Chapter 21.
The product ceiling of most industrial clusters stops at medium voltage. The Wuwei municipal government's wording is "mainly power cable below 35KV and electrical equipment wire and cable" (Wuwei Municipal Government), a sentence that makes the local product boundary clear. Below 35kV corresponds to medium and low voltage, the segment covered by the production licence, with a lower testing threshold and the largest number of enterprises; capacity at 110kV and above is concentrated in the few enterprises with vertical towers and complete test capability.
Voltage class is the primary attribute of power cable. For cable of the same nominal cross-section, 1kV and 110kV differ in insulation thickness, number of screens, test items and the list of manufacturers. If a procurement document states only the cross-section and the number of cores and omits the voltage class, the supplier cannot quote and acceptance cannot select a standard.
Chapter 3 Building Wire, Control Cable, Communication Cable and Optical Cable
What the four product categories share is a conductor plus insulation; they differ in what is transmitted, how many structural layers they have and which standards system they belong to. Their names are listed side by side because they often appear on the same table in procurement and statistics; mixing up their standards, by contrast, is the most common source of error.
Building wire is wire for buildings and home renovation at rated voltage 450/750V and below, mainly single-core, divided by insulation material into polyvinyl chloride and rubber lines. PVC-insulated products follow the GB/T 5023 series, which has seven parts. A search on the National Standards Full-Text Public System shows all seven parts to be current, with Part 6 in the 2006 edition and the rest in the 2008 edition (search results). The original texts were not opened part by part for this study, and citation follows the status shown on the search page. Rubber-insulated products follow GB/T 5013.1-2008, effective 1 September 2008 and current (National Standards Full-Text Public System).
Building wire carries the heaviest regulation of the four. It is inside the China Compulsory Certification catalogue, with rubber insulation and polyvinyl chloride insulation corresponding to categories 0104 and 0105 respectively (SAMR Announcement No. 18 of 2020). Announcement No. 9 of 2024 brought flame-retardant building wire into scope, with applications for certification accepted from 1 July 2024, and from 1 July 2025 such products may not leave the factory or be sold without certification (SAMR Announcement No. 9 of 2024). The specific rules on certification and sampling inspection are given in Chapter 21 and Chapter 23.
Control cable is used in control, signal and interlock circuits; its structure is a multi-core stranded assembly, each core separately insulated and the whole covered by a common sheath. The current standard GB/T 9330-2020, Plastic-insulated control cables, is effective 1 October 2020 (National Standards Full-Text Public System). It is not inside the China Compulsory Certification catalogue but is one of the five product units of the industrial production licence, the unit being named "plastic-insulated control cable" (Fangchenggang Municipal Administration for Market Regulation).
Control cable divides into screened and unscreened types. The screened type adds a copper-tape or copper-wire braid over the insulated cores to suppress electromagnetic interference; the unscreened type is used where interference is lower. In selection, the screening method and the earthing method must be fixed together; if only "screened control cable" is specified without the screen construction, the supplier will use the lowest-cost arrangement.
Communication cable carries voice and low-speed data; the conductor is copper, the insulation is mostly polyolefin, and the construction includes a moisture barrier and a polyolefin sheath. The current standard GB/T 13849-2025 is effective 1 May 2026, replacing GB/T 13849.1-2013 in full on the same day, with the old standard withdrawn from that date (new standard, old standard). The phrase "1MHz and below" in the standard's title is a clear technical boundary: above this frequency the attenuation of copper cable rises rapidly, and the transmission task passes to optical cable.
Copper cable is being replaced by optical cable in the access network, but demand remains in short-distance, low-rate situations and where power must be supplied along the line. The ability to carry power is an inherent advantage of copper cable over optical fibre, and this ensures that it will not disappear entirely, although demand will continue to decline as the access network is converted to fibre.
Optical cable uses silica fibre as its transmission medium, with a core, cladding, coating and outer sheath; its standards system is separate from that of copper cable. GB/T 7424.3-2025, Optical fibre cables — Part 3: Sectional specification — Outdoor cables, is effective 1 February 2026 (National Standards Full-Text Public System). China's demand for optical fibre and cable has accounted for about 50% of global consumption for ten consecutive years (industry report), a share that indicates the position of domestic optical cable capacity and market size in the world.
| Product | Object transmitted | Main standards | Compulsory certification | Among the five production-licence units |
|---|---|---|---|---|
| Building wire | Low-voltage electrical energy | GB/T 5023, GB/T 5013 | In the catalogue (0104, 0105) | No |
| Control cable | Control and signal | GB/T 9330-2020 | Not in the catalogue | Plastic-insulated control cable |
| Communication cable | Voice and low-speed data | GB/T 13849-2025 | Not in the catalogue | No |
| Optical cable | Optical signal | GB/T 7424.3-2025 | Not in the catalogue | No |
The table is compiled for this article from current standards and regulatory catalogues. The five production-licence units are overhead stranded wire, plastic-insulated control cable, extruded-insulation low-voltage power cable, extruded-insulation medium-voltage power cable and overhead insulated cable (Fangchenggang Municipal Administration for Market Regulation).
The four categories write specifications differently. Building wire is expressed by single-core cross-section, control cable is written as number of cores times cross-section, communication cable is organised by pairs or groups, and optical cable is written by fibre count. With no unified notation, quotations cannot be compared directly. If the same procurement requirement goes to three suppliers and states only the common name, without the standard number or the specification notation, the quotations returned may correspond to four different products. The type-designation rules of the four categories are also independent of one another, the same string of letters meaning different things in different categories; comparing type designations across categories is meaningless, and the standard number must be the reference.
Communication cable and optical cable are often grouped together as "communication cabling", but the grouping holds only at the level of downstream use. The two differ in transmission medium, interface, installation tools and test instruments, and their standards belong to two systems. When statistics merge them, the product unit price is pulled up by optical cable, and dividing the merged amount by the weight of copper cable yields a figure with no physical meaning.
The boundaries between the four categories come down, in the end, to the standard number. Only when a procurement document states the standard number and the voltage does the supplier know which set to follow; writing only the common product name hands the right of interpretation to the other party.
Chapter 4 Special-Purpose Cable: Flame-Retardant, Fire-Resistant, Low Smoke Zero Halogen, Photovoltaic and New Energy Vehicle Applications
Special-purpose cable is a group of products defined by additional performance requirements; the basic construction is still conductor, insulation and sheath. The word "special" refers to one or more performance requirements superimposed on a general-purpose product: resistance to flame propagation, continued energisation during a fire, low smoke and non-toxicity when burning, resistance to ultraviolet light and weathering, and resistance to high and low temperatures and to oil. A single cable may carry several requirements at once, written as a string of designations.
The designation system is unified by GB/T 19666-2019, General rules for flame-retardant and fire-resistant wires and cables or optical cables, issued on 10 December 2019, effective 1 July 2020 and replacing the 2005 edition (standard information). The common designations are Z (flame retardant), W (halogen-free), D (low smoke), U (low toxicity), N (fire resistant), J (mechanical shock) and S (water spray). Whichever letters appear in the type designation, the corresponding tests apply.
Flame retardance is tested at two levels. Single-cable flame retardance is tested to GB/T 18380.12 and GB/T 18380.13, examining whether one cable self-extinguishes after the flame is removed. Bunched flame retardance is tested to GB/T 18380.33 to GB/T 18380.36, with multiple cables laid in a bundle and then burned; classes A, B, C and D correspond to non-metallic material volumes per metre of specimen of 7, 3.5, 1.5 and 0.5 litres, with a charred height not exceeding 2.5 metres as the pass criterion. The two levels are not interchangeable: a product that passes the single-cable test may continue burning when laid in a bundle because the cables heat one another. Where a project requires bunched flame retardance and only a single-cable test report is supplied, acceptance does not hold.
Flame retardance and fire resistance can be combined. When the type designation carries both Z and N, both tests must be performed, and the order of testing is also specified: first the flame-retardance test for flame propagation, then the fire-resistance test for the ability to carry current under fire conditions. Claiming "flame-retardant and fire-resistant" on the basis of only one of the two leads to the item being judged missing at acceptance.
Fire resistance has three grades: N is fire resistance alone, NJ is fire resistance plus mechanical shock, and NS is fire resistance plus mechanical shock and water spray (GB/T 19666-2019). The difference lies in the severity of the fire scenario. Class N proves only that the cable can carry current in a flame; NJ and NS add falling-object impact and firefighting water spray, corresponding to the secondary damage more commonly seen in real fires.
Fire-resistance testing goes beyond burning. The NS grade requires the cable to remain energised while water is sprayed, and the test rig must apply the specified voltage and monitor for breakdown; the NJ grade requires the cable to withstand the impact of a specified falling mass after burning. The test equipment differs from the test line for general-purpose cable, and the number of laboratories with full capability across all three grades is limited, which is why special-purpose cable takes so long to submit for testing.
Halogen-free low smoke is a set of measurable indicators, with the pass values written in the same standard: the pH of the aqueous solution of combustion gases not lower than 4.3, and conductivity not higher than 10 microsiemens per millimetre; the content of halogen acid gases not more than 0.5% for HCl and HBr combined, and not more than 0.1% for HF (GB/T 19666-2019; the text of the standard was not obtained for this study, and the clause numbers and limit values remain to be verified). Low smoke is assessed by light transmittance, an example being a minimum transmittance of 60% where the cable outer diameter exceeds 40 millimetres, tested to GB/T 17651.2. Low toxicity is a recommended item, with a toxicity index ITC not greater than 5.
These indicators change the material system. Flame-retardant PVC insulation relies on added flame retardants and releases hydrogen chloride when burning, with high smoke density and strong corrosivity; low smoke zero halogen products switch to polyolefins with metal hydroxides, and the combustion products are mainly water and carbon dioxide, at the cost of a high filler loading, lower mechanical strength, a narrower extrusion process window and higher material cost. Where halogen-free performance is required, it must be confirmed at the same time that mechanical properties and ageing indicators still meet the product standard; otherwise the product obtained will be compromised in mechanical performance.
Photovoltaic cable is the most internationally standardised category of special-purpose cable. Mainstream products are made to EN 50618:2014 or IEC 62930, type H1Z2Z2-K, DC voltage 1.5kV; the earlier German-standard approach is TÜV 2PfG 1169/08.07, type PV1-F, AC 0.6/1kV and DC 1.0kV; China also has industry standard NB/T 42073-2016, type PV-YJYJ (TÜV testing and certification material). The three sets of standards correspond to two generations of product and two markets, and the choice differs between export and domestic sale.
The service conditions of photovoltaic cable are long-term outdoor exposure: intense ultraviolet, day-night temperature swings, a DC electric field, and possibly ozone and salt spray. The construction is usually double insulation, the insulation material mainly cross-linked polyolefin, and the sheath emphasises weather resistance and UV resistance. On the DC side the voltage is high and the current does not pass through zero, so arcs last longer; this is why photovoltaic systems place additional requirements on cable.
Cable for new energy vehicles divides into two parts, in-vehicle and charging. The GB/T 25085 series for road-vehicle cable is identical to ISO 19642:2019 and covers two voltage platforms, AC 600V/DC 900V and AC 1000V/DC 1500V (local market regulation announcement). Charging cable is governed by GB/T 33594-2025, effective 1 May 2026 (National Standards Full-Text Public System).
The service conditions of in-vehicle high-voltage harnesses are the opposite of those for photovoltaic cable: confined space, continuous vibration, temperature cycling and strong electromagnetic interference. The harness requires screening and stranding, and connectors and cable are often supplied together by the same supplier. Charging-gun cable adds two mechanical requirements, mating-cycle life and flexing life, with test methods to GB/T 33594-2025.
No credible public statistics for the segment size of special-purpose cable were obtained for this study. Figures encountered in the search, such as "low smoke zero halogen, RMB 38 billion" and "high temperature resistant, RMB 286 billion", come from document-resale sites, contradict one another and have no statistical basis, so they are not used here. The judgement in this paragraph is qualitative only: special-purpose cable is growing faster than general-purpose products and its share is rising, but no verifiable data source exists for the specific proportion.
A special requirement must be written down in three parts — designation, grade and test method — and missing any one makes acceptance impossible. The phrase "flame-retardant cable" may mean single-cable flame retardance or bunched class A; "low smoke zero halogen" may be assessed against the full set of items in GB/T 19666 or only against pH and conductivity. If the contract is not specific, the test report will not match.
Chapter 5 Downstream Structure: Power Grid, Construction, Industry and New Energy
The downstream reach of wire and cable is broad. The annual-report wording of listed companies lists power, rail transit, construction, new energy, communications, ships, intelligent equipment, metallurgy and petrochemicals, among other sectors (Shanghai Securities News report). The list serves to show that demand is dispersed; it cannot be used to allocate shares, because one company's customer list and the demand structure of the whole industry are two different things.
The power grid is the single largest buyer. State Grid investment in 2025 is expected to exceed RMB 650 billion for the first time (Xinhua). Grid procurement is characterised by centralised tendering, prequalification, delivery by batch and fixed payment terms, and the threshold for entering the supplier list is higher than the technical threshold of the product itself. UHV DC cable is one segment; according to an institutional estimate cited in the 2025 annual-report summary of Qifan Cable, its market size exceeded RMB 80 billion in 2025 (Shanghai Securities News).
Construction is the main destination for building wire. Wire consumption in home and commercial fit-out varies with completed floor area; the products are highly standardised and the channel has many tiers, and brand and certification weigh more in purchasing decisions than they do with industrial customers. Quality failures in building wire fall directly on residents, and regulation is correspondingly the heaviest; the China Compulsory Certification catalogue covers precisely this segment.
Industry is the destination for electrical equipment cable. Such cable leaves the factory with the equipment; the customer is the equipment manufacturer, certification requirements follow the finished machine, and individual cables are often not certified separately. Industrial customers are characterised by small batches, many specifications and tight delivery schedules, and they demand more flexibility from suppliers than grid customers do.
Rail transit and ships form another group of downstream users, with requirements centred on low smoke zero halogen and fire resistance. Tunnels and ship compartments are enclosed spaces where smoke and halogen acid gases are the main hazards in a fire, so these two sectors write the halogen-free low-smoke indicators of GB/T 19666 into their tender documents. Unit prices are higher than for general-purpose cable, batch sizes are smaller than grid orders, and suppliers are concentrated among the few enterprises with complete certification.
New energy divides into several lines: the DC and AC sides of photovoltaic plants, the collector lines and tower-internal cable of wind farms, the connecting cables of energy storage systems, the in-vehicle harnesses and charging-gun cable of new energy vehicles, and the array and export submarine cables of offshore wind. According to an institutional estimate cited in the same annual-report summary, the domestic submarine cable market was RMB 17.5 billion in 2025 (Shanghai Securities News). The barrier in submarine cable lies in laying vessels and repair capability, and the product's own premium comes from accessories and joints.
Communications is the fifth destination. Operator centralised procurement is mainly optical cable; demand for copper cable is concentrated in the access segment and inside equipment rooms. Such procurement is characterised by an annual framework, orders placed in batches and a long price-negotiation cycle, and the gross margin of optical cable companies moves with the procurement price. Domestic demand for optical fibre and cable has accounted for about 50% of global consumption for ten consecutive years (industry report), a share that indicates the weight of domestic demand in the global optical cable market.
The same report states that the market for dedicated cable in the new energy sector exceeded RMB 150 billion in 2025, accounting for 28% of total industry output value; these figures come from the China Submarine Cable Industry Development White Paper (2025) and are on an institutional white-paper basis. This share does not reconcile with the total output value of RMB 1.82 trillion given in the same report: dividing RMB 150 billion by 28% implies a base of about RMB 535.7 billion; if the base were RMB 1.82 trillion, RMB 150 billion would be only 8.2%. The two ratios cannot both hold, and this article cites the size figure of RMB 150 billion without citing the 28% share. Such internal inconsistencies are not rare in industry reporting; the remedy is to trace each figure back to its own denominator.
Exports form a demand channel independent of domestic downstream users. On the customs statistical basis as compiled by an industry body, wire and cable exports in 2025 were USD 32.283 billion (322.83 亿美元) and 3.15 million tonnes (315 万吨), at an average export price of USD 10,248.58 per tonne (compiled by Huajing Industry Research Institute). The original tables from the General Administration of Customs were not obtained for this study; the figures above are on the compiled basis. Imports in the same year were 130,000 tonnes and USD 4.288 billion, at an average import price of USD 32,987.82 per tonne (compiled import data), roughly three times the export price. This gap is consistent with a structure in which high-end products depend on imports, while exports consist mainly of low- and medium-voltage and general-purpose products.
Product classification and downstream classification are two different coordinate systems. The five-category shares in Chapter 1 are on a product basis; the 39.5% for power cable refers to the composition of product sales and must not be read as the grid accounting for 39.5% of industry demand. The grid also buys building wire and optical cable, and construction also buys power cable. Mixing the two coordinate systems produces conclusions such as "grid demand accounts for 40%", for which there is no source.
The influence of downstream structure on supply shows in the form of capacity. Grid orders require qualifications and track record, so capacity concentrates in a few enterprises; construction and industrial orders are dispersed, so capacity concentrates in clusters. When both kinds of order fall to the same group of factories, the product structure of those factories is stretched apart: on one side cross-linking lines for 110kV and above, on the other extrusion lines for 450/750V building wire. The two production lines of a single enterprise differ in customers, standards and gross margin.
The first step in downstream analysis is to establish which coordinate system a basis belongs to. Classification by product, by sector and by tendering entity each has its own use, and the share from any one cannot be applied directly to another. The industrial clusters discussed in Part Two use a regional basis, which differs again from the two coordinate systems used in this chapter.
Chapter 6 Four Measurement Bases: Output Value, Revenue, Copper Weight and Conductor Cross-Section
No single authoritative annual figure for "industry operating revenue" exists for this industry. The association gives "sales revenue", the statistics bureau gives "main business revenue of enterprises above designated size", and consulting firms give "market size"; the three are not interchangeable.
Set side by side, the figures differ widely. The Wire and Cable Branch of the China Electrical Equipment Industry Association states that annual sales of the wire and cable industry in 2025 are expected to exceed RMB 1.8 trillion (association release). Company annual reports, citing association data, state that total industry output value in 2025 is expected to be RMB 1.82 trillion (Shanghai Securities News). A consulting firm gives a 2024 market size of RMB 1.35 trillion (white paper report). On the National Bureau of Statistics basis, main business revenue of "wire and cable manufacturing" in 2023 was RMB 1,631.103 billion, a figure seen only in second-hand citation (second-hand citation). Another industry body gives a 2024 basis of more than RMB 1.7 trillion, including optical fibre and cable (industry body citation). Media cite a 2024 market size of about RMB 1.2 trillion (industry report).
These figures run from RMB 1.2 trillion to RMB 1.82 trillion, a spread of more than RMB 600 billion. The difference comes from three places: whether double counting of copper processing is included, whether enterprises below designated size are included, and whether optical cable and other products are included. Dividing any one figure by another yields a growth rate that means nothing.
The first level of difference between output value and revenue lies in what is calculated. Gross industrial output value is calculated on production volume and operating revenue on sales volume, with inventory and work in progress standing between them. When the copper price rises, inventory appreciation pushes revenue up even though output has not changed; when the copper price falls, the reverse occurs.
The second level of difference lies in the enterprise boundary. Above designated size means annual main business revenue of RMB 20 million (2,000 万元) or more; enterprises below that size are not included in the statistics bureau's aggregation of main business revenue. The association's sales figure covers all enterprises. The basis used for industrial clusters often includes supporting enterprises and sometimes includes copper processing. The public wording of Yixing states this level most clearly: "a core-business scale of nearly RMB 180 billion and more than RMB 300 billion on a full basis including copper processing" (Yixing Municipal Government). For the same place, including or excluding copper processing makes a difference of RMB 120 billion, a factor of about 1.67.
The third level of difference comes from the copper price itself. The average Shanghai copper price in 2025 was RMB 80,695 per tonne, up 7.52% on 2024 (Jiangxi Copper 2025 annual-report summary). Qifan Cable's 2025 annual report states that raw materials account for more than 80% of the total cost of wire and cable, of which copper and aluminium account for more than 70% (Shanghai Securities News). Assuming that copper and aluminium together account for 70% of cost and that this portion moves in step with the copper price, the metal alone would push cost up by about 5.3%. This is our estimate, not a figure disclosed by an enterprise. The conclusion is that a year's movement in the copper price is enough to move revenue by several percentage points with output unchanged, and there is no fixed conversion factor between output value and physical output.
Copper weight is the third basis, and the one most easily mistaken for a physical quantity. It is calculated as nominal cross-section times number of cores times length times the density of copper, then multiplied by a stranding factor. A copper conductor of 1 square millimetre cross-section and 1 kilometre length has a volume of 1,000 cubic centimetres and, at a copper density of 8.89 grams per cubic centimetre, weighs 8.89 kilograms. For 4 cores of 95 square millimetres over a length of 1 kilometre, the copper weighs about 3.38 tonnes and, at the 2025 average Shanghai copper price, the copper material alone costs about RMB 272,600. The above is our estimate and does not take account of stranding, compaction, insulation or sheath. The conversion table and correction factors are given in Chapter 18.
Copper weight cannot be used to infer cross-section. At the same cross-section, the number of cores determines the total copper content; stranding makes the length of each wire greater than the cable length; a compacted conductor reduces the outer diameter without changing the nominal cross-section; and screens, armour and sheaths bring in other metals. Weighing gives the total metal content of a cable, and recovering the cross-section from it requires knowledge of the construction.
Conductor cross-section is the fourth basis. What procurement and acceptance call cross-section is the nominal cross-section, whereas the standard assesses DC resistance. GB/T 3956-2008, Conductors of cables, sets the resistance limits for each type of conductor; testing is to GB/T 3048.4 and the judgement is based on DC resistance at 20 degrees Celsius, with geometric area playing no part. At the same nominal cross-section, differences in stranding, number of wires and degree of compaction produce differences in geometric area, while current-carrying capacity is determined by resistance. This also leaves room for cross-section shortfall: reducing the cross-section, introducing impurities or reducing the number of wires will all push resistance above the limit. Among the 158 batches of non-conforming products in the 2024 national supervision sampling inspection, 85 batches failed on conductor resistance, accounting for 53.8%; among the 65 non-conforming batches in 2025, 19 failed on conductor resistance, accounting for 29.2% (2024 notice reproduction, 2025 notice). Conductor resistance was the leading non-conformity item in both years; see Chapter 23.
Substituting aluminium for copper separates the two bases completely. The resistivity of aluminium is about 1.64 times that of copper, so to reach the same resistance the aluminium conductor must have a cross-section 1.64 times larger (China Electrical Equipment Industry Association). Given a copper density of 8.89 and an aluminium density of 2.70 grams per cubic centimetre, the weight per unit length of an aluminium conductor under equal resistance is about half that of copper. Both may be written as 95 square millimetres, yet copper and aluminium differ in current-carrying capacity, weight and price; the cross-section figure cannot be used apart from the metal.
| Basis | Question it can answer | Question it cannot answer |
|---|---|---|
| Output value / sales revenue | Total money in the industry, size of the tax base | Physical output, profit |
| Main business revenue above designated size | Sales scale of enterprises above designated size | The whole industry (enterprises below designated size are omitted) |
| Copper weight | Copper consumption, material cost | Output value (excludes conversion fee and other materials) |
| Conductor cross-section | Current-carrying capacity, acceptance judgement | Copper consumption (affected by number of cores, stranding and compaction) |
The table is compiled for this article from the scope of application of the four bases.
The only thing the four bases have in common is that each can be expressed in monetary or physical units; beyond that there is no conversion relationship. Converting one into another requires three sets of information: the enterprise boundary, the price basis and the product structure. If any one set is missing, the converted result is unusable.
Part Two: Scale and Geography
Chapter 7: The 103,600 Cable Entities in This Database
The Tianxia Gongchang enterprise database contains about 4.8 million manufacturing entities, with fields covering enterprise name, summary of main products, business scope, registered address, insured persons and enterprise status. Wire and cable has no ready-made industry code available in this database: the industry field only goes down to the major category of the Industrial Classification for National Economic Activities, one of which is C38, manufacture of electrical machinery and apparatus; wire and cable sits in the medium category C383 under C38, "manufacture of wires, cables, optical cables and electrical equipment", and below that in C3831, "manufacture of wires and cables". Neither of these levels exists in the database, so the stock cannot be taken by code. The entities are legal entities and exclude self-employed businesses; the reference date is October 2026.
The match runs across three fields: enterprise name, summary of main products (company_service) and business scope (business_scope). The core terms are "wire", "cable" and "cable line", and the list is extended by product name to cover power cable, control cable, building wire, overhead line, rubber-sheathed cable, mining cable, magnet wire, winding wire, optical cable, coaxial cable and others. A hit in any field counts; entities are de-duplicated by unique entity identifier, and external back-filled entities and clear outliers in insured persons are removed, giving 103,564 entities.
The nature of this figure is "legal entities whose name or text mentions a wire-and-cable term", and it must not be read as the industry stock registered under C3831. This basis does not distinguish manufacturing, sales and installation. Wherever this Part refers to "this database", it means this keyword-match basis; the shares, multiples and averages that follow all take these 103,564 entities as the denominator.
The deviation of the basis runs in two directions. Under-recall occurs with companies that write only type designations: codes such as YJV22, BV, RVV, KVV and YJLV do not necessarily carry the word "cable" when they appear in the product summary; companies that make only magnet wire write "enamelled wire", and those that make only upstream products write "copper rod" or "copper wire", none of which hits the term list. Over-recall comes from adjacent products and adjacent process steps: cable tray, cable compound, cable accessories, cable reel and cable trench cover all contain "cable"; "wire" also hits utility poles and cord sets; and traders and contractors whose business scope reads "wire and cable sales", "mechanical and electrical installation" or "cable laying" likewise enter the database.
The net effect of the two directions can be checked against local bases. Yixing officially reports nearly 500 cable companies, while this database counts 1,544, about three times as many; Ningjin officially reports more than 2,500 manufacturers and supporting enterprises, while this database counts 2,120, about 15% lower. One place is high and the other low, which shows that the keyword basis is neither equal to the full set of manufacturing enterprises nor a subset of it; when citing it, it must be placed alongside the local basis.
Total insured persons are 3,051,263, or 3.0513 million; 64,655 entities have an insurance record, 62.4% of the total; the average is 29.5 persons per entity across all entities and 47.2 persons per entity across those with a record. The first average puts the 37.6% of entities with no usable insurance data into the denominator, the second counts only entities with a record, and the two figures must be used separately. Insured persons are taken from the annual report disclosure item; a zero value does not mean there are no employees, since it may reflect non-filing for the year, non-participation, or the use of dispatched labour, and 62.4% indicates only the coverage of filing.
The table below is compiled from provincial statistics on wire-and-cable-related entities in this database (keyword-match basis, October 2026).
| Province | Entities | Share of this database's national total |
|---|---|---|
| Jiangsu Province | 15,336 | 14.8% |
| Hebei Province | 12,118 | 11.7% |
| Guangdong Province | 10,481 | 10.1% |
| Zhejiang Province | 10,167 | 9.8% |
| Shandong Province | 9,642 | 9.3% |
| Henan Province | 5,145 | 5.0% |
| Anhui Province | 4,480 | 4.3% |
| Shanghai Municipality | 3,544 | 3.4% |
| Total, eight provinces | 70,913 | 68.5% |
In this table, "national" means the national total of this database, not the industry total.
Jiangsu, Hebei, Guangdong and Zhejiang together account for 48,102 entities, or 46.5%; the top eight provinces account for 68.5%, and the remaining 23 provincial-level units share 32,651 entities. Shanghai's 3,544 entities enter the top eight, the only municipality directly under the central government to do so; Beijing's cable entities do not enter the top eight. The difference between the municipalities indicates that the distribution of entities in this industry is more closely tied to the local manufacturing base.
County-level units are more dispersed. The table below lists the eight county-level units with the most wire-and-cable-related entities in this database (including prefecture-level cities without districts, whose statistical level is the same as that of a county).
| County-level unit | Province | Entities | Share of province | Share of this database's national total |
|---|---|---|---|---|
| Yueqing | Zhejiang | 2,902 | 28.5% | 2.8% |
| Dongguan | Guangdong | 2,553 | 24.4% | 2.5% |
| Ningjin County | Hebei | 2,120 | 17.5% | 2.0% |
| Yixing | Jiangsu | 1,544 | 10.1% | 1.5% |
| Renqiu | Hebei | 1,450 | 12.0% | 1.4% |
| Yangzhong | Jiangsu | 823 | 5.4% | 0.8% |
| Tianchang | Anhui | 811 | 18.1% | 0.8% |
| Zhongshan | Guangdong | 787 | 7.5% | 0.8% |
| Total, eight places | — | 12,990 | — | 12.5% |
The eight county-level units total 12,990 entities, or 12.5%; the top four total 9,119, or 8.8%. Against the 46.5% of the top four provinces, the 8.8% of the top four counties shows how quickly concentration falls when the same set of figures moves down one level; the largest single county accounts for only 2.8% nationally. Chapters 8 to 12 examine these counties one by one, and output value is far more concentrated than entity counts.
Against published enterprise counts, 103,564 looks large. Figures reproduced by the association give more than 7,500 manufacturers in 2025, of which more than 4,500 are above designated size (Shanghai Securities News); the consultancy basis gives more than 12,000 enterprises, of which more than 4,000 are above designated size (report on a consultancy white paper); another media account cites 4,000 above designated size, more than 70 listed companies and more than 700,000 persons employed (China Jiangsu Net). This database is 8.6 times the 12,000, and its 3.0513 million insured persons are 4.4 times the "more than 700,000 persons employed". The multiples come from what is counted: the published figures are manufacturers or enterprises above designated size, whereas this database covers all legal entities whose name and text mention wire and cable, including sales, installation, supporting and part-time operations.
A second set of measures comes from customs. In 2025 exports of wires and cables were USD 32.283 billion (322.83 亿美元) and 3.15 million tonnes (315 万吨), an average export price of USD 10,248.58 per tonne; imports were USD 4.288 billion and 130,000 tonnes, an average import price of USD 32,987.82 per tonne (Huajing Industry Research Institute export data, import data). The original tables from the General Administration of Customs were not obtained this time because the website blocked access, and both figures are customs-basis statistics compiled by an industry body. The average import price is about 3.2 times the average export price, which indicates that the composition of imports and exports differs; the price gap cannot be read directly as a quality gap, but it is consistent in direction with the judgement that high-end products in this industry depend on imports (see Chapter 34). Entity counts, insured persons and customs volumes are three sets of measures answering three questions, and they cannot be converted into one another.
Chapter 8: Guanlin Town in Yixing, Jiangsu
Jiangsu has 15,336 entities, 14.8% of this database, the largest of any province and 3,218 more than second-placed Hebei. Within the province the concentration is in Yixing: this database counts 1,544, 10.1% of Jiangsu, ranking fourth among county-level units nationally. Jiangsu's remaining 13,792 entities are spread across other counties and cities in the province, of which Yangzhong's 823 enter the national county top eight. Yixing's share of its province is lower than Ningjin County's 17.5% of Hebei, so Jiangsu is less concentrated than Hebei. Yixing's cable companies are concentrated in Guanlin Town, whose administrative area is only a small part of the city; the industry mainly sits there.
A statement by the Yixing Municipal Government in June 2026 gives a set of scale figures and two share figures: the city's cable main business is worth nearly RMB 180 billion (1,800 亿元), and the full basis including copper processing exceeds RMB 300 billion; the city hosts nearly 500 cable companies, of which more than 200 are above designated size; power cable accounts for more than 15% of the national market, and the share of communication feeder cable exceeds 35% (Yixing Municipal Government).
These figures cannot substitute for one another. The RMB 180 billion is the city's cable main business, the RMB 300 billion is the full basis including copper processing, and the two differ by about RMB 120 billion, the difference lying mainly in upstream segments such as copper processing. Copper processing and cable are in the same city, which means a short purchasing radius for raw materials and, when the copper price moves, inventory and cost changing in the same area at the same time (see Chapter 14). The nearly 500 are cable companies and the more than 200 are those above designated size, on a production basis. Placed alongside the 1,544 of this database, the difference comes from what is counted: this database's keyword basis includes sales, installation, supporting and part-time operations (see Chapter 7).
Guanlin Town has a separate set of figures. In 2025 the town's industrial output value above designated size was RMB 197.8 billion (interview with the Yixing Municipal Government); in the same year the town had 165 electrical equipment enterprises above designated size with an output value of RMB 86.33 billion (Guangming Online). The RMB 197.8 billion is all industrial output above designated size in the town, not cable output; the RMB 86.33 billion is electrical equipment above designated size, a scope wider than wire and cable. Dividing one by the other, electrical equipment accounts for 43.6% of Guanlin's industrial output above designated size; spreading the RMB 86.33 billion across 165 enterprises gives an average of RMB 520 million each. Both results rest on the "electrical equipment above designated size" basis and must not be read as the average output per cable enterprise.
Guanlin Town alone has 165 electrical equipment enterprises above designated size, close to the more than 200 cable enterprises above designated size in all of Yixing. The two bases cover different industry scopes and cannot be subtracted from each other; but the two figures point in the same direction: Yixing's cable capacity is highly concentrated in Guanlin alone.
The difference between the RMB 197.8 billion and the RMB 86.33 billion, about RMB 111.5 billion, is Guanlin's industrial output above designated size outside electrical equipment. Electrical equipment accounts for 43.6%, and the other 56.4% comes from other industries. Guanlin is not a single-industry town; treating it as a pure cable town would overstate cable's weight in the local economy.
Yixing's 1,544 and Guanlin's 165 electrical equipment enterprises above designated size are not on the same level: the former includes sales, installation and supporting operations, the latter are manufacturers above designated size. Spreading the RMB 180 billion across 1,544 gives an average of RMB 117 million each; spreading the RMB 86.33 billion across 165 gives an average of RMB 520 million each, a difference of 4.5 times. The gap comes mainly from the denominator and must not be read as Guanlin enterprises being 4.5 times larger than other Yixing enterprises. Counting the city's more than 200 above designated size, the 1,544 of this database are 7.7 times the above-designated-size count; nationally, the 103,600 in this database are 23 times the more than 4,500 above designated size on the public basis. The two multiples differ considerably, which indicates that the share of above-designated-size enterprises among all entities is higher in Yixing than the national average of this database; the statistical boundaries of the two places differ, so this comparison can only serve as a reference.
The statement that power cable accounts for more than 15% of the national market needs an explicit denominator. On the association's estimate of sales above RMB 1.8 trillion (1.8 万亿元) in 2025 (China Electrical Equipment Industry Association), 15% is RMB 270 billion, which exceeds the RMB 180 billion of Yixing's entire cable main business, so total industry sales cannot serve as the denominator for this share. On a consultancy estimate of a 2024 market of RMB 1.35 trillion with power cable at 39.5% (report on a consultancy white paper), the power cable segment is about RMB 533.3 billion and 15% is about RMB 80 billion, equivalent to 44% of the city's main business; that denominator is arithmetically plausible, but the public material does not state how the share is defined, and this Part cannot confirm that it is the one being used.
Communication feeder cable refers to products such as coaxial cable and radio-frequency feeder, used in base stations, antennas and cable television networks, and falls under communication cable among the five product categories (see Chapter 3). The 35% for communication feeder cable is higher than the 15% for power cable, about 2.3 times the latter, which indicates that Yixing's national position is stronger in this segment and that its product mix is not limited to power cable. This Part has not obtained product-level output value for Yixing and cannot convert the two shares into amounts.
Another account puts Guanlin's annual cable output value at more than RMB 120 billion, or about RMB 150 billion including upstream and downstream. This Part has seen only a search snippet, has not verified the original text, and does not use it.
Yixing's industrial scale ranks near the top among county-level units, but this set of published figures belongs to several different bases, and dividing figures from different bases directly produces wrong conclusions.
Chapter 9: Gaogou Town in Wuwei, Anhui
Anhui has 4,480 entities, 4.3% of this database, ranking seventh. Wuwei does not enter the county-level top eight of this database (threshold 787), and this Part has not obtained the city's entity count separately; all published data in this section come from government releases by Wuwei and Gaogou Town.
Material released by Wuwei in September 2026 gives: more than 270 wire and cable manufacturers and supporting enterprises in the city, of which 104 are industrial enterprises above designated size; industrial output above designated size of RMB 42.6 billion in 2025, up 19.2% year on year; and cumulative output of RMB 25.83 billion from 103 above-designated-size enterprises in the first half of 2026, up 42.5% year on year (Wuwei Municipal Government). The same page gives the product positioning as "mainly power cable below 35kV and wire and cable for electrical equipment".
Another municipal government document of December 2025 gives "expected full-year industrial output above designated size of more than RMB 42 billion", with RMB 30.09 billion in the first three quarters, up 17.7% year on year (Wuwei Municipal Government). The RMB 42.6 billion and the "more than RMB 42 billion" are two versions of the same quantity; the former is the completed figure after the fact, the latter an estimate. The RMB 30.09 billion is 70.6% of the RMB 42.6 billion, broadly consistent with the time elapsed in the first three quarters; the fourth quarter alone contributed RMB 12.51 billion, 29.4% of the full year, above the 23.5% average per quarter in the first three quarters.
Of the more than 270 manufacturers and supporting enterprises, 104 are above designated size; counting 270, the above-designated-size share is about 38.5%. That is far higher than the share of above-designated-size enterprises among this database's cable entities: the more than 4,500 above designated size on the public basis are about 4.3% of the 103,600 in this database. The two shares have different numerators and denominators, so the only conclusion is that Wuwei's density of above-designated-size enterprises is markedly higher; the statistical boundaries of "more than 270" and "104" are not stated in the original, and 38.5% can only serve as a reference.
The RMB 25.83 billion of the first half of 2026 deserves separate attention: it equals 85.8% of the RMB 30.09 billion of the first three quarters of 2025, while covering only half a year. The 42.5% growth rate is more than double the 19.2% for the whole of 2025. This is the fastest half-year growth rate this Part has seen in a cable industrial cluster, but with only one half-year data point, no trend can be judged from it.
Spreading across enterprises, 103 enterprises and RMB 25.83 billion in the first half of 2026 give an average of RMB 251 million each; 104 enterprises and RMB 42.6 billion for the whole of 2025 give an average of RMB 410 million, or RMB 205 million for half a year. The first-half average is 22.4% higher than the half-year average of 2025. The enterprise count fell by only one over the year, so output growth was driven mainly by output per enterprise or by prices; this Part has no product-level price data and cannot separate the two sources.
The RMB 42.6 billion and the RMB 25.83 billion belong to two different years and cannot be added together or annualised directly. Doubling the first half gives RMB 51.66 billion, a figure that indicates only the direction of growth and cannot be taken as a full-year forecast; orders and prices in the second half may both change.
Gaogou Town hosts Wuwei's industry. In 2025 Gaogou's industrial output value was RMB 15 billion, up 9.8% year on year, with 98 industrial enterprises above designated size; the share of special-purpose cable in output value rose to 40%, and the share of military cable rose to 15% (Wuwei Municipal Government). The original discloses special-purpose cable and military cable side by side without stating whether military cable is included in special-purpose cable; this Part records them as two separate items.
Gaogou's RMB 15 billion and the city's RMB 42.6 billion cannot be divided directly. The former is "industrial output value", with no indication of whether it is limited to above designated size or whether it covers cable only; the latter is "industrial output value above designated size" for the city's cable and supporting enterprises. The 35.2% obtained from 150 ÷ 426 can only serve as an order-of-magnitude reference. Likewise, Gaogou's 98 industrial enterprises above designated size and the city's 104 cable enterprises above designated size belong to two industry scopes, and the ratio of 94.2% does not hold; but the two figures point to the same fact: Wuwei's cable industry is highly concentrated in Gaogou alone.
Spreading Gaogou's RMB 15 billion across 98 industrial enterprises above designated size gives an average of RMB 153 million each; spreading the city's RMB 42.6 billion across 104 cable enterprises above designated size gives RMB 410 million each. The former denominator includes non-cable industries, so the two averages cannot be compared directly, but they do indicate that the average output of Gaogou's above-designated-size enterprises is lower than that of the city's cable enterprises above designated size.
On product mix, Wuwei is positioned below 35kV, which differs from Yixing's structure covering power cable at several voltage levels, so the two places do not compete directly in the high-voltage market. Special-purpose cable at 40% and military cable at 15% show that the cluster's output value does not come mainly from ordinary building wire. Military cable has an access regime and a customer system different from civil cable; this Part has not obtained Wuwei's military cable qualifications or customer information, so the 15% can only be treated as one figure in the output structure. This Part has also not obtained profit or export data for Wuwei and cannot verify whether the high share of special-purpose cable translates into gross margin.
The only Anhui unit to enter this database's county-level top eight is Tianchang with 811 entities, 18.1% of the province. Tianchang and Wuwei lie in eastern and central Anhui respectively and have different product mixes; this database's count for Wuwei is below 787, which is only a ranking on the keyword basis and has nothing to do with the relative official output value of the two places.
Wuwei's published data are among the most consistent in basis that this Part has seen for a cluster: both the city and the town give enterprise counts, output value and growth rates, but the two levels do not correspond, and dividing one by the other still produces errors.
Chapter 10: Ningjin and Hejian in Hebei
Hebei has 12,118 entities, 11.7% of this database, ranking second nationally. Within the province the two points are Ningjin and Hejian, whose industrial forms differ greatly.
Ningjin had more than 2,500 manufacturers and supporting enterprises in 2025 and cluster revenue of RMB 83.96 billion, and was selected as a national-level characteristic industrial cluster of small and medium-sized enterprises; its products cover more than 30 series and over a thousand specifications in power cable, mining cable, rubber-sheathed cable and others (Xinhua). The 2024 base was 2,393 enterprises, revenue of RMB 68.9 billion, up 28.9% year on year (Hebei Provincial Department of Industry and Information Technology). Calculated across the two years, revenue in 2025 grew 21.9% over 2024; the enterprise count rose from 2,393 to more than 2,500, an increase of more than a hundred, or about 4%. Spreading the RMB 83.96 billion across 2,500 enterprises gives an average of RMB 33.58 million (3,358 万元) each.
This database counts 2,120 entities for Ningjin, 17.5% of Hebei and third among county-level units nationally. That is lower than the official more than 2,500, in the opposite direction from Yixing. The reason lies in what is counted: this database includes only legal entities, while the official "manufacturers and supporting enterprises" may include self-employed businesses; keyword matching also misses companies that write only type designations (see Chapter 7). The difference between 2,120 and more than 2,500 must not be treated as a decline in enterprises.
Using this database's 2,120 as the denominator, the RMB 83.96 billion works out to RMB 39.6 million each, 18% higher than the RMB 33.58 million on the official basis. The same revenue, a different denominator, and a gap of nearly one fifth: this is the direct consequence of the difference between the keyword basis and the official basis.
Revenue grew 21.9% in 2025 and the enterprise count about 4.5%; dividing the two, revenue per enterprise grew about 16.7%. The 2024 growth rate was 28.9%, so 2025 fell back 7 percentage points. The enterprise count kept rising while growth slowed, which indicates that the cluster's expansion has not stopped and that output per enterprise improved more slowly than in the previous year.
The number of above-designated-size enterprises among Ningjin's more than 2,500 is not disclosed in public material; the RMB 83.96 billion is cluster revenue, not output value above designated size. Cluster revenue includes supporting enterprises and differs from output value above designated size; if the above-designated-size count were obtained, it could not be used directly to divide the RMB 83.96 billion.
Hejian's figures are much smaller. The city hosts 160 specialised cable enterprises, and the total operating revenue of wire and cable enterprises in 2024 was RMB 32.31 billion, up 7.6% year on year; products are sold to 18 countries and regions including the United States, Australia and South Africa, forming more than a hundred varieties and tens of thousands of specifications (Great Wall Network, Cangzhou Channel). The 160 is a 2025 figure and the revenue is a 2024 figure; the two belong to different years.
The 160 and the RMB 32.31 billion are not the same set: the former is described as "specialised cable enterprises", the latter as "total operating revenue of the city's wire and cable enterprises". Dividing by the former gives an average of RMB 202 million each; Ningjin's division by 2,500 gives RMB 33.58 million, a difference of 6 times. The denominators differ, so the multiple only indicates that Hejian's enterprises are larger on average than Ningjin's and cannot be taken as a precise ratio. If Hejian's enterprise count including supporting firms exceeds 160, the RMB 202 million would fall accordingly.
The revenue years also differ between the two clusters: Hejian's is 2024 and Ningjin's 2025. The 38.5% comparison has not been adjusted for time; if Hejian continued to grow in 2025, its share in the same year would be higher than this and the gap would narrow accordingly.
The difference in form between the two clusters can be read from this set of figures: Ningjin wins on numbers, with more than 2,500 enterprises producing RMB 83.96 billion; Hejian wins on unit scale, with 160 enterprises producing RMB 32.31 billion, equivalent to 38.5% of Ningjin's. Ningjin's product line is wider, covering power, mining and rubber-sheathed cable; Hejian's public material gives only "more than a hundred varieties and tens of thousands of specifications" without listing leading products. Hejian's export orientation has a definite figure, 18 countries and regions; Ningjin's export data were not obtained this time, so the export orientation of the two clusters cannot be compared.
"More than 30 series" and "more than a hundred varieties and tens of thousands of specifications" are not on the same level and cannot be compared directly. A series is a product major category and a variety is a specification combination under a series; Hejian gives varieties and specifications while Ningjin gives series, and the two lack a common unit of count.
In this database's county-level top eight, Hebei's second entry is Renqiu with 1,450 entities, 12.0% of the province, and Hejian does not enter the top eight (threshold 787). Renqiu and Hejian both belong to Cangzhou but to different county territories; this database does not merge the counts, and the two cannot substitute for each other. Ningjin and Renqiu together have 3,570 entities, 29.5% of Hebei; the remaining 8,548 are spread across other counties and cities. Hejian's absence from the top eight means that this database's keyword count for Hejian is below 787, which is compatible in order of magnitude with the official 160 specialised cable enterprises; this Part has not obtained the exact figure and draws no further inference.
Ningjin and Hejian represent two kinds of cluster: one piles up revenue through the number of enterprises, the other achieves nearly 40% of the former's revenue with far fewer enterprises.
Chapter 11: Zhejiang: Nanxun in Huzhou and Yueqing in Wenzhou
Zhejiang has 10,167 entities, 9.8% of this database, ranking fourth nationally. The two points in the province make two different products: Nanxun in Huzhou makes magnet wire, and Yueqing in Wenzhou makes electrical supporting products.
Nanxun's magnet wire cluster is in Lianshi Town. The town has about 20 magnet wire enterprises with annual output value above RMB 16 billion, close to 20% of the national annual output value, making it one of the country's important magnet wire clusters (Huzhou Municipal Administration for Market Regulation). Dividing the RMB 16 billion by 20 gives an average of RMB 800 million each. The 20 is "about" and the RMB 16 billion is "above", so this average can only serve as an order-of-magnitude reference.
The RMB 16 billion is close to 20% of the national figure, which implies national annual magnet wire output value of about RMB 80 billion. This is back-calculated from a single cluster's share, and "close to 20%" is itself approximate, so the RMB 80 billion can only serve as an order-of-magnitude reference and cannot be treated as the statistical value for national magnet wire.
Magnet wire is winding wire among the five product categories (see Chapter 1), not power cable. Its downstream is electric motors, transformers, household appliances and drive motors for new energy vehicles, a customer structure different from that of cable plants. Nanxun's 20 enterprises produce close to 20% of national output value, so single-plant scale is far above that of the cable assembly stage. The two product types have different process lengths: magnet wire is closer to the material end and cable assembly closer to the finished-product end; this difference is consistent in direction with the difference in single-plant scale, but this Part has no product-level cost data and cannot quantify it.
Magnet wire is often mixed with bare conductor and building wire in statistical work. This database's keyword list includes "magnet wire" and "winding wire" but not companies that write only "enamelled wire" (see Chapter 7), so Nanxun's 20 cannot be set directly against this database's count for Zhejiang; Nanxun District also does not enter this database's county-level top eight, and this Part has no keyword count for the district.
Yueqing is the other extreme. This database counts 2,902 entities for Yueqing, 28.5% of Zhejiang and first among county-level units nationally, 349 more than second-placed Dongguan. The 349 is equivalent to 13.7% of Dongguan, a narrow gap; but the error of the keyword basis cannot be quantified, so the first and second places should not be ranked as better or worse. Yueqing's public basis is the "electrical industry cluster", which combines low-voltage electrical apparatus, wire and cable, and electronic components; this time the original material for that cluster was not obtained, nor were wire and cable output value and enterprise counts listed separately. The figure of 2,902 can therefore be compared only with other counties in this database and cannot be set alongside Yueqing's official output value to calculate an average per enterprise.
Among Yueqing's 2,902, there are very likely many entities whose main business is low-voltage electrical apparatus and whose business scope also lists wire and cable. Yueqing has a high density of low-voltage electrical apparatus enterprises, and cord sets, wiring harnesses and plug cords are often produced on the same lines as circuit breakers and contactors within one company. This database does not provide a breakdown of main business revenue, so the part-time portion cannot be separated out; this judgement can only serve as a note on basis and cannot be quantified.
In terms of product relationships, cord sets and plug cords fall under "cord sets (0101)" in the CCC catalogue (State Administration for Market Regulation Announcement No. 18 of 2020, see Chapter 21), and production in the same plant as low-voltage electrical apparatus is a common arrangement. The output value of such enterprises is recorded mainly under low-voltage electrical apparatus, yet on the keyword basis they are counted as cable entities. Within Yueqing's 2,902, the weight of part-time operations determines how far the figure represents cable capacity; this Part has no product-level enterprise counts and cannot give a proportion.
The contrast between Nanxun and Yueqing shows the limits of county rankings. Yueqing's 2,902 is 145 times the 20 of Lianshi in Nanxun, but the two do not make the same kind of product: the former's count includes part-time operations, while the latter's 20 specialise in magnet wire. Placing the two figures on one list indicates enterprise density only, not industrial strength.
Lianshi is a town-level unit and is not listed separately in this database's county ranking; the figure of 20 comes from public material of the Huzhou Municipal Administration for Market Regulation, is not on the same basis as this database's county counts, and cannot be ranked alongside Yueqing or other counties.
County rankings are sensitive to changes in basis. Yueqing is first, Dongguan second and Ningjin third, all above 2,000 entities; fourth-placed Yixing has 1,544 and fifth-placed Renqiu 1,450, with no cliff between the top five. If any one place removed its part-time enterprises, the ranking would change.
Zhejiang's 10,167 is 66.3% of Jiangsu's count; the two provinces together have 25,503 entities, 24.6% of the national total. Yueqing alone accounts for 28.5% of Zhejiang, and the remaining 7,265 are spread across other cities and prefectures, which this Part has not counted one by one.
Chapter 12: Guangdong: Humen in Dongguan and Zhongshan
Guangdong has 10,481 entities, 10.1% of this database, ranking third nationally. The two points in the province are Dongguan and Zhongshan, which together have 3,340 entities in this database, 31.9% of the province.
Dongguan has 2,553 entities, 24.4% of Guangdong, the largest in the province; Zhongshan has 787, or 7.5%. Both Dongguan and Zhongshan are prefecture-level cities without districts, so their statistical level is the same as that of a county and both can be compared with county-level units; but Dongguan's size is that of a prefecture-level city, which must be noted in any cross-sectional comparison. The two together account for 31.9% of Guangdong, and the remaining 68.1% is spread across other cities and prefectures in the province.
Humen Town is where Dongguan's cable industry is concentrated. In 2024 the Humen cable industrial cluster's total industrial output value exceeded RMB 15 billion and it gathered more than 120 enterprises above designated size; in September 2025 the cluster was selected in the fourth batch of national-level characteristic industrial clusters of small and medium-sized enterprises; the planning targets are to exceed RMB 30 billion by the end of 2028 and to reach more than RMB 50 billion by 2030; the chain covers cable materials, precision equipment, connectors and wiring harness manufacturing (Southern Daily). The RMB 15 billion is a 2024 figure, the designation came in September 2025, and the target base was not updated with the designation.
Dividing the RMB 15 billion by the 120 above designated size gives an average of RMB 125 million each. This average is likewise affected by the basis: the RMB 15 billion is the cluster's total industrial output value while the 120 are only the above-designated-size enterprises, so the non-above-designated-size portion is in the numerator but not in the denominator, and the actual average is below RMB 125 million.
The two targets can be converted into average annual growth rates. On the 2024 base of RMB 15 billion, reaching RMB 30 billion in 2028 means doubling in four years, equivalent to 18.9% average annual growth; reaching RMB 50 billion in 2030 means growing 2.33 times in six years, equivalent to 22.2% average annual growth. These are compound growth rates calculated directly from the targets and the base; they contain no price factor and do not represent actual outcomes.
The pace differs between the two targets. From 2028 to 2030, going from RMB 30 billion to RMB 50 billion means 66.7% growth in two years, equivalent to 29.1% a year, above the 18.9% of the first four years. The target curve accelerates, and the last two years are harder than the first four. If the above-designated-size count stays at 120, RMB 50 billion in 2030 corresponds to RMB 417 million each, 3.3 times the RMB 125 million of 2024; if the enterprise count rises, the average growth falls accordingly.
In absolute increments, the first four years go from RMB 15 billion to RMB 30 billion, an average of RMB 3.75 billion added each year; the last two years go from RMB 30 billion to RMB 50 billion, an average of RMB 10 billion added each year. The annual increment in the last two years is 2.7 times that of the first four, and the target's demands on investment and capacity are concentrated in that stretch.
Humen's RMB 15 billion covers cable materials, precision equipment, connectors and wiring harnesses, a scope wider than wire and cable itself. Wiring harnesses and connectors are electronic components, and materials and equipment are upstream; placing them together with cable output value overstates the scale of cable. Conversely, if only the cable portion were taken, the public material provides no breakdown. The word "cluster" must be attached whenever this figure is cited.
A cluster that puts materials, equipment, connectors and wiring harnesses into a single statistical basis shows that the local advantage lies in the completeness of supporting operations, not in cable output alone. This structure differs from the power cable clusters of Yixing and Wuwei: Yixing is mainly power cable and Wuwei mainly power cable below 35kV and wire and cable for electrical equipment, whereas Humen's products fall across materials, equipment, connectors and wiring harnesses.
Zhongshan's public material contains no dedicated statistics for a cable cluster. This database counts 787 entities for Zhongshan, 7.5% of Guangdong and eighth among county-level units nationally, 30.8% of Dongguan's 2,553. Zhongshan's industrial structure is dominated by lighting, household appliances and hardware, and cable mostly exists as a supporting step; this database provides no downstream attribution field, so this can only serve as background and a supporting share cannot be derived from the 787 entities. Zhongshan's 787 is a city-wide count while Humen's 120 is the above-designated-size count for one town, so the two cannot be subtracted; Dongguan's city-wide 2,553 includes towns and subdistricts outside Humen, and this Part has no count for Humen in this database.
Guangdong's 10,481 is 68.3% of Jiangsu's. Jiangsu, Hebei and Guangdong together have 37,935 entities, 36.6% of the national total. Guangdong's remaining 7,141 are spread across other cities and prefectures in the province, which this Part has not counted one by one. The share of Dongguan's city-wide above-designated-size enterprises held by Humen's 120 was not obtained by this Part and cannot be derived by dividing 120 by 2,553. Humen's supply-chain description and Zhongshan's lighting and appliance structure both point to supporting-type products; Yixing and Wuwei are mainly power cable, while Humen is mainly connectors, wiring harnesses and cable materials; the two kinds of cluster serve different downstreams and their business cycles are not synchronised. This Part has no product-level enterprise counts and cannot give the share of supporting-type products.
Chapter 13: Verification Results for "China Cable Hometown"-Type Titles
Introductions to industrial clusters commonly use titles such as "China Cable Hometown", "China Wire and Cable Hometown" and "China Cable City". This Part traced the relevant appellations for Yixing and Wuwei and for Guanlin Town one by one, and states the conclusion first: these titles are indeed in use on local government web pages and in media reports, but none of the three items (the awarding body, the year of award and the designation document) has been verified against an original source.
The table below is compiled from verification of public web pages (verification date 18 October 2026).
| Title | Used by | Awarding body and year | Original source verified by this Part |
|---|---|---|---|
| China Wire and Cable Hometown | Yixing | Not obtained | Used directly on the Yixing Municipal Government web page (link) |
| China Cable Hometown | Wuwei | Not obtained | Used directly both on the Wuwei Municipal Government web page and in an Anhui Daily report (link) |
| China Cable City | Guanlin Town, Yixing | Not obtained | Used directly in an interview with the Guanlin Town government (link) |
| National Wire and Cable Industry Base | Guanlin Town, Yixing | Guanlin describes itself as "awarded", without a body or a year | As above |
| One of China's four major cable production bases | Yixing, Wuwei and others | Only an account of a 2007 inspection and designation; no original document seen | Search snippet only; not used as a basis |
The Yixing Municipal Government writes "China Wire and Cable Hometown" directly when introducing the local industry, without stating who awarded it or when. The Wuwei Municipal Government's industry introduction page and an Anhui Daily report write "China Cable Hometown" directly, likewise without award information. The Guanlin Town government interview uses "China Cable City" and describes the town as "awarded the National Wire and Cable Industry Base", with a body and a year given in neither case. These three cases belong to one class: the title is in use, the provenance cannot be verified.
The claim of "one of China's four major cable production bases" is weaker still. Searches turn up a second-hand account of an on-site inspection and designation in 2007 by the Wire and Cable Branch of China Electrical Equipment Industry Association, the Shanghai Electric Cable Research Institute and others, but this Part has found neither the original document nor a complete list. Under this Part's evidence rules, a designation supported only by second-hand accounts cannot enter the main text; this Part does not use it and does not speculate about the list.
The usual path by which such titles spread is: local media use them first, higher-level media republish, and investment-promotion material and industry reports cite them in turn; as the number of citations grows, the source becomes harder to trace. The several appellations this Part retrieved all follow this path: the body text of the page carries the title, with no award information and no attached designation document. This Part does not speculate about the awarding body; the association names that appear in the searches relate only to the 2007 "four major bases" account and have no established correspondence with the "hometown"-type titles.
Verifying a title requires at least four items of information: the awarding body, the year of award, the document name or list, and the period of validity. If any one of the four is missing, it cannot be determined whether the title is still valid. This Part checked the five titles in the table one by one, and none has all four.
The scope of application also differs between titles and designations. A national-level characteristic industrial cluster of small and medium-sized enterprises corresponds to a defined park, township or county territory, while a title corresponds to an administrative area; the former's spatial unit is more specific, while the latter may cover an entire city or county. The two cover different spaces and cannot substitute for each other.
Contrasting with these titles is another kind of designation: national-level characteristic industrial clusters of small and medium-sized enterprises have a list, and the Humen entry even gives the batch. Ningjin's cluster was selected as a national-level characteristic industrial cluster of small and medium-sized enterprises (Xinhua), and Dongguan Humen's cable cluster was selected in the fourth batch in September 2025 (Southern Daily). These two can be verified to the name of the designation, and the Humen entry also to the batch and year, so they are citable designations; "hometown", "city" and "base" titles are not in this class.
A title cannot serve as evidence because it does not point to any testable threshold. The production licence covers only 5 product units (Fangchenggang Municipal Administration for Market Regulation); CCC covers only cord sets and building wire of 450/750V and below (State Administration for Market Regulation Announcement No. 18 of 2020, see Chapter 21); sampling inspection looks at specific items such as conductor resistance and insulation thickness (see Chapter 23). A place being called a "hometown" neither shows that local enterprises all hold production licences nor that its sampling pass rate is higher than elsewhere. If procurement and research work use such titles, they can only be taken as a lead to a local industry calling card, not as proof of capacity, quality or qualification.
This Part's rule is therefore: wherever "China Wire and Cable Hometown", "China Cable Hometown", "China Cable City", "National Wire and Cable Industry Base" or "one of China's four major cable production bases" appears, it is qualified as "a title used by the local government and media"; the text does not state that a body awarded it in a given year, does not list the awardees, and does not infer industrial scale from it.
This Part handles these titles in one way only: it records that they are in use, not that they were awarded.
Part Three: Raw Material Pass-through
Chapter 14 The Copper Price: Pricing Mechanism and Pass-through Path
Wire and cable is one of the industries across the whole of manufacturing most sensitive to the price of a single commodity. Qifan Cable's 2025 annual report states that "raw materials account for more than 80% of the total cost of wire and cable, and copper and aluminium alone account for more than 70% of the cost of wire and cable" (summary of Qifan Cable's 2025 annual report published by Shanghai Securities News). The same annual report describes the company's pricing model as "copper material price + target gross margin", with the copper material price referenced to the "average price of spot or futures copper material" on the "Yangtze River non-ferrous metal spot market, the Shanghai Futures Exchange and other markets". Zhongchao Holdings, in an investor relations question-and-answer, states that the company adopts a pricing method of "raw material cost + conversion fee + target gross profit", with raw material cost accounting for about 80% of operating cost (investor relations record on Eastmoney). The two companies' statements point to the same structure: in a quotation the copper price is an explicitly agreed baseline, and the conversion fee and target gross margin are stacked on top of that line.
There are at least three public copper prices. The cathode copper futures contract on the Shanghai Futures Exchange reflects the domestic forward price denominated in renminbi; the three-month copper contract on the London Metal Exchange reflects the international price denominated in US dollars; and the Yangtze River non-ferrous metal spot market quotation reflects domestic spot transactions. The three move in the same direction, but their absolute levels and amplitudes of fluctuation do not coincide, and which one is taken as the benchmark directly determines how high or low a quotation sits.
On annual average prices, the average Shanghai copper price for the whole of 2025 was RMB 80,695 per tonne, up 7.52% on 2024; the average three-month copper price on the London Metal Exchange was USD 9,976 per tonne, up 7.63% on 2024 (summary of Jiangxi Copper's 2025 annual report published by China Securities Journal). In December 2025, Shanghai copper broke through RMB 98,000 per tonne at its intraday high, and LME copper broke through USD 12,000 per tonne. In the previous year, the spot copper price rose 9.69% year on year, with a full-year average of RMB 75,000 per tonne (7.5 万元/吨) (industry operating analysis republished on the website of the Zhejiang Provincial Administration for Market Regulation). Two independent sources can be cross-checked: 80,695 divided by 1.0752 is approximately 75,050 per tonne, which matches the average of RMB 75,000 per tonne.
Looking further ahead, in September 2026 an industry media outlet described the copper price as having "recently settled firmly above the historic threshold of 100,000 yuan per tonne" (industry report on China Wire and Cable Network). The statement specifies neither a trading day nor a contract, so it can serve only as a description of a trend and should not be cited as a point-in-time price.
The path by which the copper price enters cable cost falls into four segments. The first is from the copper price to the cost of conductor material, and the two are strictly linear: one tonne of copper makes one tonne of conductor, so however much the copper price rises, the cost of conductor material rises by the same amount. The second is from the cost of conductor material to the cost of the complete cable, which depends on copper's share of complete-cable cost; that share varies with the product, with the copper share of small-cross-section building wire close to the upper bound of the range, the copper cost share of bare copper stranded wire close to 100% and thus outside the range, and that of armoured medium-voltage cable far lower. The third is from the cost of the complete cable to the quotation, which depends on how the copper price benchmark is taken and at what point in time under the pricing formula. The fourth is from the quotation to gross margin, which depends on whether a change in the copper price can be absorbed at the conversion-fee end.
The rule of thumb offered by industry media is that "for every RMB 1,000 per tonne rise in the copper price, cable cost rises by 1% to 1.5%" (industry report on China Wire and Cable Network). This coefficient is often applied directly, but in fact it carries a premise about the price level. Derived from the formula, the effect of a change in the copper price on complete-cable cost equals the copper cost share multiplied by the rate of change in the copper price. Taking the 2025 average of RMB 80,695 per tonne as the base, a change of RMB 1,000 per tonne amounts to only 1.24% of the copper price, and even if copper accounts for nine tenths of complete-cable cost, complete-cable cost rises by only 1.12%. For a change of RMB 1,000 per tonne to bring about a 1.5% rise in cost, the copper cost share would have to exceed 100%, which does not hold arithmetically. Working backwards, the Institute finds that the coefficient holds over a price range of roughly RMB 60,000 to 70,000 per tonne: at that level RMB 1,000 per tonne is 1.4% to 1.7% of the copper price, and with a copper share of eight or nine tenths the result falls exactly in the 1% to 1.5% range. Citing this coefficient without also giving the copper price level produces a large error.
The lag in pass-through is more easily overlooked than the proportion of pass-through. The copper price changes every day, a cable quotation takes an average over one period agreed in the contract, and settlement for copper rod on the purchasing side takes a price over another period. The three periods do not coincide, and the gaps between them are the exposure; Chapter 17 treats the sales side and the purchasing side separately.
The effect of a rise in the copper price on industry profit can be seen in the data of listed companies. In 2024 the average spot copper price rose 9.69%, while the main business revenue of about 80 listed cable companies grew 7.78% year on year and net profit fell 25.71% year on year (industry operating analysis republished on the website of the Zhejiang Provincial Administration for Market Regulation). Another set of figures has about 66% of companies increasing revenue in 2024 and about 57% recording a decline in net profit attributable to shareholders (report on China Jiangsu Network). Revenue growth below the rise in the copper price is consistent with at least one of sales volume or conversion fee making a negative contribution; a fall in net profit far larger than the rise in revenue is consistent with the copper price increase not being fully passed through and the difference being absorbed by profit. Aluminium prices, energy, product mix and impairment also affect profit, and public statements do not allow the respective shares to be separated.
This set of data supports one judgement: the copper price determines the absolute level of operating revenue in this industry, and the conversion fee determines profit. Reading a rise in the copper price as industry prosperity is the most common misreading in this industry.
Chapter 15 Aluminium and Aluminium-Alloy Conductors
The price advantage of aluminium as a conductor comes from the absolute price gap. In 2024 the spot aluminium price rose 6.58% year on year, with a full-year average of RMB 19,900 per tonne (industry operating analysis republished on the website of the Zhejiang Provincial Administration for Market Regulation); in the same year the average spot copper price was RMB 75,000 per tonne. On a per-tonne basis, copper is 3.77 times aluminium. This multiple is often taken directly as the room for substitution, but the proportion actually substitutable is far smaller, and the reason lies in resistivity.
A set of resistivity data published on the website of the China Electrical Equipment Industry Association is the starting point for judgement. Under GB/T 3953-2009, the resistivity of copper is no more than 0.017241 ohm·square millimetre per metre in the soft state and no more than 0.01796 in the hard state; under GB/T 3954-2009, the resistivity of ordinary aluminium is no more than 0.02759 in the soft state and no more than 0.028264 in the hard state; under GB/T 30552-2013, the resistivity of aluminium alloy is no more than 0.028264 in the soft state and no more than 0.028976 in the hard state (wire and cable industry information from the China Electrical Equipment Industry Association). The association's conclusion on the same page is that "the conductivity of aluminium alloy wire ... still falls short of two thirds of the conductivity of copper wire". Checking aluminium alloy in the soft state at 0.028264 against copper in the soft state at 0.017241 gives a ratio of 1.64, a conductivity of about 61% of copper, which is indeed below two thirds.
The conversion more commonly used in transmission and distribution engineering is made on the basis of equal DC resistance. Taking ordinary aluminium in the hard state at 0.028264 against copper in the soft state at 0.017241 gives a ratio of 1.64, meaning that the aluminium conductor cross-section must be enlarged to 1.64 times that of copper to obtain the same DC resistance.
Looking at cross-section alone overstates the weight penalty of aluminium. What follows is an illustrative calculation by the Institute; the assumptions are listed after the worked example.
The table is an illustrative calculation by the Institute and is not data published by any institution
| Item | Copper conductor | Aluminium conductor | Aluminium-to-copper ratio |
|---|---|---|---|
| Resistivity (ohm·square millimetre per metre) | 0.017241 | 0.028264 | 1.64 |
| Cross-section required for equal resistance (relative value) | 1.00 | 1.64 | 1.64 |
| Density (grams per cubic centimetre) | 8.89 | 2.70 | 0.30 |
| Weight per unit length at equal resistance (relative value) | 1.00 | 0.50 | 0.498 |
| Material cost per unit length at equal resistance (relative value) | 1.00 | 0.13 | 0.132 |
| Conductor diameter at equal resistance (relative value) | 1.00 | 1.28 | 1.28 |
The assumptions of the worked example are: the 2024 average spot copper price of RMB 75,000 per tonne and average spot aluminium price of RMB 19,900 per tonne; a copper density of 8.89 grams per cubic centimetre and an aluminium density of 2.70 grams per cubic centimetre; and equal DC resistance, the same length, and no allowance for differences in stranding factor on either side. Under these assumptions, the aluminium conductor at equal resistance weighs about 49.8% of copper and costs about 13.2% of copper in material, that is, roughly one seventh. The conductor diameter, being the square root of the cross-section ratio, is 1.28 times that of copper.
This set of figures explains two things. First, the economic motive for replacing copper with aluminium is very strong at the conductor material line, where material cost can fall to one seventh of copper. Second, the cost gap at complete-cable level is markedly smaller than one seventh. With the conductor diameter enlarged 1.28 times, the circumference of the insulation and the sheath increases in step, and at the same insulation thickness the volume of non-metallic material rises accordingly; the welding and connection requirements for aluminium are also more demanding than for copper, and the cost structure of accessories differs. The one seventh on the material ledger usually shrinks to a modest percentage by the time it reaches a complete-cable quotation.
The applicable boundary of aluminium-for-copper substitution is drawn by engineering conditions, not by price. Aluminium has high resistivity and needs a larger cross-section for the same current-carrying capacity, and this determines that it is viable in fixed-installation scenarios where space is not constrained: building main lines, the DC side of photovoltaic plants and wind farm collector lines all have enough cable tray and duct space to accommodate a larger outer diameter. In scenarios with constrained space, frequent bending or high-reliability connections, this condition does not hold. Industry media put it thus: "in new energy, flexible circuit boards and ultra-high-speed transmission, copper's conductivity and ductility have no very perfect substitute" (industry report on China Wire and Cable Network). High-voltage harnesses for new energy vehicles, industrial robot cables and high-speed data lines all fall on this side.
Aluminium alloy and pure aluminium must be treated separately. Pure aluminium has low mechanical strength and poor creep resistance, and joints loosen easily after long-term fastening, which is the main reason aluminium conductors have suffered accidents historically. Aluminium alloy improves mechanical and creep properties by adding alloying elements, at the cost of a slight rise in resistivity, which is also why aluminium alloy in the soft state at 0.028264 in the table is higher than ordinary aluminium in the soft state at 0.02759. At product level the standard is the GB/T 31840 series, Power Cables with Aluminium Alloy Conductor and Extruded Insulation for Rated Voltages from 1kV to 35kV (《额定电压1kV到35kV铝合金芯挤包绝缘电力电缆》) (drafting document for the national standard), and the voltage range this series covers coincides with the main battleground of copper-core extruded-insulation power cables.
At the policy level, what has been verified and what has not must be distinguished. The Implementation Plan for the High-Quality Development of the Aluminium Industry (2025–2027) (《铝产业高质量发展实施方案（2025—2027年）》) was issued as MIIT Joint Raw Materials Document [2025] No. 62 (工信部联原〔2025〕62 号), jointly by ten departments (publication page of the Ministry of Industry and Information Technology). This verification opened only as far as the issuing notice, with the text of the plan inside an attachment, and the original text was not obtained. Specific proportions in circulation, such as "the proportion of aluminium alloy cable used in distribution network projects shall be no less than 30%" and "aluminium cable shall account for more than 90% of new transmission lines", were not found in the original MIIT text in this verification and are not adopted in this article.
The conclusion that can be written is limited to direction: the policy lists aluminium alloy cable as one of the directions for expanding aluminium consumption, without giving a mandatory substitution ratio. The actual pace of substitution is determined jointly by project cost, route conditions, operating practice and tender documents, not unilaterally by price.
Chapter 16 Insulation and Sheath Materials: PVC, XLPE and LSZH Compounds
Working backwards from the two ratios in Qifan Cable's annual report, insulation compound, sheath compound, screen compound, armour steel tape and wire, and filling material together account for no more than one tenth of complete-cable cost. The method of back-calculation is: raw materials total more than 80%, copper and aluminium together more than 70%, and subtracting one from the other gives the upper bound for non-copper-aluminium raw materials. This figure is inconspicuous in complete-cable cost but plays the leading role in quality accidents. In the 2025 national supervision sampling inspection, 15 batches failed the tensile test after ageing of insulation, 11 batches failed the heat elongation test of insulation, 6 batches failed the weight-loss test of sheath, and 5 batches failed the shrinkage test of insulation (notice No. 24 of 2026 from the State Administration for Market Regulation, republished on the website of the Hunan Provincial People's Government). All of these items point to material formulation and process, not to the conductor.
Another set of figures gives different proportions. A white paper jointly published by a consulting firm and an association states that raw material cost accounts for 90% and copper and aluminium for more than 60% (interpretation of the white paper republished by NetEase Finance). Working backwards on this basis, non-copper-aluminium raw materials account for close to three tenths of complete-cable cost, three times the previous figure. The gap between the two sets of numbers comes from a difference in the statistical object: one is listed companies whose product mix is weighted towards copper, the other the industry as a whole. Any citation must state which set is being used.
Polyvinyl chloride is the main material for building wire insulation and low-voltage cable sheaths. Its formulation consists of resin, plasticiser, stabiliser, filler and flame retardant; the plasticiser proportion determines softness and also the volatility in long-term use. The sheath weight-loss test specified in GB/T 2951.32 measures precisely the weight a specimen loses at elevated temperature; excess weight loss means the plasticiser ratio is too high or the stabiliser system inadequate, and the sheath hardens and cracks with years in service. In the earlier edition of the Implementation Specification for National Supervision Sampling Inspection of Wire and Cable Product Quality (《电线电缆产品质量国家监督抽查实施规范》, CCGF 708.1—2015) this test is listed as a Class B inspection item (CCGF 708.1—2015, an earlier system, for reference). The other set of assessment indicators for PVC is tensile strength and elongation at break before and after ageing, with test methods specified in GB/T 2951.11 and GB/T 2951.12.
Cross-linked polyethylene is used for the insulation of power cables above 1kV. Cross-linking links linear polyethylene molecular chains into a three-dimensional network, so that the insulation keeps its shape at the conductor's long-term operating temperature, and the permissible short-circuit temperature is also higher than that of thermoplastic materials. Whether cross-linking is sufficient is judged by the heat elongation test. The heat elongation test carried out under GB/T 2951.21 measures the elongation of the specimen and its permanent deformation after cooling, under a specified temperature and load; a specimen with insufficient cross-linking stretches markedly here. This too is a Class B inspection item. Among low-voltage products there is also a class of XLPE-insulated cable whose cross-linking process is mainly silane warm-water cross-linking, with lower equipment investment than the dry-process cross-linking used for medium voltage.
Low smoke zero halogen compound is the class with the most difficult formulation. GB/T 19666-2019, General Rules for Flame Retardant and Fire Resistant Electric Wires and Cables or Optical Fibers (《阻燃和耐火电线电缆或光缆通则》), gives separate indicators for halogen-free, low-smoke and low-toxicity performance: halogen-free requires the aqueous solution of combustion gases to have a pH of not less than 4.3 and a conductivity of not more than 10 microsiemens per millimetre, with hydrogen halide content of HCl and HBr together not more than 0.5% and HF not more than 0.1%; low smoke requires the minimum light transmittance to be met at a specified specimen diameter, for example 60% where the specimen diameter is greater than 40 millimetres; low toxicity recommends a toxicity index of not more than 5 and a critical CO concentration of 1750 milligrams per cubic metre (the text of the standard was not obtained in this verification, and the limits remain to be checked). At material level the standard is GB/T 32129-2015, Halogen-Free Low-Smoke Flame-Retardant Cable Compounds for Wire and Cable (《电线电缆用无卤低烟阻燃电缆料》); this standard number comes from a second-hand citation by a testing body, the original text was not opened in this verification, and it is treated as pending verification when cited.
There is tension among the three sets of indicators. Flame retardancy is achieved by adding metal hydroxide fillers: the higher the filler proportion, the better the flame retardancy, and the narrower the window for mechanical properties and extrusion. The halogen-free requirement excludes halogenated flame retardants, which amounts to abandoning the technical route with the highest flame-retardant efficiency. For one formulation to meet at the same time the bunched-tray burning requirements of the GB/T 18380 series, the halogen-free indicators of GB/T 17650.2 and the low-smoke indicators of GB/T 17651.2 leaves little room for adjustment. This explains why LSZH compounds are markedly more expensive than PVC compounds of the same grade, and why this category is often listed as a separate tier in tender documents.
Cables at medium voltage and above have a further layer of semiconductive screening compound. There is a layer of semiconductive material on the conductor surface and another on the insulation surface, whose function is to even out the electric field and eliminate burrs and air gaps at the conductor surface. This layer accounts for a very small share of complete-cable cost, yet has the highest technical threshold. A report on China Jiangsu Network states that for 220kV and above, extra-high-voltage XLPE insulation compound and extra-high-voltage semiconductive screening compound "still have to use imported materials" (report on China Jiangsu Network). This is the hardest external dependence on the material side of the industry, and its location is not the conductor but insulation and screening.
There is also a regulatory change on the material side. Announcement No. 9 of 2024 from the State Administration for Market Regulation brings flame-retardant wire and cable into compulsory product certification: from 1 July 2024 designated certification bodies began accepting applications, and from 1 July 2025 flame-retardant wire and cable that has not been certified and marked with the CCC mark may not leave the factory, be sold or be imported (announcement of the State Administration for Market Regulation). The scope is limited to rubber-insulated and PVC-insulated flame-retardant wire and cable at rated voltages of 450/750V and below. This ties the formulation of flame-retardant building wire to certification, so that a change of formulation requires new samples to be submitted.
The buyer's judgement on the material side can rest on three things. Ask for the grade and supplier of the materials, not only the test report on the finished cable; check the type test data for insulation heat elongation and sheath weight loss, the two items that best reveal whether a formulation has been cut; and for low smoke zero halogen products, require all three reports — halogen-free, low smoke and bunched flame retardancy — since with one missing the set is incomplete.
Chapter 17 How Copper Price Movements Enter the Quotation: Pricing Against a Quoted Date, Price Locking and Hedging
A cable factory faces two price mechanisms at once, one towards downstream and one towards upstream. The two are often conflated, but in fact they differ in what is priced, in the time window and in who bears the risk. The direct consequence of conflating them is to mistake "the quotation follows the copper price" for "the risk has been transferred".
The downstream mechanism is the quotation. Qifan Cable's annual report writes "copper material price + target gross margin", with the copper material price based on the average spot or futures price on the Yangtze River non-ferrous metal spot market and the Shanghai Futures Exchange, and the gross margin determined by factors such as the number of processing passes, process complexity and transport cost (annual report summary published by Shanghai Securities News). Zhongchao Holdings writes "raw material cost + conversion fee + target gross profit" (investor relations record on Eastmoney). What the two formulations have in common is that the copper price part moves with the market while the conversion fee part is set by the cable factory itself. What is really compared on a quotation sheet is the conversion fee column.
Three variables in the quotation mechanism need to be written into the contract to have effect. The first is the object from which the benchmark price is taken: the Yangtze River non-ferrous spot price and the SHFE futures price can differ by several hundred yuan per tonne on the same trading day. The second is the averaging period: the price on the day before quotation, the five-day average, the previous month's average and the average for the delivery month differ widely under a one-way copper market. The third is the validity period: from issue of a quotation through signing to delivery several months may pass, and who bears the change in the copper price over that time depends on the small print on the quotation.
The upstream mechanism is purchasing. A cable factory buys copper rod, whose price consists of the copper price plus a conversion fee, and the way the copper price is fixed falls into several types within the industry. Pricing against a quoted date means the buyer takes delivery first or agrees a quantity first, and the price is fixed by the buyer choosing the futures price of a particular trading day within an agreed time window, with the length of the pricing window and the contract months available written into the purchase contract. Price locking means fixing the price at signing, with the seller hedging by establishing a corresponding position in the futures market. Hedging means the cable factory establishes a position in the futures or options market opposite in direction to its spot exposure, using the profit or loss on the futures side to offset that on the spot side.
These three practices address the same problem: the price risk over the interval between purchase pricing and sales pricing. Their difference lies in who retains the risk. Pricing against a quoted date gives the choice to the buyer, who bears the responsibility for the timing chosen; price locking fixes the price at the signing date, and the seller bears the basis risk thereafter; hedging does not change the price of the spot contract but establishes an offsetting position outside the books, and its effect depends on the correlation between futures and spot prices.
There is a structural mismatch between the two mechanisms. The quotation benchmark on the sales side is usually an average over a period, for example "the average price of the month before the delivery month", whereas pricing against a quoted date or price locking on the purchase side is usually a point-in-time price. The difference between the average and the point-in-time price is the basis borne by the cable factory. When the copper price rises one-way, the average is below the point-in-time price at the end of the period, and the copper price compensation received on the sales side is less than the cost paid on the purchase side; when the copper price falls one-way the reverse holds. This difference is not reflected in the conversion fee but in gross margin.
The 2024 data show the order of magnitude of this difference. That year the average spot copper price rose 9.69%, while the main business revenue of about 80 listed cable companies grew 7.78% year on year and net profit fell 25.71% year on year (industry operating analysis republished on the website of the Zhejiang Provincial Administration for Market Regulation). Revenue growth below the rise in the copper price is consistent with a negative contribution from sales volume or conversion fee; a fall in profit far larger than the rise in revenue is consistent with both basis and conversion fee eroding profit. Aluminium prices, energy, product mix and impairment also affect profit, and public statements do not allow the respective shares to be separated. The risk of copper price fluctuation ultimately falls on the two openings of conversion fee and basis, not on the copper price itself.
The practical effect of hedging depends on whether three conditions hold at the same time. The first is that the exposure is measurable, that is, the enterprise can calculate at a given point how much copper is signed but not purchased, purchased but not sold, and sold but not delivered. The second is maturity matching: the expiry month of the futures contract must line up with the delivery month of the spot, and where it does not the position must be rolled, which carries a cost. The third is accounting treatment: whether hedging gains and losses go to current profit or loss or are deferred affects the pattern of fluctuation in reported profit. With one of the three missing, hedging turns from reducing risk into increasing it.
For the buyer, this chapter yields three operational conclusions. At enquiry, require the supplier to show the copper price and the conversion fee separately on the quotation sheet, since only two separate columns allow horizontal comparison; at signing, set out clearly the object, the period and the validity period of the copper price benchmark, leaving no room for vague formulations such as "settled at market price"; and in the annual framework agreement, agree the trigger conditions and adjustment formula for price revision, turning renegotiation under a one-way market into automatic execution within the contract.
For the cable factory, the room for design in the quotation mechanism lies on the conversion fee side. The conversion fee is determined by the number of processing passes, process complexity and transport cost (annual report summary published by Shanghai Securities News), and all three can be substantiated on the quotation sheet: how many passes, whether cross-linking is included, whether armour is included, and the distance in kilometres. The more detailed the listing, the steadier the bargaining position when the copper price moves.
Chapter 18 Converting Conductor Cross-Section into Copper Consumption
The copper consumption of a cable can be calculated with one formula, provided three assumptions are stated clearly.
W = S × ρ × k × (1 + δ)
Here W is the conductor weight per unit length, S the nominal cross-section, ρ the density of copper, k the stranding factor and δ the rate of positive tolerance on cross-section. ρ is taken as 8.89 grams per cubic centimetre, which converts to 8.89 kilograms of copper per square millimetre of cross-section per kilometre of length. k reflects the fact that when stranded, individual wires follow a helical path and their actual length is greater than the axial length of the cable; the example in this article takes 1.03. δ reflects the positive tolerance left on the nominal cross-section in production; the example in this article takes 0, that is, calculation on the nominal cross-section. In the formula the copper density is a material constant, while k and δ are set values in the example; no standard prescribes their values, and they change with the product and with the factory.
Worked example one: copper-core XLPE-insulated power cable, size 4×240 plus 1×120 square millimetres, length 1 kilometre. There are four phase conductors of 240 square millimetres, each weighing 240 times 8.89, equal to 2,133.6 kilograms per kilometre; one neutral conductor of 120 square millimetres, 1,066.8 kilograms per kilometre; the total on nominal cross-section is 9,601.2 kilograms. Multiplying by the stranding factor 1.03 gives 9,889.2 kilograms, or about 9.89 tonnes.
Multiplying these 9.89 tonnes by the copper price: at the 2025 average Shanghai copper price of RMB 80,695 per tonne, conductor copper cost is about RMB 798,000 per kilometre; at the 100,000 yuan per tonne described by industry media in September 2026, about RMB 989,000 per kilometre. Between the two prices, the conductor cost of the same cable size differs by about RMB 191,000 per kilometre, a rise of 23.9%, exactly equal to the 23.9% by which the copper price rises from 80,695 to 100,000. The copper price is passed through to conductor cost in the same proportion, with no discount at this layer.
Worked example two: copper-core PVC-insulated building wire, size 2.5 square millimetres, length 1 kilometre. On the nominal cross-section, 2.5 times 8.89 equals 22.225 kilograms; multiplying by the stranding factor 1.03 gives 22.89 kilograms. At RMB 80,695 per tonne, copper cost is about RMB 1,847 per kilometre, or about RMB 185 per 100 metres. In this product tier copper cost accounts for the highest proportion of complete-cable cost, while the conversion fee and insulation compound together take a relatively small share, so the quotation for small-cross-section building wire reacts most sensitively to the copper price.
The table is an illustrative calculation by the Institute and is not data published by any institution; the assumptions are a copper density of 8.89 grams per cubic centimetre, a stranding factor of 1.03, calculation on the nominal cross-section, and no allowance for positive tolerance on cross-section
| Size | Total nominal cross-section (square millimetres) | Conductor copper weight (kilograms per kilometre) | Copper cost (yuan per kilometre, at RMB 80,695 per tonne) |
|---|---|---|---|
| 4×240＋1×120 | 1,080 | 9,889.2 | 798,000 |
| 1×240 | 240 | 2,197.6 | 177,300 |
| 1×120 | 120 | 1,098.8 | 88,700 |
| 2.5 (building wire) | 2.5 | 22.9 | 1,847 |
The values in the 1×240 and 1×120 rows of the table can be converted directly from 8.89 kilograms per square millimetre per kilometre: a 240 square millimetre single core is 240 times 8.89 times 1.03, equal to 2,197.6 kilograms, and multiplying by RMB 80.695 per kilogram gives RMB 177,300; the 120 square millimetre core comes to 1,098.8 kilograms and RMB 88,700.
Cross-section is not proportional to current-carrying capacity, and this determines the economics of copper consumption. The current-carrying capacity of a conductor is constrained by heat dissipation: cross-section grows with the square while heat-dissipating surface area grows with the circumference, so the two grow at different rates. Double the cross-section and the current-carrying capacity increases by less than double. The implication for selection is that when cross-section is chosen by current-carrying capacity, copper consumption grows faster than transmission capacity. Engineering therefore has the concept of economic current density, which converts line losses into electricity charges, compares them with conductor investment and finds the cross-section with the lowest total cost. This point has already appeared in the discussion of definitions in Chapter 6; what is given here is its counterpart at the level of the formula.
There is a difference in definition, laid down by the standard, between nominal cross-section and actual copper consumption. GB/T 3956-2008, Conductors of Insulated Cables (《电缆的导体》), takes the upper limit of DC resistance at 20 degrees Celsius as the basis for assessing a conductor, and does not directly prescribe geometric cross-section (full text of CCGF 708.1—2015). This means that "240 square millimetres" is a type designation, and what is actually mandatory is the resistance value. Reduce the cross-section by 1% and DC resistance rises by about 1%, a proportional relationship determined by resistance being inversely proportional to cross-section. How far nominal cross-section can diverge from actual cross-section therefore depends on the margin left by the resistance limit, not on any provision about geometric dimensions.
This difference in definition is one reason conductor resistance has long headed the list of failures in sampling inspection. In the 2024 national supervision sampling inspection, 85 batches failed on conductor resistance, and in 2025, 19 batches, the top failure item in both years (2024 notice, 2025 notice). Conductor resistance failure has two causes: one is a shrunken cross-section, the other insufficient copper purity or improper stranding. The two cannot be distinguished from the measurement result, both appearing as elevated resistance.
The practical implication for the buyer is that price comparison should be made per kilometre, because a price per tonne does not reflect conductor consumption. For cable of the same nominal 4×240 plus 1×120, the difference in per-kilometre price between two factories may come from a different benchmark date for the copper price, from a different conversion fee, or from a different actual cross-section. The first two are visible on the quotation sheet; the third can be verified only by measuring the resistance of every drum. Chapter 31 gives the specific practice for the acceptance stage.
One further point about the Institute's calculation needs to be stated. The stranding factor of 1.03 in the table is an illustrative value; the actual stranding factor varies with the lay length, the number of individual wires and the number of stranding layers, and for compacted conductors the stranding factor and the fill factor also interact. All the weights and amounts given in this section serve to illustrate orders of magnitude and proportional relationships, and cannot be used directly in commercial quotations.
Chapter 19 Cost Structure and the Actual Form of the Conversion Fee
There are four public statements of cost structure, and they do not fully agree. Qifan Cable's 2025 annual report states that raw materials account for more than 80% of total cost and copper and aluminium for more than 70% (annual report summary published by Shanghai Securities News). Zhongchao Holdings states that raw material cost accounts for about 80% of operating cost (investor relations record on Eastmoney). A white paper by a consulting firm states that raw material cost accounts for 90% and copper and aluminium for more than 60% (interpretation of the white paper republished by NetEase Finance). Industry media state that copper material accounts for 60% to 80% of production cost, while another states that it is generally as high as more than 70% (industry report on China Wire and Cable Network, industry report on Sohu).
The four sets of figures can be arranged into one range: copper alone accounts for 60% to 80%, copper and aluminium together for more than 70%, and all raw materials for 80% to 90%. The range moves with the product mix: the copper share of small-cross-section building wire is close to the upper bound; the copper cost share of bare copper stranded wire is close to 100% and thus outside the range; armoured medium-voltage cable is close to the lower bound. Citing any single number requires stating which product class it corresponds to, otherwise the same enterprise will produce two different "cost shares" on two different products.
Subtracting the 70% for copper and aluminium from the 80% for raw materials leaves 10%, which is insulation compound, sheath compound, screen compound, armour steel tape and wire, filler rope and wrapping tape. Subtracting the 60% for copper and aluminium from the 90% for raw materials leaves 30%, which is the same set of things. The two sets differ threefold on this one subtraction, which suggests that the white paper's "raw materials" very likely also include fuel and power and part of manufacturing expense. The difference points to one thing: cost shares from different sources cannot be used together, since once mixed the total does not hold.
The remaining part consists of the conversion fee and gross margin. Qifan Cable's annual report describes the gross margin as "determined by the number of processing passes, the degree of process complexity, transport cost and various other factors" (annual report summary published by Shanghai Securities News). These twelve characters give the three determinants of the conversion fee, each of which can be expanded.
The number of process passes is the most direct item. The environmental impact assessment public notice for the Henan Rongyao Aluminium project gives the original process sequence for three product lines: aluminium stranded wire as "aluminium rod - wire drawing - take-up - semi-finished product inspection - stranding - take-up - inspection - acceptance into warehouse"; YJV-type cable as "raw material - extrusion - take-up - cabling - sheath extrusion - coiling - packing - acceptance into warehouse"; and BV and BVR type wire as "raw material - extrusion - take-up - inspection - coiling - packing - acceptance into warehouse" (EIA public notice on the website of the Yanshi Municipal People's Government). This notice corresponds to low-end building wire and aluminium stranded wire projects, without cross-linking or armour. Comparing the three lines, BV needs only one extrusion pass, YJV needs three — extrusion, cabling and sheath extrusion — and aluminium stranded wire needs two, wire drawing and stranding. The more passes, the more labour, electricity and equipment depreciation are consumed, and the higher the conversion fee.
Voltage class widens the gap in process passes further. The insulation of medium-voltage cable must be cross-linked, and the cross-linking equipment itself is a heavy asset; higher still, conductor screening and insulation screening must be added. The division of work usual in the industry is that high and extra-high voltage use vertical cross-linking lines, while medium voltage uses catenary cross-linking lines (industry article reprint). The examples of vertical towers cited in the industry article are 60-odd metres and 90 metres; tower height varies with voltage class and line design, and the investment scale and depreciation burden far exceed those of a catenary line, so this part of the cost must be allocated into the conversion fee for high-voltage products.
Transport cost is the third item, and it carries no small weight in the cable category. Cable is heavy freight: for copper-core 4×240 plus 1×120 cable the conductor alone comes to nearly 10 tonnes per kilometre, and the complete cable is heavier still, so the transport radius directly determines freight as a proportion of goods value. The formation of the clusters at Yixing, Wuwei, Ningjin, Hejian and Humen is mutually causal with the constraint of transport radius: the local and surrounding market is enough to support scale, while distant markets are served by higher voltage classes, special-purpose products and exports to offset freight.
The separation of the conversion fee from the copper price is the foundation of the division of labour in this industry. The copper price part is "pass-through" in the financial statements: revenue rises when copper rises and falls when copper falls, with little relation to operating capability. The conversion fee part is what the enterprise actually earns. The separation has a clear confirmation in the data: in 2024 the copper price rose 9.69%, the main business revenue of 80 listed cable companies grew 7.78% and net profit fell 25.71% (industry operating analysis republished on the website of the Zhejiang Provincial Administration for Market Regulation). Revenue followed the copper price; profit did not.
Pressure from capacity utilisation on the conversion fee is supported by public data. A report on China Jiangsu Network states that the manufacturing output of high-voltage and extra-high-voltage cable is about 2 times market demand, and that of mid- and low-end products reaches 3 times (report on China Jiangsu Network). If manufacturing output is twice demand, this shows only that output is in excess relative to demand; capacity utilisation will be lower still, and it cannot be converted directly into half the equipment standing idle. The same holds for the threefold case. In this structure the pricing power over the conversion fee lies with the buyer, especially for mid- and low-end products.
The comparability of conversion fees has to be examined item by item. Of two factories' quoted conversion fees, one may include freight and the other not, one may include packing and the other not, one may include factory testing and the other not, and one may include 13% VAT while the other is a pre-tax price. Once these items are separated out and aligned, the gap is often smaller than it appears. The first thing a buyer should do on receiving a quotation is to establish clearly the boundary of the conversion fee column.
The separation of the copper price from the conversion fee is the starting point for understanding every price phenomenon in this industry. The copper price determines the absolute level of operating revenue, the conversion fee determines profit, and capacity utilisation determines the conversion fee. Only by linking the three can one explain why industry scale is growing while profit is falling.
Chapter 20 Contract Risk When Raw Materials Rise in Price
When the copper price runs at a high level, the risk in a cable purchase contract concentrates at four positions: price benchmark, price revision mechanism, delivery schedule and quality acceptance. The clauses corresponding to these four positions can all be written very specifically; a formulation such as "strengthen risk management" amounts to writing nothing.
The first position is the price benchmark. Writing "settled at market price" in a contract amounts to no agreement at all. Market price has three sources — Yangtze River non-ferrous spot, SHFE futures and LME — and the three are not of equal value on the same trading day. Writing "settled at the copper price" likewise leaves the period undefined, and the copper price on the quotation date, the signing date and the delivery date may differ by several thousand yuan per tonne.
An executable formulation sets out all three of object, period and algorithm. Example:
"The benchmark copper price under this contract shall be the arithmetic mean of the settlement prices of all trading days from 【 】 Year 【 】 Month 【 】 Day to 【 】 Month 【 】 Day of the most-active contract for cathode copper futures on the Shanghai Futures Exchange. The conversion fee shall be RMB 【 】 per tonne, fixed and unchanged for the term of the contract."
The function of this clause is to compress uncertainty from a range into a single number. The object specifies the exchange and the contract, the period specifies the start and end dates, and the algorithm specifies the arithmetic mean. If any one of the three is left blank, dispute will follow.
The second position is the price revision mechanism. A fixed-unit-price contract is a disaster for one party under a one-way copper market: the seller loses when the copper price rises, the buyer loses when it falls. A two-way linkage formulation:
"Where the change in the benchmark copper price on the settlement date relative to the contract benchmark copper price is within ±3% (inclusive), the contract unit price shall not be adjusted; the part exceeding ±3% shall be adjusted on an actual basis by multiplying the net copper weight per unit length of conductor listed in the annex to this contract by the amount of the change."
This clause has three elements. The ±3% threshold is the band within which the two parties share normal fluctuation, and the figure can be adjusted to project scale; the measurement basis is "net copper weight per unit length of conductor", which must be listed in an annex, otherwise the adjustment cannot be calculated; and the direction of adjustment is two-way, applying to rises and falls alike. A clause that provides only for rises and not for falls becomes a loss for the buyer when the copper price declines.
The third position is the delivery schedule. The most common gap in an annual framework agreement is to agree only a total volume and not a monthly offtake volume. When the copper price falls the buyer delays taking delivery and the seller's inventory depreciates; when the copper price rises the buyer takes delivery early and the seller's capacity is stretched. A formulation:
"The annual framework volume shall be 【 】 tonnes, and the buyer's monthly offtake may fluctuate within ±20% of the agreed volume. Where delivery is delayed by more than 30 days for reasons attributable to the buyer, the delayed portion shall be re-priced at the benchmark copper price on the actual date of offtake; where delivery is delayed by more than 30 days for reasons attributable to the seller, the delayed portion shall be priced at the contract benchmark copper price, and the seller shall additionally pay liquidated damages at 【 】% of the value of the delayed portion."
This clause covers delay by either party. A one-sided constraint is hard to get signed in actual negotiation, whereas a two-way clause is easier to reach.
The fourth position is quality acceptance, whose importance rises when the copper price is high. The higher the copper price, the stronger the economic incentive to shrink the cross-section. In the 2024 national supervision sampling inspection, 85 batches failed on conductor resistance, and 19 batches in 2025, the top failure item in both years (2024 notice, 2025 notice). Over the same period the copper price rose 9.69% in 2024 and 7.52% in 2025 (industry operating analysis republished on the website of the Zhejiang Provincial Administration for Market Regulation, summary of Jiangxi Copper's 2025 annual report published by China Securities Journal). The two sets of numbers cannot be directly linked as cause and effect, but for a buyer to list conductor resistance as a mandatory inspection item while the copper price is high is a practice with grounds.
An executable formulation of the quality clause:
"The DC resistance of the conductor shall be measured drum by drum in accordance with GB/T 3956-2008, with limits applied as 【 】. If any drum fails, that drum shall be returned and compensation paid at 【 】 times the value of that drum; if the non-conformity rate within the same batch exceeds 5%, the whole batch shall be returned, and the seller shall bear the transport and testing costs arising thereby."
Writing "drum by drum" into the contract is the key. Sampling inspection of drums has low coverage in practice, and the scatter of conductor resistance is enough for sampling to miss the problem drums. The cost of measuring every drum falls on the buyer, but it is the only way to stop a shrunken cross-section at the goods-receipt stage.
Beyond these four, two more are worth adding.
One is an abnormal-fluctuation clause. When the copper price moves beyond a certain amplitude within a single month, no formula can cover it, and a channel for renegotiation must be left:
"Where the benchmark copper price moves by more than 15% within a single month (the threshold may be adjusted to product category and project scale), either party may propose consultation in writing within 5 working days, and the consultation period shall not exceed 15 days. If consultation fails, either party may terminate the unperformed part, neither party bearing liability for breach, and the performed part shall be settled on an actual basis."
The value of this clause lies in providing an exit mechanism. A price-linkage clause without an exit mechanism will be unilaterally delayed in execution by one party under extreme market conditions.
The other concerns the handling of inspection disputes:
"Where the parties dispute an inspection result, the sample shall be sent to 【 】 testing body for re-inspection, whose result shall be final, and the cost shall be borne by the party responsible."
The testing body must be designated at signing, not found once a dispute has arisen. When designating it, confirm that it has the testing capability for GB/T 3956 and the GB/T 12706 series, and that the scope of its accreditation covers the items in dispute.
Beyond these six clauses there is one contractual boundary easily overlooked: which document prevails when the quotation and the contract disagree. In practice a quotation often carries a line such as "this quotation is valid for 7 days" or "prices float with the copper price"; if these conditions are not written into the body of the contract at signing, the contract prevails in a dispute and the limiting clauses on the quotation lose effect. Conversely, if the floating clause on the quotation is written into the contract, it becomes a contract term and must set out object, period and algorithm clearly in the manner described above.
In contract risk management for rising raw material prices, the point of application is the calculability of clauses. One clause from which an amount can be calculated is worth more than ten statements of principle.
Part Four: Access, Quality and Sampling Inspection
Chapter 21 Production Licensing and China Compulsory Certification (CCC)
Access for wire and cable is administered by two separate regimes: the industrial product production licence and China Compulsory Certification (CCC). Their coverage does not overlap: one governs the qualification to produce, the other the admission of products, and both define their scope by product category.
The current basis for the production licence is the Decision of the State Council on Adjusting and Improving the Catalogue of Industrial Products Subject to Production Licence Administration (《国务院关于调整完善工业产品生产许可证管理目录的决定》, Guofa [2024] No. 11). The catalogue contains 14 categories and 27 varieties, with wire and cable listed as the sixth category (Guofa [2024] No. 11, 2024-05-09; catalogue note, 2026-04-16). The implementing authority is the provincial department responsible for industrial product production licences, and the decision also specifies that the relevant approval powers may not be delegated. The scope of licence issuance is divided by product unit; the rules in force before Announcement No. 49 of 2025 listed five units: overhead stranded conductor, plastic-insulated control cable, extruded-insulation low-voltage power cable, extruded-insulation medium-voltage power cable and overhead insulated cable (Fangchenggang Municipal Administration for Market Regulation, 2025-04-17); the new rules took effect on 1 April 2026, and the product units are as set out in the annex to the announcement of the State Administration for Market Regulation. An administrative reconsideration document describes the scope as follows: "wire and cable products subject to the industrial product production licence system include only … 5 product units" (administrative reconsideration document, 2025-05-30).
The boundaries of the five units must be read word by word. Power cable covers only low and medium voltage; 110 kV and above falls outside the production licence. Overhead stranded conductor and overhead insulated cable are two separate units, and bare conductor and overhead insulated wire cannot substitute for each other. Building wire is not among the five units. Announcement No. 49 of 2025 of the State Administration for Market Regulation published the implementation rules for production licences covering 24 products, effective from 1 April 2026 (Announcement No. 49 of 2025, 2025-12-09); certificate renewal and the addition of units follow the new rules.
CCC covers far less. The current catalogue is based on Announcement No. 18 of 2020 of the State Administration for Market Regulation: 17 broad categories and 103 products, with only three listed under the wire and cable category — cord sets (0101), rubber-insulated wires and cables for rated voltages of 450/750 V and below (0104), and PVC-insulated wires and cables for rated voltages of 450/750 V and below (0105) (Announcement No. 18 of 2020, 2020-04-21). Announcement No. 9 of 2024 brought flame-retardant wire and cable into the 0104 and 0105 categories: designated certification bodies began accepting applications on 1 July 2024, and from 1 July 2025 flame-retardant wire and cable that has not obtained certification and bears the CCC mark may not leave the factory, be sold or be imported (Announcement No. 9 of 2024, 2024-04-07).
A dividing line therefore emerges: CCC governs building wire and cord sets at 450/750 V and below, including flame-retardant products within these two categories; power cable, control cable, communication cable and optical fibre cable are not in the CCC catalogue. A common formulation in public reporting is that "wire and cable must pass CCC certification", which extends a requirement for building wire to all products. Checking the catalogue against the specific product at the time of purchase is more reliable than relying on a supplier's verbal assurance.
Table 1 Coverage of the Two Access Regimes Compared (table compiled from public documents)
| Item | Industrial product production licence | China Compulsory Certification (CCC) |
|---|---|---|
| Basis | Guofa [2024] No. 11 | Announcement No. 18 of 2020 of the State Administration for Market Regulation |
| Products covered | Overhead stranded conductor, plastic-insulated control cable, extruded-insulation low-voltage power cable, extruded-insulation medium-voltage power cable, overhead insulated cable | Cord sets, rubber-insulated wires and cables at 450/750V and below, PVC-insulated wires and cables at 450/750V and below |
| Flame-retardant products | Administered by product unit and implementation rules | Flame-retardant products in the 0104 and 0105 categories, mandatory from 2025-07-01 |
| Implementing authority | Provincial production licence authority | Designated certification bodies |
| Not covered | Power cable at 110kV and above, building wire, communication cable, optical fibre cable | Power cable, control cable, communication cable, optical fibre cable |
The "Not covered" row is the one most easily skipped when checking at the time of purchase. Procurement of medium-voltage power cable must check the product unit and voltage ceiling on the production licence; procurement of building wire must check the CCC certificate and mark; for high-voltage and extra-high-voltage cable, neither regime applies, and the constraints come from product standards, type tests and tender documents, as discussed in Chapter 26.
An enterprise producing both building wire and low-voltage power cable must hold both qualifications. Building wire requires a CCC certificate, low-voltage power cable requires the low-voltage power cable unit on the production licence, and the two types of certificate cannot substitute for each other. When tender documents require "provision of a production licence", it is necessary to check whether the product unit on the certificate covers the product being bid; an enterprise holding only a CCC certificate is not qualified to produce medium-voltage power cable. Conversely, an enterprise holding only a production licence is equally non-compliant in producing 450/750 V building wire.
Certificate status also has a temporal dimension. In the 2025 national supervision sampling inspection, 2 batches of products were suspected of being produced without a CCC certificate, 1 batch did not bear the CCC mark and 1 batch's certificate had been revoked (2025 national supervision sampling inspection notice, 2026-03-23). All these cases involved products within the catalogue, which shows that enforcement gaps remain in the mandatory requirements. When checking a certificate, it is necessary to see whether the certificate number, model and specification, and voltage class cover the product purchased, and then whether the certificate is in valid status.
The boundary of the access regime is drawn by product category; the first step in procurement checking is to establish which regime the product falls under.
Chapter 22 The Standard System: GB/T 12706, GB/T 5023 and Industry Standards
The function of a product standard is to turn "conforming" into a set of measurable quantities. The wire and cable standard system has three layers: product standards specify the model, construction, performance and test items; method standards specify how to test; and material standards specify the indices for insulation compounds, sheath compounds and conductor materials. Procurement contracts cite product standards, while acceptance applies method standards, and the two must be used as a pair.
Product standards for power cable are divided by voltage. The GB/T 12706 series covers extruded-insulation power cable from 1 kV to 35 kV: Part 1 corresponds to 1 kV and 3 kV (GB/T 12706.1-2020, 2020-03-31), Part 2 to 6 kV to 30 kV and Part 3 to 35 kV, a division stated explicitly in the earlier national supervision sampling inspection specification (CCGF 708.1—2015) (CCGF 708.1—2015, 2015-04-29). The 66 kV and 110 kV levels are covered by a separate set of standards: the product standard is GB/T 11017.2-2024 and the test methods follow GB/T 11017.1-2024, both issued on 29 September 2024 and effective from 1 April 2025 (GB/T 11017.1-2024, 2024-09-29). At higher voltage levels, product standards are revised more slowly; the draft revision for 220 kV was still at the drafting stage in January 2026 (standard draft, 2026-01-27).
Standards for building wire are listed by insulating material. The GB/T 5023 series covers PVC insulation in 7 parts, all listed as current on the catalogue page, with Parts 1 to 5 and Part 7 in the 2008 edition and Part 6 in the 2006 edition (GB/T 5023 standard catalogue, 2008-06-30). Rubber insulation follows the GB/T 5013 series, Part 1 in the 2008 edition, effective from 1 September 2008 (GB/T 5013.1-2008, 2008-01-22). The product scope of both sets is limited to 450/750 V and below, corresponding to categories 0104 and 0105 of the CCC catalogue.
Control cable, communication cable and optical fibre cable each have their own standards. Control cable is GB/T 9330-2020, effective from 1 October 2020 (GB/T 9330-2020, 2020-03-31). Local communication cable has just completed a revision: GB/T 13849-2025 was issued on 5 October 2025 and takes effect on 1 May 2026 (GB/T 13849-2025, 2025-10-05), and the previous GB/T 13849.1-2013 is repealed on the same date (previous edition page, 2013-07-19). Outdoor optical fibre cable is GB/T 7424.3-2025, effective from 1 February 2026 (GB/T 7424.3-2025, 2025-10-05).
Flame retardance and fire resistance have one umbrella standard. GB/T 19666-2019 specifies the code designation system, classification and test methods, effective from 1 July 2020 (GB/T 19666-2019, 2019-12-10). This standard does not replace product standards; it is an additional set of requirements layered on top of them, and the specific classes are set out in Chapter 24. Cable for charging electric vehicles is covered by GB/T 33594-2025, effective from 1 May 2026 (GB/T 33594-2025, 2025-10-05).
Table 1 Current Versions of the Main Product Standards (table compiled from public standard information)
| Product | Standard number | Status and effective date |
|---|---|---|
| Low-voltage and medium-voltage power cable | GB/T 12706.1-2020 | Current, 2020 edition |
| High-voltage power cable | GB/T 11017.2-2024 (product), GB/T 11017.1-2024 (test methods) | Current, effective 2025-04-01 |
| Building wire (PVC insulation) | GB/T 5023 series | All seven parts listed as current |
| Building wire (rubber insulation) | GB/T 5013.1-2008 | Current, effective 2008-09-01 |
| Control cable | GB/T 9330-2020 | Current, effective 2020-10-01 |
| Local communication cable | GB/T 13849-2025 | Current, effective 2026-05-01 |
| Outdoor optical fibre cable | GB/T 7424.3-2025 | Current, effective 2026-02-01 |
| General standard for flame retardance and fire resistance | GB/T 19666-2019 | Current, effective 2020-07-01 |
| Cable for charging electric vehicles | GB/T 33594-2025 | Current, effective 2026-05-01 |
Industry standards supplement national standards for specific categories. The 2025 edition of the Yueqing municipal implementation rules for supervision sampling inspection of wire and cable products cites GB/T 5023.3 and GB/T 5023.5 alongside JB/T 8734.2 and JB/T 8734.3 among its bases for judgement (Yueqing 2025 edition implementation rules, 2025-09-24). The JB/T 8734 series covers some categories of PVC-insulated wires and cables for rated voltages of 450/750 V and below, and specifies sizes that national standards do not list separately. Enterprise standards may be stricter than national and industry standards, but may not be looser than mandatory requirements.
A reference that omits the year from a standard number automatically points to the new edition when the standard is revised. If the contract says "GB/T 13849", cable produced after 1 May 2026 must be assessed against the 2025 edition, and reports under the old edition cease to be valid; if the contract says "GB/T 13849.1-2013", it still cites a repealed standard after the new edition takes effect. Neither wording is appropriate; the correct approach is to state the full designation of the current edition and agree how a revision of the standard will be handled.
For higher voltage levels, standards update out of step with product iteration. The product standard for 110 kV cable was revised only in 2024, the draft revision for 220 kV was still being drafted in January 2026, and above 35 kV there is no production licence constraint. In procuring such products, the correspondence among the standard edition, the type test report and the tender documents matters more than whether the standard itself is strict or lenient.
A recommended standard is not mandatory in itself, but once written into a contract it becomes binding. GB/T standards are all recommended standards; if a procurement contract states "supply in accordance with GB/T 12706.1-2020", all the test items of that standard become the basis for acceptance; if the contract says only "complies with national standards", the first task in a dispute is to determine which edition applies. A superseded standard is no longer the basis for acceptance even if products conforming to it are still in circulation. The time lag in a standard revision is another risk point: GB/T 13849 had a seven-month transition from issue in October 2025 to implementation in May 2026, during which products to either edition may be in production, and the contract should state the standard number and year.
The text of standards is publicly available. The National Standards Full-Text Public System provides online access, and the standard number, issue date, effective date and replacement relationship can all be checked on the standard page. Three things should be checked: whether the standard number carries a year, whether the status is current, and whether the superseded standard has been repealed. Only when all three check out is the citation a valid edition. When a procurement document cites a standard, writing the standard number, year and applicable model on the same line leaves much less room for later dispute.
A standard number plus the year is the smallest unit in a contract; a citation without the year cannot be tied to an edition when the standard is revised.
Chapter 23 Common Non-Conformities in Market Supervision Sampling Inspection
The State Administration for Market Regulation organises a national supervision sampling inspection of product quality every year, and wire and cable is one of the categories always sampled. The 2025 results are contained in the Notice on the National Supervision Sampling Inspection of the Quality of 24 Products including Indoor Heaters in 2025 (《2025年室内加热器等24种产品质量国家监督抽查情况通报》, Issue No. 24 of 2026), issued on 23 March 2026 (2025 national supervision sampling inspection notice, 2026-03-23). The 2024 results are contained in the Notice on the National Supervision Sampling Inspection of the Quality of 27 Products including Lithium-ion Batteries and Battery Packs for Electronic Products in 2024 (《2024年电子产品用锂离子电池和电池组等27种产品质量国家监督抽查情况通报》) (2024 national supervision sampling inspection notice, 2025-04-08). The two notices have different sampling structures, so the non-conformity rates cannot be compared directly, though the ranking of item categories can.
In 2025 a total of 2,204 batches were sampled. The production segment covered 1,763 batches from 1,761 production entities in 31 provinces, the distribution segment covered 161 batches from 158 sellers in 31 provinces, and the online-sales segment covered 280 batches from 259 sellers on 8 platforms. 65 batches were found non-conforming, a non-conformity rate of 2.95% (65 ÷ 2,204, calculated in this article).
In 2024 a total of 3,464 batches were sampled. The production segment covered 2,320 batches from 2,320 production entities and the distribution segment 1,144 batches from 1,046 sellers. 158 batches were found non-conforming, a non-conformity rate of 4.56% (158 ÷ 3,464, calculated in this article). The dedicated online-sales sampling inspection in the same year was notified separately: 272 batches from 255 sellers on 8 platforms, with 41 batches found non-conforming, a non-conformity rate of 15.07% (41 ÷ 272, calculated in this article) (2024 online-sales sampling inspection notice, 2025-04-08).
Table 1 Overall Results of Three Sampling Inspections (table compiled from public notices; non-conformity rates calculated in this article)
| Inspection | Batches sampled | Non-conforming batches | Non-conformity rate |
|---|---|---|---|
| 2025 national supervision sampling inspection | 2,204 | 65 | 2.95% |
| 2024 national supervision sampling inspection | 3,464 | 158 | 4.56% |
| 2024 dedicated online-sales sampling inspection | 272 | 41 | 15.07% |
The 15.07% non-conformity rate in the 2024 dedicated online-sales sampling inspection is 3.3 times the 4.56% of the contemporaneous offline (production plus distribution) national supervision sampling inspection. The gap indicates a higher risk at the online-sales end; the notice does not give non-conforming batch counts separately for the production and distribution segments, so it cannot be used to judge whether the manufacturing end or the distribution end is higher. Operators on online-sales channels are mostly sellers, with mixed sources, small batches and fast turnover, which makes sampling coverage costly.
The sampling structures of the two inspections differ greatly. The distribution segment sampled 1,144 batches in 2024, falling to 161 batches in 2025, while 280 batches in the online-sales segment were added. The focus shifted from offline distribution to online sales, and the offline distribution sample shrank to one seventh of its former size, so a direct comparison of the two years' non-conformity rates does not hold. What is comparable is the ranking of item categories: conductor resistance ranked first in both years.
Looking at the item categories of non-conformity, the first place was identical in both years. Among the 65 non-conforming batches in 2025, conductor resistance ranked first with 19 batches, insulation tensile strength after ageing second with 15 batches, followed by insulation hot set with 11 batches, sheath thickness with 9 batches, insulation tensile strength before ageing with 8 batches, insulation thickness with 7 batches, sheath loss of mass with 6 batches and insulation shrinkage with 5 batches (2025 notice, 2026-03-23). The eight items sum to 80, higher than the 65 non-conforming batches, because a single batch can fail more than one item at the same time.
Among the 158 non-conforming batches in 2024, conductor resistance ranked first with 85 batches and the insulation tensile test after ageing second with 46 batches; insulation thickness and insulation tensile strength before ageing tied for third with 36 batches each, followed by sheath thickness with 27 batches, insulation hot set with 23 batches, bunched cable flame retardance with 7 batches and the single-cable flame test with 4 batches (2024 notice, 2025-04-08). The item counts sum to more than the number of non-conforming batches because a single batch can fail several items at once.
Table 2 Non-Conforming Batches by Item in the Two Inspections (table compiled from public notices)
| Non-conforming item | 2024 | 2025 |
|---|---|---|
| Conductor resistance | 85 | 19 |
| Insulation tensile test after ageing | 46 | 15 |
| Insulation thickness | 36 | 7 |
| Insulation tensile test before ageing | 36 | 8 |
| Sheath thickness | 27 | 9 |
| Insulation hot-set test | 23 | 11 |
| Sheath loss-of-mass test | 5 | 6 |
| Insulation shrinkage test | 1 | 5 |
| Bunched cable flame retardance | 7 | Not listed in the notice |
| Single-cable flame test | 4 | Not listed in the notice |
Table note: for 2024 there were in addition 14 batches for sheath tensile strength after ageing and 9 batches for flexing, plus a combined 15 batches for three items — insulation tensile strength after air-bomb ageing, sheath tensile strength before ageing and sheath hot set; the item counts sum to more than the number of non-conforming batches because a single batch can fail several items at once.
Conductor resistance ranked first for two consecutive years, and this signal is more valuable than the non-conformity rate. A high conductor resistance has two sources: a conductor cross-section smaller than nominal, or insufficient copper purity with a high impurity content. Both point to the same thing: the amount of copper per unit length is below the design value. The direct consequence of high resistance is a rise in operating temperature and a fall in current-carrying capacity, and long-term operation accelerates insulation ageing. This is a failure path that can cause fire, and its severity is higher than non-standard marking.
Insulation tensile strength after ageing ranked second, with 46 batches in 2024 and 15 batches in 2025; insulation thickness had 36 batches in 2024 and 7 batches in 2025. The earlier national supervision sampling inspection specification classified the minimum insulation thickness as a Class A item and the average insulation thickness as Class B (CCGF 708.1—2015, 2015-04-29). The thinnest point is the weak point where breakdown occurs, and insufficient thickness directly reduces the voltage withstand margin. Insulation tensile strength after ageing and insulation hot set point to material formulation and degree of cross-linking: a fall in tensile strength after ageing indicates insufficient resistance to thermo-oxidative ageing, and a hot-set value above the limit indicates an insufficient degree of cross-linking.
Violations of law and regulation are counted separately. The 2025 notice lists 4 batches of counterfeit goods, 2 batches suspected of production without a CCC certificate, 3 batches with non-conforming core colours, 1 batch with a revoked certificate, 1 batch without the CCC mark and 1 batch whose sample body bore another enterprise's information, a total of 12 batches across six items. The 2024 notice lists 2 batches suspected of counterfeiting, 12 batches suspected of production without a CCC certificate, 6 batches with a manufacturer name and address that does not exist and 4 batches without a manufacturer name and address. A problem with the manufacturer name and address cuts the traceability chain directly, as discussed in Chapter 27.
What the sampling data supports is the judgement that conductor resistance and insulation tensile strength after ageing are the two main points at which wire and cable quality fails.
Chapter 24 Classification Tests for Flame Retardance and Fire Resistance
Flame retardance and fire resistance are two different groups of performance, differing in test methods, classification and acceptance indices. Flame retardance means that the flame does not continue to spread after the material is removed from the ignition source; fire resistance means that the cable can maintain circuit continuity for a period under fire conditions. GB/T 19666-2019 combines both groups in a single general standard and distinguishes them by code designation (GB/T 19666-2019, 2019-12-10).
The code designation system is built from letters: Z for flame retardant, W for halogen-free, D for low smoke, U for low toxicity, N for fire resistant, J for mechanical impact and S for water spray. Codes can be combined, for example WDZ for halogen-free low-smoke flame retardant and NH for fire resistant. The order of letters in a product model and the meaning of the codes must be checked against the standard; a supplier's verbal explanation cannot serve as the basis.
Flame retardance has two routes, single cable and bunched cable. Single-cable flame retardance is tested to GB/T 18380.12 and GB/T 18380.13, assessing whether one insulated wire extinguishes itself after the flame is removed. Bunched-cable flame retardance is tested to GB/T 18380.33 to GB/T 18380.36, with classes A, B, C and D corresponding to different volumes of non-metallic material per metre: class A 7 litres per metre, class B 3.5 litres per metre, class C 1.5 litres per metre and class D 0.5 litres per metre, with the acceptance criterion that the charred height does not exceed 2.5 metres. The test conditions of the two differ greatly: the single-cable test applies a small fire load, while the bunched test simulates mutual ignition when cables are laid densely together.
Passing the single-cable test cannot substitute for passing the bunched test, and this is the most common misuse in procurement. Basements, risers and cable trays in a building are cable-dense locations; where the design documents require bunched class B or class C, a single-cable flame-retardant test report has no probative value. In the 2024 national supervision sampling inspection, 7 batches failed bunched cable flame retardance and 4 batches failed the single-cable flame test (2024 notice, 2025-04-08), which shows that both types of test have failures.
Class A has the largest number of test pieces and the highest fire load, and class D the smallest. Design documents choose the class according to the cable density of the location: underground substations, tunnels and risers usually require a higher class, while surface wiring in ordinary rooms may use a lower class. Writing the class required by the design into the contract at the time of purchase is more useful than the two words "flame retardant".
Fire resistance has three grades: N for fire resistance alone, NJ for fire resistance plus mechanical impact, and NS for fire resistance plus mechanical impact plus water spray (GB/T 19666-2019). The three grades correspond to different installation environments. Circuits such as fire pumps, smoke-extraction fans and emergency lighting are usually designed to require NJ or NS; cable meeting only grade N may fail in a fire after impact from falling objects or spraying with fire water.
Halogen-free, low smoke and low toxicity are three independent indices and cannot substitute for one another. The determination of halogen-free includes a pH of the combustion gases not lower than 4.3, a conductivity not higher than 10 microsiemens per millimetre, and hydrogen halide content: HCl and HBr together not more than 0.5%, and HF not more than 0.1% (GB/T 19666-2019, with test methods in GB/T 17650.2). Low smoke is determined by light transmittance, with a minimum transmittance of 60% where the test-piece outer diameter is greater than 40 millimetres (test method GB/T 17651.2). The recommended index for low toxicity is a toxicity index ITC of not more than 5, with a critical carbon monoxide concentration of 1750 milligrams per cubic metre.
Table 1 Classification and Acceptance of Flame Retardance and Fire Resistance (table compiled from GB/T 19666-2019)
| Item | Classification | Acceptance index | Test method |
|---|---|---|---|
| Single-cable flame retardance | Not classified | Self-extinguishing after removal from the flame | GB/T 18380.12, GB/T 18380.13 |
| Bunched-cable flame retardance | A, B, C, D | Non-metallic material volume 7, 3.5, 1.5, 0.5 litres per metre; charred height not exceeding 2.5 metres | GB/T 18380.33 to GB/T 18380.36 |
| Fire resistance | N, NJ, NS | Circuit continuity maintained under fire conditions | GB/T 19666-2019 |
| Halogen-free | Not classified | pH ≥ 4.3, conductivity ≤ 10 μS/mm, HCl + HBr ≤ 0.5%, HF ≤ 0.1% | GB/T 17650.2 |
| Low smoke | Not classified | Minimum transmittance 60% where outer diameter is greater than 40 mm | GB/T 17651.2 |
| Low toxicity | Not classified | Toxicity index ITC ≤ 5 | GB/T 19666-2019 |
There is a further relationship between the indices. To achieve flame retardance, halogen-free compounds usually require large amounts of inorganic flame-retardant filler; the filler lowers the mechanical and processing performance of the material, so the sheath is made thicker or a cross-linkable material is used instead. Low smoke and halogen-free are often required together, but low smoke assesses smoke density while halogen-free assesses the corrosivity of the combustion gases; when a report tests only one of them, the other has no conclusion.
How the procurement document is worded determines whether acceptance can be carried out. A requirement written as "flame-retardant cable" is insufficient for acceptance; the code combination, single or bunched, bunched class and fire-resistance grade should be stated, for example "WDZBN-YJY23, bunched class B, grade NS". The codes in the model must correspond one to one with the test items in the test report. The test report must be read for the test category and the test-piece outer diameter; test pieces of different outer diameters cannot be covered by the same bunched-test conclusion.
When submitting samples, four things must be stated clearly on the commissioning form: the full model, the code combination, the bunched class or fire-resistance grade, and the test-piece outer diameter. The testing body determines the number of test pieces for the bunched test from the outer diameter; different outer diameters mean different test conditions, and reports cannot substitute for one another. In the 2025 national supervision sampling inspection, 11 batches failed the insulation hot-set test (2025 notice, 2026-03-23); hot set assesses the degree of cross-linking (the extent to which polyethylene molecular chains change from linear to network form) and belongs to a different dimension from flame-retardant class, so a flame-retardant report cannot prove cross-linking quality.
Every index of flame retardance and fire resistance has its own test method; acceptance requirements must go as far as the classification, and stopping at "flame retardant" is equivalent to having no requirement.
Chapter 25 Identifying Inferior Cable: Copper Core, Cross-Section, Insulation Thickness
Inferior cable can be identified through four actions that can be carried out on site: look at the copper core, measure the cross-section, measure the insulation thickness and check the marking. None of the four requires a laboratory, only a micrometer, a steel rule and a section of sample. They cannot replace type tests, but they are enough to screen out most obviously non-conforming products at goods receipt.
The first action is to look at the colour and lustre of the copper core. Conforming annealed copper is purplish red, with a metallic lustre at the fracture surface and an even colour. A whitish, yellowish or greyish colour, or obvious oxidation spots, slag inclusions or burrs on the section, indicates a problem with copper purity or the annealing process. If the surface of the copper core is black and cannot recover its metallic lustre after wiping, this is usually related to oxidation or impurities. If the section is silvery white and remains silvery white on the inner layer after scraping the surface, copper-clad aluminium should be suspected; the resistivity of aluminium is about 1.64 times that of copper (China Electrical Equipment Industry Association material), and at the same cross-section the current-carrying capacity is clearly lower than that of a copper conductor. This judgement is empirical and can serve only as an initial screen: a copper core of normal colour may still have an insufficient cross-section, while a copper core of abnormal colour should be sent directly for testing.
The second action is to measure the cross-section. For a single-strand conductor, measure the wire diameter with a micrometer and calculate from the circle-area formula, the cross-section being pi times the square of the diameter divided by four. For a stranded conductor, measure the wire diameter and count the number of wires in the same way, then multiply by the number of wires. Taking a single-strand wire of nominal 2.5 square millimetres as an example, if the measured wire diameter is 1.70 millimetres, the calculated cross-section is 2.27 square millimetres, about 9% below the nominal value. The number of wires in a compacted stranded conductor is not easy to count, so an electrical method is used instead: measure the DC resistance of a length of conductor with a DC resistance tester, convert it to 20 degrees Celsius and compare it with the limit for that size. Conductor resistance is a Class A item in national supervision sampling inspection, with 85 non-conforming batches in 2024 and 19 in 2025, ranking first in both years (2025 notice, 2024 notice).
The third action is to measure the insulation thickness. Insulation thickness must be measured at the thinnest point, not at a single point. Using a reading microscope or an insulation thickness gauge, take several points evenly around the circumference of the test piece and compare the minimum value with the nominal value. The earlier national supervision sampling inspection specification classified the minimum insulation thickness as a Class A item and the average insulation thickness as Class B (CCGF 708.1—2015, 2015-04-29). Measuring only the average thickness misses eccentricity; where eccentricity is severe, the average thickness may meet the requirement while the thinnest point no longer does. The same applies to sheath thickness: in the 2025 sampling inspection, 9 batches failed sheath thickness.
The fourth action is to check the marking. Marking includes the CCC mark, manufacturer name and address, model and specification, rated voltage, date of manufacture and metre marking. The CCC mark is meaningful only for products in categories 0104 and 0105; the absence of a CCC mark on power cable, control cable, communication cable and optical fibre cable is normal and cannot be taken as evidence of counterfeit goods. The manufacturer name and address must lead to an identifiable entity; in the 2024 sampling inspection, 6 batches bore a manufacturer name and address that does not exist and 4 batches bore no manufacturer name and address (2024 notice). Metre markings must be continuous and the length of a whole drum must match the nominal length; skipped or repeated metre numbers are a sign of rewinding and repacking. The model on the certificate of conformity must match the inkjet marking on the cable itself.
Beyond the four actions, there are several rough checks that can be made on site. The surface of the insulation and sheath should be smooth, free of pores and foreign particles, should spring back normally after being stretched by hand, and should not take a deep mark when scratched with a fingernail. If a section of sample is frozen in a refrigerator for several hours and then bent, cracking of the sheath indicates insufficient low-temperature performance; the Yueqing municipal implementation rules for supervision sampling inspection list low-temperature bending of insulation, low-temperature bending of sheath, low-temperature impact of finished cable and non-flame-propagation as inspection items (Yueqing 2025 edition implementation rules, 2025-09-24). If a small section of insulation is burned with a lighter and the sample continues to burn after removal from the flame, it should be sent for testing; this rough check cannot replace the single-cable flame test.
Table 1 The Four On-Site Identification Actions and Their Corresponding Indices (table compiled in this article)
| Action | Tool | Object observed | Corresponding inspection item |
|---|---|---|---|
| Look at the copper core | Visual | Colour, lustre, fracture surface | Copper purity and annealing quality; no separate inspection item |
| Measure the cross-section | Micrometer, DC resistance tester | Wire diameter, number of wires, conductor resistance | Conductor resistance (Class A) |
| Measure the insulation | Reading microscope, thickness gauge | Thinnest point of insulation, sheath thickness | Minimum insulation thickness (Class A), sheath thickness (Class B) |
| Check the marking | Visual, tape measure | CCC mark, manufacturer name and address, metre marking | Marking and traceability requirements |
The limits of the four actions must be stated clearly. They can detect only two classes of problem, undersizing and missing marking, and cannot detect internal defects such as material formulation, degree of cross-linking and partial discharge. The insulation hot-set test reflects the degree of cross-linking and the sheath loss-of-mass test reflects the formulation, and both require a laboratory; in the 2025 sampling inspection these two items failed in 11 batches and 6 batches respectively. On-site identification serves to screen out obviously non-conforming batches and cannot replace incoming inspection.
The four actions can identify only cut corners; formulation and process defects require laboratory testing.
Chapter 26 Qualification and Track-Record Thresholds in Tendering
Qualification review of cable suppliers in tendering is in substance the application of the two access regimes of Chapter 21 and the standard system of Chapter 22 to a single set of documents. The review can be grouped into four categories: legal-entity status, product qualifications, test reports and track record.
Legal-entity status is checked against the business licence and the production address. The bidder's name and production address must match the enterprise name and address on the certificates. In cases of commissioned processing and own-brand labelling (OEM), the authorisation documents and the ownership of the certificates must be examined; in the 2025 national supervision sampling inspection, 1 batch had another enterprise's information marked on the body of the sample (2025 notice, 2026-03-23), and such products point to a different enterprise when traced.
Product qualifications are checked by product unit. The production licence must cover the unit and voltage class of the product being bid; the five units are overhead stranded conductor, plastic-insulated control cable, extruded-insulation low-voltage power cable, extruded-insulation medium-voltage power cable and overhead insulated cable (Fangchenggang Municipal Administration for Market Regulation, 2025-04-17). Cable at 110 kV and above has no production licence, so the qualification review turns to type test reports and track record. Building wire must be checked against the CCC certificate, whose model and specification must cover the product being bid, and from 1 July 2025 flame-retardant building wire must also hold CCC certification (Announcement No. 9 of 2024, 2024-04-07).
Three common formulations of qualification requirements in tender documents cause misjudgement. The first requires CCC certificates for all cables, bringing power cable and control cable into the CCC scope. The second requires a production licence for cable at 110 kV and above, a voltage class outside the scope of licence issuance. The third says only "provide a test report", without stating the standard number, test items or validity period of the report. All three should be clarified in writing with the tendering party before bidding, and the materials should be prepared on the clarified basis.
When an agent bids, in addition to the agent's business licence it must provide an authorisation letter issued by the manufacturer, the manufacturer's production licence or CCC certificate, and the manufacturer's type test report for the corresponding model. The authorisation letter must state the authorised products and the authorisation period, and the scope of authorisation must cover the model being bid. Where the manufacturer and the bidder are different and the chain of authorisation cannot be provided, the qualification review should not be passed.
Three things are checked in a test report: the issuing body, the report date and the scope of coverage. A type test report should be issued by a testing body with the corresponding qualification, and the model and specification, voltage class and conductor cross-section in the report must correspond to the product being bid. Reports have a validity period, and an old report may cease to apply after a standard is revised; for example, GB/T 13849 was replaced from the 2013 edition by the 2025 edition on 1 May 2026 (GB/T 13849-2025, 2025-10-05). A report covers only the sample submitted, not all batches, and this gap must be filled by sampling inspection during contract performance.
Track-record thresholds are set by the tendering party and have no uniform values. The common dimensions are similar products, similar voltage class, similar project type and contract value. The review must check whether the contract and the acceptance certificate correspond; a track record with only a contract and no acceptance certificate has weaker probative force than one with both. Bidding for a medium-voltage lot with a track record in low-voltage products, or for a grid lot with a track record in building projects, both count as a mismatch between track record and lot.
Table 1 The Four Categories of Material in Qualification Review and the Points to Check (table compiled in this article)
| Category | Material | Points to check |
|---|---|---|
| Legal-entity status | Business licence, production address | Whether the bidder and the certificate holder are the same |
| Product qualifications | Production licence, CCC certificate | Whether the product unit, voltage class and model and specification are covered |
| Test reports | Type test report, third-party test report | Body's qualification, report date, models covered |
| Track record | Contracts, acceptance certificates | Similar products, similar voltage class, similar projects |
Price evaluation must leave room for copper price exposure. A cable quotation consists of two parts, the copper price and the conversion fee; the copper price follows the market, while the conversion fee is determined by the number of process passes and process complexity, as set out in Chapter 17 and Chapter 19. If bids from different bidders are compared directly at evaluation, then in a one-way copper market the low-price winner may seek a price adjustment or supply inferior goods during performance. The contract should state the copper price benchmark, the adjustment formula and the trigger conditions.
Constraints during performance come from sampling inspection and certificate status. In the 2025 national supervision sampling inspection, 1 batch's certificate had been revoked (2025 notice, 2026-03-23), and products produced while a certificate is revoked lack access qualification. The purchaser may provide in the contract that, where the supplier's licence or CCC certificate is revoked or suspended, the purchaser has the right to terminate the contract and require replacement. Incoming inspection items follow the acceptance procedure in Chapter 31, with emphasis on conductor resistance and insulation thickness.
The core of qualification review is the coverage relationship: the certificate covers the product, the report covers the model and the track record covers the lot; if any one does not cover, the qualification review fails.
Chapter 27 Quality Responsibility and Incident Traceability
Traceability of quality responsibility is held together by four things: marking, batch, records and documentary evidence. With all four in place, the production batch and the responsible entity can be located after an incident; if one is missing, traceability breaks at the gap.
Marking is the first link. The cable itself should bear the manufacturer's name, model and specification, rated voltage, date of manufacture and metre marking, and the packaging or certificate of conformity should state the factory address, the standard applied and the inspector. In the 2024 national supervision sampling inspection, 6 batches bore a manufacturer name and address that does not exist and 4 batches bore no manufacturer name and address (2024 notice, 2025-04-08), and these 10 batches could not be traced to a producer after an incident. In the 2025 sampling inspection there were 4 batches of counterfeit products (2025 notice, 2026-03-23); the marking on counterfeit products points to a different enterprise, and traceability is led in the wrong direction.
Batch is the second link. Cable is produced in batches, and the same batch uses the same lot of copper rod, the same lot of insulation compound and the same set of process parameters. The metre marking is the physical carrier of the batch: metre markings are continuous within one drum of cable, and different batches have a joint or a restart of the metre count at the junction. After an incident, the faulty section can be cut out and its metre marking traced back to the production drum number and date of manufacture. Recording the metre-marking range and drum number of every drum at the time of purchase is the only way to fix batch information.
Records are the third link. Factory test records should include the results of conductor resistance, insulation thickness, voltage tests and partial discharge tests; the partial discharge test is used to find voids and impurities inside the insulation. Factory test items follow the GB/T 12706 series; the earlier national supervision sampling inspection specification separately classified conductor resistance, the bending test and the subsequent partial discharge test, the dielectric loss tangent and the 4-hour voltage test as Class A serious items, and the judgement at sampling inspection is stricter where an item is missing or non-conforming, so the two are not the same list of items (CCGF 708.1—2015, 2015-04-29). Agreeing in the contract that factory test reports are provided with the goods is the way for the purchaser to close this link.
Documentary evidence is the fourth link. The procurement contract, invoice, delivery note and incoming acceptance record form the chain of civil liability. In construction projects, installation records, supervision records and acceptance records for concealed works must be added. Cable is laid underground or in trays and is hard to observe directly after an incident, so installation records are the basis for locating the faulty section.
Table 1 The Four Links in the Traceability Chain and the Consequences of Failure (table compiled in this article)
| Link | Carrier | Consequence if missing |
|---|---|---|
| Marking | Inkjet marking on the cable, certificate of conformity | Cannot locate the producer |
| Batch | Metre marking, drum number | Cannot locate production date and batch |
| Records | Factory test report | Cannot determine whether the product conformed at the factory |
| Documentary evidence | Contract, invoice, acceptance records | Cannot determine the transaction relationship and allocation of responsibility |
Allocation of responsibility follows roles. The producer is responsible for conformity of the product to standards, the seller for incoming goods inspection, the installer for installation quality and the user for operation and maintenance. An incident investigation must distinguish four situations: the product itself is non-conforming, installation damage, overload operation and an external ignition source. The responsible entity differs among the four, and so does the direction of evidence gathering. To establish that the product itself is non-conforming, the faulty section must be sent for testing, with analysis of conductor resistance, insulation thickness and material composition, and compared with the retained sample from the same batch.
Retaining a sample is the step most easily overlooked. At the time of purchase, a length of cable sufficient for a complete test should be retained per batch, sealed and recorded with its metre marking and drum number. After an incident, the retained sample can be compared with the faulty section to determine whether the problem is batch-wide or local damage. The retention period should cover the quality warranty period; in construction projects the quality warranty period and the product warranty period are two separate periods and should be stated separately in the contract.
The order of sampling after an incident affects the conclusion. First seal the faulty section and the retained sample from the same batch, then dismantle the joint, and finally send the samples for testing. If dismantling takes place before testing, the form of the fusion mark at the fracture is destroyed, and the determination of whether the short circuit was the cause or the result loses its basis. The nature of the fusion mark is determined by a qualified body; what the purchaser and the installer must do is keep the physical evidence intact and record the dismantling process.
The contractor's incoming acceptance is an independent link in the traceability chain. Incoming re-inspection should use witnessed sampling, with the sampling location, quantity and sealed-sample number recorded in the acceptance record, and the sample sent to a qualified testing body. A batch that fails re-inspection should be rejected from site in its entirety and must not be picked over for use; picking over lets non-conforming products flow into concealed works, and after an incident the scope of responsibility cannot be defined by batch.
There is a further practical difficulty for traceability. The distribution chain involves resale through multiple hands, and one drum of cable may pass through three levels of distributor, each of which may change the packaging and labels. The non-conformity rate in the online-sales sampling inspection was 15.07%, more than three times the non-conformity rate of the contemporaneous offline (production plus distribution) sampling inspection (2024 online-sales sampling inspection notice, 2025-04-08); the longer the channel, the more easily markings and documents are lost in the intermediate links. For the purchaser to shorten the channel and buy directly from the producer is itself to shorten the traceability chain.
Traceability capability is established at the procurement stage; after an incident it can only be used, not created retroactively.
Part Five: Equipment and Procurement Verification
Chapter 28 Main Equipment: Wire Drawing, Stranding, Extrusion, Cross-Linking, Cabling and Armouring
Wire and cable production lines are linked in process sequence. From the entry of copper rod and aluminium rod into the plant to the dispatch of finished drums, the main line consists of six processes, namely wire drawing, stranding, extrusion, cross-linking, cabling and armouring, while the sheath is completed in two passes, the bedding layer and the outer sheath. An equipment list can be used to infer a plant's product positioning: a plant with only drawing, extrusion, cabling and sheathing makes building wire and low-voltage power cable; adding a cross-linking line takes it into medium voltage; only with armouring and copper-tape screening added does high voltage come into question.
The boundaries of the six processes are set out clearly in a notice of proposed approval for an environmental impact assessment in Henan. The notice lists the processes of three product lines in a project with an annual output of 6 million metres (600 万米): aluminium stranded conductor as "aluminium rod—wire drawing—take-up—semi-finished product inspection—stranding—take-up—inspection—acceptance and warehousing"; YJV cable as "raw material—extrusion—take-up—cabling—sheath extrusion—coiling—packaging—acceptance and warehousing"; BV and BVR wire as "raw material—extrusion—take-up—inspection—coiling—packaging—acceptance and warehousing" (notice text). This is a combination of low-voltage building wire and aluminium stranded conductor, without cross-linking or armouring. The same document separates "drawing—stranding" and "extrusion—cabling—sheathing" into two lines, which corresponds exactly to the way the two groups of equipment are organised: the front section is organised around the conductor, and the rear section around insulation and sheathing.
Wire drawing comes first. Copper rod or aluminium rod is drawn down step by step through multiple dies in a drawing machine; the metal work-hardens in cold working and loses ductility, so annealing must be completed during or after drawing before stranding and extrusion can proceed. The industry calls in-line continuous annealing combined with drawing "continuous drawing and annealing"; the annealing power supply passes a heavy current directly through the wire, which is heated by its own resistance. Drawing requires a continuous supply of drawing lubricant for lubrication and cooling; the concentration and temperature of the bath affect the surface quality of the single wire and the life of the dies. If take-up tension is unstable, the single wire is wound unevenly on the drum and tangles readily during stranding. Machines are classified by inlet and outlet diameter as coarse, medium, fine and extra-fine drawing machines, and one multi-wire drawing machine can draw several wires at the same time. When inspecting equipment, a distinction must be made between in-line annealing and off-line pot annealing: the former runs in step with drawing and gives uniform wire properties; the latter loads wire in batches, has a long cycle, is prone to mechanical damage to the wire in handling, and the bare storage between annealing and stranding allows further oxidation.
Stranding addresses cross-section and flexibility. The cross-section of a single conductor is limited by drawing and take-up; large-section conductors are built up from many single wires stranded in a regular pattern, and stranding also reduces the skin effect under alternating current. The main machines are cage stranders and drum stranders, with tubular stranders for small cross-sections. Stranding has two process points: lay length and lay direction. The lay length is the distance advanced by a single wire in one revolution around the core; too long a lay leaves the conductor loose, while too short a lay increases wire consumption and reduces the effective conductive length of the conductor. Adjacent layers are stranded in opposite directions to prevent loosening between layers. Compacted conductors pass through a compacting die after stranding, pressing the round strands into a compacted circular or sector shape to reduce the finished diameter and raise the fill factor. Where the cross-section is larger still, a segmental conductor is used: the whole conductor is divided into several segments, each stranded separately and insulated from the others, then assembled in cabling, in order to suppress the skin effect. Equipment capability is judged on three points: the largest cross-section that can be stranded, the single-length capacity of the take-up and pay-off drums, and whether tooling for compaction and segmental conductors is available. The single length determines the number of joints; one more joint means one more jointing operation and one more test.
Capacity must be balanced once between the front and rear sections. The line speed of one cross-linking line corresponds to the output of several extruders and stranders; if any section among drawing, stranding, extrusion and cabling is narrow, the actual capacity of the whole line is set by the narrowest section. During a factory audit, list the number of machines, the line speeds and the shifts for each process and compare them with the declared capacity; the shortfall usually appears in stranding and cabling, because the equipment for these two processes is cheap per unit and easy to expand, and is therefore the most easily overlooked.
Extrusion is the forming process for insulation and sheath and the focus of quality control. The equipment is a screw extruder with a head and dies; the barrel is temperature-controlled in zones along the heating sections, and the length-to-diameter ratio and compression ratio of the screw are selected to suit the compound. Polyvinyl chloride (PVC), polyethylene, cross-linked polyethylene (XLPE) and low-smoke zero-halogen (LSZH) compounds differ in melt viscosity and processing window, so a change of compound on the same machine requires a change of screw or a reset of the temperature zones. Medium-voltage and higher cross-linked cables use triple-layer co-extrusion: conductor screen, insulation and insulation screen are extruded together in one head, and the interfaces between the three layers bond in the molten state, avoiding delamination and contamination. The two critical control variables in extrusion are thickness and eccentricity: thickness determines the insulation margin, and eccentricity determines where the thinnest point lies.
Cross-linking converts thermoplastic polyethylene into thermosetting cross-linked polyethylene. The equipment falls into two types, vertical tower and catenary, and the difference between them in capacity and quality is covered in Chapter 29.
Cabling is used for multicore cables. Insulated cores are stranded into a cable in a regular pattern, filler ropes are added in the interstices to round out the core, and tapes are wrapped around the outside to bind it. The cabling lay length and lay direction affect bending performance and the difference in core lengths, and sector-shaped cores also require control of phase sequence and of the direction in which they are turned. After cabling the core passes to armouring or sheathing.
Armour falls into two kinds, steel tape and steel wire, usually distinguished by digits in the product designation: YJV22 is steel-tape armoured, YJV32 and YJV42 are steel-wire armoured, and the position of the digit corresponds to the combination of armour layer and outer sheath. Steel-tape armour resists lateral pressure and suits direct burial; steel-wire armour resists tension and suits shafts, underwater laying and routes with large drops. A bedding layer is extruded before armouring and the outer sheath after it; the sheath compound is selected to GB/T 12706.1-2020 and GB/T 19666-2019 as polyvinyl chloride type ST1 or ST2, polyethylene, or low-smoke zero-halogen polyolefin.
An equipment list is only an upper bound. On the same extruder, worn screws, inaccurate temperature control and uneven die clearance put the resulting insulation thickness and eccentricity into different distributions. The operating rate, die-change time and process-parameter records of the equipment say more about actual capability than the nameplate does. Equipment capability and material supply are two independent constraints: insulation compound and semi-conductive screen compound for high-voltage and extra-high-voltage cross-linked cables still depend on imports, and for products of 220kV and above the insulation compound and screen compound must use imported materials. A plant with a cross-linking tower does not necessarily obtain a stable supply of high-voltage compound; a plant with a complete equipment list does not necessarily fix the parameters of every process.
Chapter 29 The Cross-Linking Process: Capacity and Quality Differences between CCV and VC
Cross-linked polyethylene is based on low-density polyethylene compounded with a peroxide cross-linking agent. Under high temperature and high pressure the peroxide decomposes to produce free radicals, which link the linear polyethylene molecular chains into a three-dimensional network structure. After cross-linking the material changes from thermoplastic to thermosetting, its long-term permissible operating temperature rises, and its heat resistance under short-circuit conditions also rises, so almost all medium-voltage and higher power cables use cross-linked polyethylene insulation.
The core equipment of the cross-linking process is the cross-linking line, made up of the extrusion head, the heating section, the cooling section and the take-up and pay-off equipment. The heating section is a pressure-bearing steel tube filled with high-pressure nitrogen to suppress bubbles, and the cable completes heating, cross-linking and initial cooling inside the tube. According to the path of the cable through the tube, lines fall into two types, vertical tower and catenary. The vertical tower type is commonly written VCV (Vertical Continuous Vulcanization), abbreviated VC in some literature, and the catenary type CCV (Catenary Continuous Vulcanization).
In the vertical tower type the cable travels vertically downwards. The triple-layer co-extrusion head is mounted on the upper part of the tower body; the conductor passes vertically down through the cross-linking tube and the cooling tube and is finally turned horizontal at the tower base by a turning wheel. The engineering value of the vertical arrangement lies in eliminating transverse gravity: insulation is relatively thick in the molten state, and if it dwells in a horizontal or inclined position, gravity makes the insulation distribution around the conductor uneven, producing eccentricity. When travelling vertically downwards, gravity acts along the cable axis, and the distribution of insulation around the circumference is determined mainly by the concentricity of the head and the stability of traction.
The cost of the vertical tower lies in its height. According to an industry article (written by a person from Far East Holding Group), Furukawa of Japan built the first tower more than 60 metres high in the early 1980s and later built a 90-metre-high tower (original text). Tower height determines the usable length of the cross-linking tube and the cooling tube, and length determines the heating and cooling time of the cable at a given line speed. If tower height is limited, line speed can only be reduced, otherwise the inside of the cable is not fully cooled when it leaves the tube. The consequence of insufficient cooling appears at the tower base: a cable running vertically must be turned horizontal by the turning wheel, and insulation that is not fully cooled is bent under pressure at the turn, readily producing deformation and internal stress. The article lists this as an engineering constraint of vertical lines.
The heating medium in the cross-linking tube is of two kinds, pressurised nitrogen and steam. Steam transfers heat quickly and costs less, but moisture enters the insulation, so high-voltage products mostly use pressurised nitrogen; the nitrogen pressure also suppresses bubbles, since gas produced by decomposition of the cross-linking agent dissolves or is compressed under high pressure and micropores rarely form inside the insulation. Between the head and the cross-linking tube is the inlet seal, where the cable passes from atmospheric pressure into the high-pressure zone; the condition of the seal directly affects whether the pressure remains stable. In operation the three variables of temperature, pressure and line speed must be kept matched: raising the line speed shortens the heating time, so the temperature or the tube length must be adjusted accordingly; if any one of the three is changed on its own, the degree of cross-linking departs from the process window.
The catenary type arranges the cross-linking tube along a curve approximating a catenary; the cable enters from below and passes along the curve through the heating and cooling sections, with no need for a tall tower. The article states that a CCV line "is not limited by factory building height" and that its "efficiency in producing high-voltage cable is higher than that of a VCV line". With the building-height constraint removed, the cross-linking tube can be made longer and the line speed raised, so one extruder produces more per unit time than on a tower line. The difference on the investment side is equally clear: a tower involves civil works plus steel structure, with the tower body, the lift and the load-bearing platforms all counted, and the capital expenditure of a tower line is far higher than that of a catenary line of the same specification.
The operating variable of the catenary type is the catenary itself. When producing different cross-sections and different insulation thicknesses, the support roller positions, tension and tube inclination must be readjusted so that the actual path of the cable inside the tube matches the set catenary curve. If the adjustment is inaccurate, the conductor shifts up or down in the tube and the eccentricity increases accordingly. This adjustment depends on experience and trial running, so scrap at start-up is higher than on a tower line; the more frequent the change of specification, the more start-up scrap, and this cost must enter the quotation for small-batch orders.
The cost of the catenary type comes back to eccentricity. The cable is not vertical inside the catenary tube, so the molten insulation still has a transverse component of gravity, and the position of the conductor in the tube is maintained by tension and support rollers. The thicker the insulation and the larger the conductor cross-section, the harder it is to maintain concentricity. Hence the division of labour usual in the industry: CCV for medium-voltage cable, VCV for high-voltage and extra-high-voltage cable. This division is an engineering convention, not laid down in any standard clause; cases exist of another type of line being used for the same voltage class, and the judgement rests on the product standard and the type test.
The quality difference finally rests on two inspection items. The first is insulation eccentricity. In Table 3 of the older CCGF 708.1—2015, eccentricity is listed as an inspection item for 6kV to 30kV and for 35kV products; the second is the thickness at the thinnest point of the insulation, which the same table lists as a Class A serious non-conformity, whereas average insulation thickness is only Class B. The relationship between the two is geometric: the greater the eccentricity, the thicker the average thickness must be to satisfy the lower limit at the thinnest point, and material consumption rises accordingly; without such compensation, the thinnest point becomes the starting point for breakdown and partial discharge. High-voltage cable has thick insulation, so the same relative eccentricity corresponds to a larger absolute deviation, and the electric field strength is also higher; this is the direct reason why high-voltage products adhere to the vertical tower arrangement.
Degree of cross-linking is another independent indicator. Insulation that is under-cross-linked elongates beyond the limit in the hot set test, and the same table lists hot set of cross-linked polyethylene insulation as an inspection item. The degree of cross-linking is determined jointly by the cross-linking agent formulation, the heating temperature curve and the dwell time in the tube, and does not necessarily correspond to the tower type; tower type affects eccentricity and capacity, while degree of cross-linking affects heat resistance and service life.
Capacity structure diverges accordingly. Tower lines involve heavy investment and long downtime for specification changes, and suit few varieties, large batches and high voltage; catenary lines involve light investment and start and stop quickly, and scheduling for many medium-voltage varieties is more flexible. The cross-linking threshold for medium voltage is therefore lower than for high voltage, and capacity expands faster. According to an industry report from Jiangsu, manufacturing output of high-voltage and extra-high-voltage cable is about 2 times market demand, and that of mid- and low-end products reaches 3 times. This multiple explains the quotation behaviour of the two kinds of line: medium-voltage products move towards the conversion fee when capacity is surplus, while the prices of high-voltage products remain constrained at both the equipment and material ends.
On-site judgement can start from the tower. The tower is the most conspicuous structure on a plant site; whether there is one, and whether its height matches the product voltage class, can be seen at a glance. A plant with catenary lines and no tower that makes medium-voltage products is a normal configuration; a plant claiming to produce the 66kV and 110kV products covered by GB/T 11017.2-2024 but with no tower line needs to be asked where its insulated cores come from. Buying in cores and then cabling and sheathing them exists in the industry, and in such cases the equipment list and the product catalogue do not agree. Tower type is only circumstantial evidence of product positioning; quality still has to come back to per-drum factory test reports and type test reports.
Chapter 30 Ten Items for On-Site Verification at a Cable Plant
On-site verification must answer three questions: whether the plant has equipment matching the products it claims, whether it has verifiable process records, and whether it has qualifications covering the products being purchased. In half a day to a day of factory auditing, a single walk along the route arranged by the plant makes problems hard to find; only by checking the ten items below one by one, each with a pass line and follow-up points, can the conclusions be compared across plants.
Two things should be done before entering the plant: check the business licence, the production licence and the CCC certificate, and confirm that the address on the certificates matches the actual production address; and designate at random a batch in production, requiring that it be traced from the production release order to the factory test report. The tour route usually avoids problem areas; only when the sampled batch is designated by the auditing party are the conclusions reliable.
- Wire drawing and stranding
What to check: whether the drawing machine has in-line annealing and whether the annealing power supply is interlocked with the take-up; the model, number of cages and maximum stranding cross-section of the strander; whether the compacting die is in use; and how wire joints are spliced.
Pass line: drawing and annealing are completed in the same unit; strander capability covers the declared maximum cross-section; adjacent layers of the stranded conductor are laid in opposite directions; the diameter of the compacted conductor matches the process card; wire joints are made by cold welding or butt welding and dressed, with no burrs and no annealing discolouration at the joint.
Follow-up points: whether annealing is in-line or pot type, and for pot annealing the charge per pot and the turnaround cycle; what the stranding lay length is based on and whether there is a process card; whether the number and position of wire joints are recorded; whether large-section conductors are bought in.
- Insulation extrusion
What to check: the extruder screw and head, the temperature zone displays, the dies, the triple-layer co-extrusion head, and on-line thickness or eccentricity measuring devices.
Pass line: the screw is configured for the compound; actual zone temperatures match the process card and are recorded; the dies are not severely worn; medium-voltage and higher products use triple-layer co-extrusion, with conductor screen, insulation and insulation screen extruded simultaneously; thickness and eccentricity data are available from on-line measurement or sampling.
Follow-up points: whether the screw is changed and the barrel cleaned when the compound is changed; whether the temperature controllers are calibrated; whether eccentricity data are averages or thinnest-point values; how a plant without triple-layer co-extrusion makes medium-voltage insulation, and the interlayer interface in wrapping or two-pass extrusion needs to be questioned.
- Cross-linking
What to check: the tower or catenary line, the cross-linking tube length, the nitrogen system, the take-up and pay-off equipment, and the cross-linking process records.
Pass line: the cross-linking line matches the product voltage class; nitrogen pressure and temperature curve records exist; the line speed matches the process card; each batch has hot set or degree-of-cross-linking test records.
Follow-up points: the actual length of cross-linking tube in use; how nitrogen purity and pressure are monitored; whether the first piece after a specification change undergoes the hot set test; whether high-voltage insulated cores are bought in.
- Cabling
What to check: the cabling machine, fillers, wrapping tapes, and the device for turning sector-shaped cores.
Pass line: cabling lay length and lay direction follow the process card; filling is full and the core is round; the number of tape layers and the overlap rate conform to the process; the phase sequence of sector-shaped cores is correct.
Follow-up points: whether the difference in core lengths is measured after cabling; whether the filler in flame-retardant cable is itself flame retardant; whether the tape in low-smoke zero-halogen cable is halogen-free; how the cabling lay length is matched to the bending radius.
- Armouring and sheathing
What to check: the steel-tape or steel-wire armouring head, the bedding layer, the sheath extrusion line, and the thickness measuring points.
Pass line: the number of armour layers, the gap and the wrapping direction conform to the process; the bedding layer is intact; the thickness at the thinnest point of the sheath meets the requirement; the sheath compound matches the contract.
Follow-up points: whether the steel tape is galvanised and how the coating is accepted; the number and diameter of steel wires; whether the sheath compound grade corresponds to the purchase records; whether the weight loss test records for type ST2 polyvinyl chloride sheath can be produced.
- Test laboratory
What to check: the partial discharge shielded room, power-frequency withstand voltage equipment, conductor resistance measuring equipment, ageing ovens, hot set equipment, tensile testing machines, and thickness gauges.
Pass line: the partial discharge room is shielded and independently earthed, and the background discharge level can be checked; the capacity of the withstand voltage equipment matches the highest test voltage; conductor resistance measurement has a constant-temperature environment or temperature correction; ageing ovens and hot set equipment are within their calibration validity.
Follow-up points: whether the equipment is in use, with operating logs and the date of the most recent test; whether test personnel are full-time; whether raw records agree with the reports issued; which items are outsourced and what qualifications the outsourced laboratory holds.
- Raw materials
What to check: the goods-receipt notes and quality certificates for copper rod and aluminium rod; the grade, batch and shelf life of insulation and sheath compounds; copper tape and steel tape.
Pass line: copper rod has a quality certificate and a composition report; insulation compound is in original packaging from the cable compound manufacturer, with grade and batch; the material store is protected against moisture and dust; first-in-first-out is recorded in a ledger.
Follow-up points: whether copper rod of unknown origin or recycled copper is used; whether insulation compound is supplied in bulk or has absorbed moisture; the grade and origin of high-voltage insulation compound; how masterbatch, flame retardant and cross-linking agent are metered.
- Qualifications
What to check: the industrial product production licence, the CCC certificate, and management system certification certificates.
Pass line: the five product units of the production licence cover the products being purchased; the CCC certificate is within its validity period and covers the model; flame-retardant building wire has been certified in accordance with Announcement No. 9 of 2024 of the State Administration for Market Regulation.
Follow-up points: whether the certificate has been suspended or withdrawn, which can be checked against announcements of the State Administration for Market Regulation; whether the licence address matches the actual production address; how contract-manufactured and private-label products are marked; power cable is not in the CCC catalogue, so a CCC certificate cannot substitute for a production licence.
- Quality records and sampling inspection history
What to check: incoming inspection, in-process inspection and factory test records, the retained-sample cabinet, and national supervision sampling inspection records.
Pass line: every drum has a factory test report corresponding to its drum number; retained samples correspond to batches and the retention period is defined; there are no non-conformity records in national supervision sampling inspections in recent years.
Follow-up points: whether records are filled in after the event; how non-conforming products are segregated and disposed of; whether the plant has refused a sampling inspection, a 2025 notification having named three enterprises that refused inspection; whether rectification reports and re-inspection records can be produced.
- Capacity and equipment ledgers
What to check: the equipment ledger, operating rate records, die-change records, and the production schedule.
Pass line: the declared capacity matches the number of machines, line speeds and shifts; key equipment has backup or an outsourcing arrangement; the production schedule corresponds to the committed lead time.
Follow-up points: which process is the bottleneck; the actual operating rate of the cross-linking line and extruders; the outsourcing ratio and the list of subcontractors; how scheduling priorities are set in peak season.
When the ten items are done, retain the evidence for each: photographs, copies of records, report numbers. Verbal explanations do not count; only records that can be traced to drum number and batch are evidence. Among the ten items, the conductor, cross-linking and test laboratory items carry the greatest weight: the first two determine whether the product can meet the standard, and the third determines whether the plant itself knows whether it meets the standard.
Chapter 31 Acceptance: Factory Tests, Type Tests and Third-Party Testing
Purchase contracts often say "provide test reports", a phrase that does not distinguish between kinds of report. Cable tests fall into three classes, differing in what they attest, in coverage and in who performs them; mixing them creates blind spots. The factory test attests that this drum of product conforms; the type test attests that this design is qualified; third-party testing attests that the submitted sample or the sampled batch conforms to the standard. The three cannot substitute for one another.
| Item | Factory test | Type test | Third-party testing |
|---|---|---|---|
| What it attests | Every drum of product conforms to the standard | A given frozen design conforms to the standard | The submitted sample or sampled batch conforms to the standard |
| Coverage | Every drum | Model, voltage class and cross-section range | The submitted sample |
| Performed by | The manufacturer | The manufacturer or a designated body | An independent testing body |
| Trigger | Before every drum leaves the plant | New product design freeze, change of material or structure, revision of the standard | Tendering, key projects, quality disputes |
| Effect of the report | Mandatory delivery document | Demonstrates design capability | Applies only to the sample submitted |
The table is compiled from public materials, based on the older CCGF 708.1—2015 (the former system, for reference) and on national supervision sampling inspection practice.
The factory test is a routine test carried out on every drum. The routine items for power cable include DC resistance of conductor and the 4h voltage test, with a partial discharge test added for medium-voltage and higher products. Conductor resistance and the 4h voltage test are listed as Class A in Table 3 of the older CCGF 708.1—2015, and the bending test and the partial discharge test that follows it are likewise listed as Class A. The name of the 4h voltage test already states its time cost: every drum must be energised on the test bench for four hours. At acceptance, check whether reports are per drum, whether drum numbers correspond to the physical goods, and whether the test date precedes dispatch. A report that gives only the model and not the drum number cannot prove that the drum delivered was tested.
Factory tests for building wire have separate local rules for comparison. The 2025 edition of the sampling inspection rules of Yueqing City lists 19 inspection items, including insulation thickness, sheath thickness, overall dimensions, conductor resistance, voltage test on finished cable, tensile strength of insulation before and after ageing, weight loss of insulation, tensile strength of sheath before and after ageing, weight loss of sheath, low-temperature bending of insulation and sheath, low-temperature impact on finished cable, thermal shock of insulation and sheath, flexing, and flame propagation resistance; the criteria are those of Parts 3 and 5 of the GB/T 5023 series and Parts 2 and 3 of the JB/T 8734 series cited in those rules. A single drum of building wire undergoes fewer test items than power cable, but sampling coverage is broader, and service properties such as low-temperature performance and flexing account for a larger share.
The product standards also include sampling tests, positioned between factory tests and type tests. Sampling tests draw specimens by batch, with more items than the per-drum test and fewer than the type test, and are used to verify batch consistency. The buyer can stipulate the frequency and items of sampling tests in the contract as a tightening measure beyond the factory test. Factory tests, sampling tests and type tests form a chain of proof running from the single drum to the batch to the design, and third-party testing provides independent verification outside that chain.
The type test is a one-off design verification covering all items. The bending test and the partial discharge test that follows it, and the dielectric loss tangent, all belong to this class: the former checks whether the insulation and screen are damaged after the cable has been bent, and the latter checks the overall dielectric loss level of the insulation. A type test report covers one model series; beyond the voltage classes, cross-section ranges or material combinations stated in the report, the report does not apply. Circumstances that trigger a repeat include freezing the design of a new product, changing insulation or screen compound, changing the conductor structure, and revision of the standard. At purchase, check whether the model, voltage and cross-section on the report agree with the product being bought, and whether the report date is later than the implementation date of the current standard.
Third-party testing is performed by an independent body, and its value lies in the sampling step. Who draws the sample, where it is drawn and whether the sealed-sample record is complete determine how representative the report is. A sample sent in by the manufacturer can attest only the batch submitted, not the batch delivered. Third-party reports are used for tender qualification, re-inspection on entry for key projects and quality disputes; routine items can give a result the same day, while items involving ageing and burning must be scheduled according to the test cycle. When commissioning, confirm that the body holds the relevant qualification, that the report carries a sampling record and photographs of the sealed sample, and that the standard version on which the test is based is the current version. A third-party report cannot replace the per-drum factory test; the two have different coverage.
The checklist at the delivery stage includes: drum numbers corresponding to factory test report numbers; certificates of conformity, nameplates and length markings all present; the production licence number and product units covering the products purchased; for building wire, checking the CCC mark and certificate number, flame-retardant building wire having required certification since 1 July 2025; conductor resistance re-measured on a sampling basis by batch; conductor cross-section and weight checked by the method in Chapter 18; drums and appearance free from damage and moisture. Re-measurement of samples uses a Kelvin bridge or a DC resistance tester; ambient temperature is recorded during measurement and converted according to the standard, otherwise the same cable gives different conclusions at different temperatures.
Data from supervision sampling inspections show why all three kinds of report must be examined. The 2025 national supervision sampling inspection of wire and cable sampled 2204 batches of product and found 65 non-conforming batches, of which 19 batches failed on conductor resistance, 7 batches on insulation thickness and 15 batches on tensile strength of insulation after ageing. The 2024 inspection sampled 3464 batches, with 158 non-conforming batches and 85 batches failing on conductor resistance, a non-conformity rate of 4.56% (158÷3464, calculated for this article). The 2024 online-sold product sampling inspection covered 272 batches, with 41 non-conforming batches, a non-conformity rate of 15.07%, about three times the non-conformity rate of offline inspections in the same period. Many of these non-conforming products came with test reports when they left the factory.
There are five common mismatches: using a type test report in place of a factory test report; one report covering all specifications; a third-party report whose sample was sent in by the manufacturer with no sampling record; a report dated earlier than the implementation date of the current standard; and a CCC mark that does not match the scope of the certificate. The act of acceptance is to match the three kinds of report to drum number, batch and product model respectively, and to fill in whichever link is missing.
Chapter 32 Lead Time, Minimum Batch Size and Transport Radius in Procurement
The unit price on a quotation consists of the copper price and the conversion fee, and the cost delivered to the plant adds three more items: the capital tied up by the lead time, the inventory created by the minimum order quantity, and the freight generated by distance. These three are the items most easily passed over at the enquiry stage and most easily turned into disputes during contract execution.
Lead time is a serial chain: price locking and copper rod purchasing, drawing and stranding, insulation extrusion, cross-linking, cabling, armouring and sheathing, factory testing, packing and transport. The tempo of the chain is set by the bottleneck process. For routine medium- and low-voltage specifications the bottleneck is usually extrusion and cabling, and scheduling is fast when capacity is surplus; for high-voltage products the bottleneck is cross-linking, since a tower line has long downtime for a specification change, and scheduling follows the order book. Non-standard products add material preparation time: the grades and batch sizes of flame-retardant, fire-resistant and low-smoke zero-halogen compounds differ from those of routine compounds, and the armour layer requires a separate armouring head and steel tape. Testing also takes time, since every drum must undergo the 4h voltage test, and when the batch is large the test bench becomes the bottleneck.
A lead-time commitment must be tied to the production schedule. At the enquiry stage, ask the plant for the most recent production cycle for that specification and the volume of orders currently in process, then compare these with the committed lead time; an answer giving only a vague number of days does not allow the scheduling margin to be judged. The lead-time clause in the contract should state the starting point, whether it runs from the entry into force of the contract or from the locking of the copper price and confirmation of the drawings, since in practice the two may differ by several days. Where delivery is in batches, the specification, quantity and date of each batch must be listed in the contract, to avoid easy specifications being delivered first and difficult ones being left to the end.
The copper price turns lead time into a price problem. In 2025 the average price of Shanghai copper over the year was RMB 80,695 per tonne, and in September 2026 the copper price rose above RMB 100,000 per tonne, about 24% higher than the 2025 average (calculated for this article from the two prices). The longer the lead time, the greater the exposure to copper price movements in the interval. Orders with long lead times should fix a mechanism for pricing against a quoted date or for price locking in the contract, pricing copper and the conversion fee separately; the specific approach is given in Chapter 17.
The minimum batch size is determined by the cost of a specification change. An extrusion specification change requires changing dies, cleaning the barrel and reheating; each start-up of a cross-linking line produces start-up material; a drum change in cabling requires re-threading; and a programme change on the test bench requires rewiring. These costs are spread over the batch, so the smaller the batch, the higher the unit price. Drum length is another variable: for the same length, a larger cross-section means a heavier drum and a larger drum size; many short lengths on many drums add joints, drums and handling operations. Products with colour, printing or flame-retardant requirements take a further step up in minimum order quantity, because the changeover batch for masterbatch and flame-retardant compound is larger than for ordinary compound. At the enquiry stage the minimum order quantity for a specification change should be asked directly; the unit price of a standard specification says nothing about the cost of a small batch.
Drum length and the number of lengths must be fixed together at the time of ordering. A cable may not contain a joint in the middle; lengths exceeding the capacity of one drum must be divided into several sections, each sealed at both ends, and joined with intermediate joints during laying. As the number of joints increases, accessory cost and the number of failure points increase together, particularly with high-voltage cable. The ordered length should be measured along the actual laying route, with allowances added at both ends and for bends, then rounded to the capacity of the drum; too much allowance wastes material, while too little requires a supplementary order, and a supplementary order returns to the question of minimum batch size.
Transport radius is determined by the ratio of value to weight. Cable is priced by weight; in 2025 the average export price was USD 10,248.58 per tonne and the average import price was USD 32,987.82 per tonne, the import price being about 3.2 times the export price (calculated for this article from the two averages). This price gap shows that cable of the same weight can differ several times over in value, and freight as a proportion of value differs several times over as well. Building wire has low value and high weight, so freight accounts for a high proportion and the transport radius is short; high-voltage cable has high value and a low freight proportion, so the radius can be extended. Qifan Cable, in its annual report, lists transport cost as one of the factors determining gross margin, alongside the number of processing passes and process complexity, which indicates that freight has already entered the conversion-fee side of the quotation.
The cable drum is a hard constraint in the transport stage. Drum diameter increases with cross-section and length; beyond the vehicle limit an oversize transport permit must be obtained, and both freight and time rise. Drums are usually supplied by the plant against a deposit, and who bears the freight for returning empty drums must be stated in the contract. Cable drums cannot be stored in the open for long periods; the buyer's unloading capacity and warehouse conditions determine the size of a delivery batch, and the batch in turn affects the unit price.
Freight is calculated in several ways, per tonne-kilometre, per axle, per vehicle and per trip, and the basis must be clarified when quoting. Less-than-truckload transport has a high unit price and a greater risk of damage; full-truckload transport has a low unit price but requires a delivery large enough to fill a vehicle. Where the delivery point is in the city centre, requires night entry or lies on a restricted route, freight is further increased. When several quotations are compared side by side, only by converting freight to a per-tonne or per-metre basis does the comparison become meaningful. Export orders carry additional packaging requirements; the wooden drums commonly used require fumigation treatment, and the ordinary drums used for inland transport are not suitable.
Industrial clusters determine the geography of supply. Yixing has nearly 500 cable enterprises, including more than 200 above designated size; Wuwei has more than 270 manufacturing and supporting enterprises, including 104 above designated size; Ningjin has more than 2500 manufacturing and supporting enterprises; Hejian has 160 specialist cable enterprises; and Humen has gathered more than 120 enterprises above designated size. Manufacturers within one cluster share supporting facilities in copper rod purchasing, drum circulation and outsourced processes, and are usually stronger than isolated plants in short lead times and small batches. When buyers compare prices within a cluster, they are comparing conversion fees under one and the same set of supporting conditions.
The contract clauses must fix four variables. For the copper part, agree a benchmark price, a price-adjustment cycle and a method of pricing against a quoted date; state the conversion fee separately and agree whether it varies with batch size and lead time. For freight and unloading, state who bears the cost and who provides the delivery point and unloading equipment. For drums, agree the deposit, the return period and compensation for loss. For lead time, agree the starting point, the batch plan and liability for delay. For categories with surplus capacity, compensation for late delivery can be agreed; for categories with tight capacity, the seller usually requires a higher prepayment percentage. The more specific the clauses, the fewer the disputes at the execution stage.
To compare three quotations side by side, they must be reduced to four figures: the tax-inclusive unit price delivered to the plant, the committed lead time, the minimum order quantity for a specification change, and the allocation of drum and freight costs. Only when all four are present is a quotation comparable; with one missing, the source of a low price cannot be judged.
Part Six: Outlook and Boundaries
Chapter 33 Demand-Side Changes in Grid Investment and Renewable Energy Installations
Demand for wire and cable is spread across several downstream sectors. One listed company's annual report lists its customer industries as power, rail transit, construction, renewable energy, communications, ships, intelligent equipment, metallurgy and petrochemicals Qifan Cable 2025 annual report, cited via Shanghai Securities News. Because the downstream base is broad, price movements in any single sector do not directly determine industry-wide sales; and for the same reason, any statement of "demand growth" must first specify which downstream segment is growing and on what basis the magnitude is measured. The classification of downstream structure is given in Chapter 5.
The publicly reported figure for grid investment is one company's investment plan, and cannot be treated as national equipment procurement spending by the grid.
In January 2025, Xinhua reported that State Grid's full-year investment was "expected to exceed RMB 650 billion (6500 亿元) for the first time" Xinhua. The figure carries three qualifications. The first is the entity: State Grid is one enterprise, while national grid investment also includes China Southern Power Grid and local grids, with generation-side investment counted separately. The second is timing: this is a plan figure set at the start of the year, and the completed figure at year-end was not obtained in the evidence gathering for this report. The third is content: grid investment covers civil works, installation, land acquisition and equipment purchase; cable is only one item of equipment purchase, and there is no public conversion coefficient between investment value and cable procurement value.
Taken together, the three qualifications leave this figure able to establish only a direction: the scale of grid-side investment moved up a step in 2025. It cannot be used to calculate cable demand, nor converted into conductor volume at any ratio. Any treatment that multiplies the RMB 650 billion by a coefficient to derive the size of the cable market lacks a verifiable coefficient.
UHV and submarine cable are often singled out. An industry report of April 2026, citing industry data in the summary of Qifan Cable's 2025 annual report, states that in 2025 the UHV DC cable market exceeded RMB 80 billion, submarine cable reached RMB 17.5 billion, and special-purpose cable for renewable energy reached RMB 150 billion Shanghai Securities News. None of the three figures states its statistical boundary: whether it is cable proper or includes accessories and installation; whether it is an ex-works price or a project price; whether the unit of account is a product or a project. Sub-markets with unclear boundaries, used side by side, will result in double counting.
The three should not be compared directly in order of magnitude either. RMB 150 billion is the whole of special-purpose cable for renewable energy, RMB 80 billion is UHV DC cable, and RMB 17.5 billion is submarine cable; their product scopes overlap, and while the reference year is the same the statistical scope is not. Added together, the resulting RMB 247.5 billion would be neither a sub-total under any single basis nor comparable with total industry output value. This report lists them side by side only; it does not sum them, and it does not divide them into the RMB 650 billion investment figure.
The same report also says that "special-purpose cable for renewable energy accounts for 28% of total industry output value". The statement is incompatible with another set of figures in the same text: if total industry output value is taken as RMB 1.82 trillion (1.82 万亿元), RMB 150 billion corresponds to 8.24%; for 28% to hold, the denominator would have to be RMB 535.7 billion (1500÷0.28). That denominator does not appear in the original report, and is inconsistent with the total output value given in the same text. This report therefore does not use the 28% figure; it uses only the absolute amount of RMB 150 billion, noting that it follows the basis of an institutional white paper, the China Submarine Cable Industry Development White Paper (2025).
The pull of renewable energy installations on cable can be seen in changes to product standards. For DC cable used in photovoltaic systems, the mainstream European standards are EN 50618 and IEC 62930, type H1Z2Z2-K, voltage class DC 1.5kV; the domestic industry standard is NB/T 42073-2016 TÜV Rheinland wire and cable testing and certification material. Charging facilities use GB/T 33594-2025, Cables for Electric Vehicle Charging, implemented on 1 May 2026 National Standards Full-Text Public System. The GB/T 25085 series for road vehicle cables identically adopts ISO 19642:2019 and covers two voltage platforms, AC 600V/DC 900V and AC 1000V/DC 1500V Qinhuangdao Municipal Administration for Market Regulation announcement. The existence of these standards shows that volumes in these applications have stabilised to the point of requiring separate regulation; the standards themselves give no volumes.
The three standards correspond to three operating conditions, and the products are not interchangeable. Photovoltaic cable is exposed to ultraviolet light and temperature cycling over the long term and is assessed for weathering resistance and DC withstand voltage; charging cable is repeatedly flexed and dragged between vehicle and charging post and is assessed for flex life and abrasion resistance; vehicle cable sits near the battery pack and drive motor and is assessed for temperature resistance, oil resistance and shielding. The single phrase "renewable energy is driving cable demand" resolves, on a purchase order, into three categories of product with different specifications, and the supplying enterprises are often not the same set.
The annual increment in installed capacity was not obtained in the evidence gathering for this report. The consequence of this gap is definite: cable is installed once with the project and replaced over a long cycle, so demand is driven mainly by new installations. Without an installation series, "renewable energy demand is growing" can only be a directional judgement, with no magnitude attached. This report omits any specific statement involving installation growth rates.
Communications is another independent demand line. China's fibre-optic cable demand has accounted for about 50% of global consumption for ten consecutive years Jiangsu News, reprint of Xinhua Daily. This share indicates the scale of domestic communications construction, and also indicates that conditions in this segment depend more on carrier centralised procurement and overseas markets, and do not move in step with grid investment.
Export data are more complete than the domestic sub-market data. In 2025 wire and cable exports were USD 32.283 billion (322.83 亿美元), up 22.8% year on year; export volume was 3.15 million tonnes (315 万吨), up 13.6%; and the average export price was USD 10,248.58 per tonne, against USD 9,493.31 per tonne in 2024 Huaon Industry Research Institute, compiled from customs data. The three figures check against one another: dividing the value growth rate by the volume growth rate gives 1.228÷1.136≈1.081, an average price rise of about 8.1%; dividing the two average prices directly gives 10,248.58÷9,493.31≈1.0796, a rise of about 8.0%, the difference between the two routes falling within rounding. Recalculating the volume growth rate from 2.77 million tonnes in 2024 and 3.15 million tonnes in 2025 gives 315÷277≈1.137, consistent with 13.6% 2024 data.
The two years moved in different directions. In 2024 export volume grew 19.2% and value 13.7% Shanghai Electric Cable Research Institute basis, with value growth below volume growth and the average price down about 4.6%; in 2025 volume grew 13.6% and value 22.8%, and the average price turned to a rise. Part of the 2025 price rise can be explained by the copper price: the average Shanghai copper price in 2025 was RMB 80,695 per tonne, up 7.52% year on year Jiangxi Copper 2025 annual report summary, via China Securities Journal. The rise in the copper price and the rise in the average export price are close, but the two are not on the same basis, and changes in product mix also enter the per-tonne price. This report records only this set of same-direction movements; it does not attribute the entire rise in the average price to the copper price.
Value and volume diverge in a period of rising copper prices. In 2025 export value growth was 9.2 percentage points above export volume growth, the difference being carried by the per-tonne price. In September 2026 the copper price stood at RMB 100,000 per tonne industry media; on the same logic, export value growth in 2026 will continue to exceed volume growth. In reading export data, setting the two growth rates side by side is more reliable than looking at value alone; in judging the strength of external demand, export volume and export value should be read together.
Three things on the demand side can be confirmed. Grid-side investment moved up a step, supported by a public figure, though on a plan basis only. Renewable-energy-related products have separated from general-purpose cable to form independent standards and categories, supported by standard numbers. Exports rose in both volume and price in 2025, supported by customs-compiled data. None of the three answers the same question: what share of total industry sales these demands together account for.
Demand structure is bifurcating, and the size of each branch after the bifurcation cannot yet be determined.
Chapter 34 Overcapacity and the Shape of Price Competition
Judgements about capacity in this industry have long rested on a single multiple-based statement. The statement is quoted repeatedly, but it is neither a statistical result nor a series that can be tracked by year.
The multiple appears in a provincial media report of July 2025: "the manufactured production volume of high-voltage and extra-high-voltage cable is about twice market demand, and mid- and low-end products reach three times" Jiangsu News, reprint of Xinhua Daily. The original gives no statistical year, sample scope or calculation method, and its wording is not rigorous: the first clause uses "manufactured production volume" and the second "products", neither of which is the same concept as "capacity". Output is the quantity already produced; capacity is the upper limit of what can be produced. If output is twice demand, the implied capacity utilisation rate can only be lower. The original does not make this distinction.
For the statement to become a trackable indicator, three series are required: national capacity, national output and apparent consumption (output plus imports minus exports). The evidence gathering for this report obtained only imports and exports, not output and capacity. The statement can therefore be cited only as an industry estimate, not used as a statistical conclusion.
The first indirect indicator is concentration. An association-cited basis puts the market share of the top ten enterprises at about 12% Shanghai Securities News; another consultancy basis puts the top ten rising from about 19% to more than 20% Stockstar. The two figures differ by more than 8 percentage points. Low concentration is not in itself evidence of overcapacity; a market with rapidly growing demand can equally be unconcentrated. Its significance lies in this: when demand growth falls back, a fragmented market lacks coordination for output cuts, and price is set by the highest-cost capacity in operation.
The second indicator is profit. Around 80 listed companies saw main-business revenue grow 7.78% in 2024 while net profit fell 25.71% Shanghai Electric Cable Research Institute basis. Another basis reports that about 66% of companies grew revenue in 2024 and about 57% saw attributable net profit decline Jiangsu News. The samples differ; the direction is the same.
Part of this direction comes from accounting treatment and cannot be read entirely as a price war. Revenue contains copper; profit does not. The average spot copper price rose 9.69% year on year in 2024, to an annual average of RMB 75,000 per tonne (7.5 万元/吨) Shanghai Electric Cable Research Institute basis; the average Shanghai copper price rose a further 7.52% in 2025 Jiangxi Copper 2025 annual report summary. Raw materials account for more than 80% of the total cost of wire and cable, of which copper and aluminium account for more than 70% Qifan Cable 2025 annual report, cited via Shanghai Securities News. A rise in the copper price raises both revenue and cost, and with the conversion fee unchanged it compresses gross and net margins. "Revenue up without profit up" is the joint result of price competition and the rise in the copper price; the share of each cannot be separated from public financial statements.
The third indicator is the import-export price gap. In 2025 the average export price was USD 10,248.58 per tonne and the average import price USD 32,987.82 per tonne, the latter 3.22 times the former export data, import data. In 2024 the same ratio was 3.15 times: exports of 2.77 million tonnes and USD 26.296 billion, an average price of USD 9,493 per tonne; imports of 140,000 tonnes and USD 4.19 billion, an average price of USD 29,929 per tonne; the difference of USD 20,435 per tonne is consistent with the USD 204.4 million per 10,000 tonnes given in the compiled version 2024 import and export data. The gap indicates that exports are mainly mid- and low-end and imports mainly high-end, and is consistent in direction with the industry judgement that the high-end market is dominated by foreign firms. The gap itself cannot be read as a quality premium: the product composition on the two sides differs, and the copper price and exchange rate also enter the per-tonne price.
The fourth indicator is quality. The 2025 national supervision sampling inspection covered 2,204 batches, with 65 non-conforming batches, a non-conformity rate of 2.95% SAMR Notice No. 24 of 2026, reprinted on the Hunan provincial government website. In 2024, 3,464 batches were inspected, with 158 non-conforming, 4.56% 2024 notice reprint. In both years the leading non-conformity was conductor resistance (85 batches in 2024, 19 batches in 2025) and the second was tensile strength after ageing of insulation (46 batches, 15 batches); insulation thickness accounted for 36 batches and 7 batches respectively and was not in second place. A conductor resistance non-conformity means directly that the conductor cross-section is insufficient or the copper is impure; an insulation thickness non-conformity means directly that insufficient material was used in extrusion; both point to trading material for price. Online sales show more concentrated problems: the 2024 sampling inspection of online-sold products covered 272 batches, with 41 non-conforming, 15.07% online-sold product sampling notice reprint, 3.3 times the 4.56% for offline (production plus circulation) in the same period.
Price competition falls on the conversion fee; the copper price is not a competitive variable. Statements from two listed companies corroborate each other: Zhongchao Holdings says it adopts a pricing method of "raw material cost + conversion fee + target gross margin" investor relations Q&A, via Eastmoney; Qifan Cable says it uses a pricing model of "copper price + target gross margin", with the copper price benchmark taken as the average of the Changjiang Nonferrous Metals spot market and the Shanghai Futures Exchange, and the gross margin determined by the number of processing passes, the degree of process complexity and transport cost annual report, via Shanghai Securities News. The copper price follows the market; only the conversion fee and the target gross margin are contestable. Pricing mechanisms and cost structure are covered in Chapters 17 and 19.
Cost pressure in 2026 is more direct than in 2025. The average Shanghai copper price in 2025 was RMB 80,695 per tonne Jiangxi Copper 2025 annual report summary, and in September 2026 the spot price stood at RMB 100,000 per tonne industry media, a difference of about RMB 19,300 per tonne. On the rule of thumb in the latter, each RMB 1,000 per tonne rise in the copper price raises cable cost by 1% to 1.5%; converted on this basis, cost rises by about 19% to 29%. This conversion carries three uncertainties: the coefficient itself is a media basis, the annual average and the monthly spot price are not at the same point in time, and the cost base includes the conversion fee. It indicates direction and order of magnitude; it does not represent the actual cost of any individual enterprise.
Competition is fiercest in the segment with the lowest entry threshold. The production licence covers five product units: overhead stranded conductor, plastic-insulated control cable, extruded-insulation low-voltage power cable, extruded-insulation medium-voltage power cable, and overhead insulated cable Fangchenggang Municipal Administration for Market Regulation; China Compulsory Certification covers only three categories, namely cord sets and rubber-insulated and PVC-insulated building wire at 450/750V and below Certification and Accreditation Administration Announcement No. 18 of 2020. The power cable unit of the production licence stops at medium voltage; 110kV and above is outside the licence scope. This segment is not therefore without thresholds: vertical towers, ultra-clean insulation compound, type tests and engineering track record form other thresholds, and the extra-high-voltage insulation compound and shielding compound used at 220kV and above still have to be imported. The access system is covered in Chapter 21.
Competition also operates within clusters. Yixing concentrates nearly 500 cable enterprises and more than 200 enterprises above designated size Yixing Municipal Government; Ningjin has more than 2,500 production and supporting enterprises, and the cluster's revenue in 2025 was RMB 83.96 billion Xinhua; Wuwei has more than 270 production and supporting enterprises, of which 104 are above designated size Wuwei Municipal Government. Within a single cluster, products are homogeneous and transport radii overlap, and price comparison is completed locally.
Expansion expectations at the local level remain. The cable and wire industrial cluster of Humen town saw gross industrial output value exceed RMB 15 billion in 2024, with more than 120 enterprises above designated size, and plans to exceed RMB 30 billion by the end of 2028 and reach more than RMB 50 billion in 2030 Southern Daily. Planning targets are government expectations, not capacity under construction, but they indicate that resources at cluster level are still being invested and that exit pressure has not translated into contraction at local level.
If capacity were indeed two to three times demand, four things should be observable: a sustained low capacity utilisation rate, a continuing decline in conversion fees, a net reduction in the number of enterprises, and a lengthening receivables cycle. Of the four, only the third can be partly tested with public data, and the conclusion is contrary to expectation: cluster enterprises in Ningjin rose from 2,393 in 2024 to more than 2,500 in 2025 Hebei Provincial Department of Industry and Information Technology, Xinhua. The number of enterprises has not fallen net, which may be because the cluster's statistical scope expanded, or because those exiting are small plants that never enter the statistics. This item can neither be disproved nor confirmed on its own.
Overcapacity in this industry is currently supported only by a single multiple-based statement, and lacks capacity utilisation and output series; the results of price competition can be confirmed, but how much of their cause is capacity and how much is demand cannot yet be determined.
Chapter 35 Data Boundaries of This Report
The proprietary data in this report all come from the Tianxia Gongchang enterprise database. Under the wire and cable keyword scope, the number of entities is 103,564, with insured persons totalling 3.0513 million (305.13 万人), of which 64,655 entities have insurance records, accounting for 62.43%.
The basis must be stated: the industry field in the database goes only down to the major class of the Industrial Classification for National Economic Activities, and wire and cable manufacture is a small class (C3831) under the manufacture of electrical machinery and equipment (C38); the database has no marker at this level. The 103,564 entities are obtained by keyword matching against enterprise names, main products and service scope, not by counting industry codes. The keyword set covers cable, wire and cable, wire, magnet wire, optical cable and winding wire; an enterprise that matches several words is counted only once, so the 103,564 cannot be split by product category, and this report gives no proprietary figure for "how many power cable enterprises there are".
Keyword matching has two kinds of error, pulling in opposite directions.
On the over-counting side: entities whose name or business scope contains "cable" also include cable tray (a metal product), cable compound (a plastics product), cable accessories and cable reels (electrical components), and wholesalers that trade only and do not manufacture; entities whose name contains "wire and cable" also include data cable, power cord and wiring harness plants, products that belong to other industries. In-house workshops of downstream enterprises are also counted if they write cable into their business scope.
On the under-counting side: enterprises that describe their main business as "conductor", "winding wire" or "enamelled wire" are not matched if their name and business scope do not contain cable, wire or wiring; enterprises that sell mainly for export and use a foreign-language name or "Wire & Cable" are likewise not matched; workshops that only do contract manufacturing and have no name of their own are not in the database.
The scale of either kind of error cannot be measured. The figure can therefore be treated neither as the number of C3831 enterprises nor as the number of cable plants.
Geographic distribution can be read; capacity distribution cannot. The top eight provinces are Jiangsu 15,336, Hebei 12,118, Guangdong 10,481, Zhejiang 10,167, Shandong 9,642, Henan 5,145, Anhui 4,480 and Shanghai 3,544, totalling 70,913, or 68.47% of the 103,564. The top eight districts and counties are Yueqing 2,902, Dongguan 2,553, Ningjin 2,120, Yixing 1,544, Renqiu 1,450, Yangzhong 823, Tianchang 811 and Zhongshan 787, totalling 12,990, or 12.54%. These proportions indicate the density of keyword hits: inside industrial clusters, traders, supporting plants and downstream workshops are concentrated, and the same keywords match more entities. Treated as capacity shares, they would systematically overstate industrial clusters and understate non-cluster areas.
Insured persons are on a declared basis. Dividing 3.0513 million by the 64,655 entities with insurance records gives an average of 47.2 persons per entity; this average covers only enterprises with declarations and excludes zero-declaration and non-declaring entities, and cannot be used to calculate output value per person or size per plant. The conflict with public figures must be stated: some media report that the industry currently has 4,000 enterprises above designated size and more than 700,000 persons employed Jiangsu News. 3.0513 million is 4.36 times 700,000. The denominators differ: the former covers all entities in this database, the latter only enterprises above designated size. The cause of the difference can be explained; the ratio itself cannot be verified.
The fifth National Economic Census (end-2023) is the most recent stock anchor, but its public bulletin summarises by major class of the Industrial Classification for National Economic Activities; the manufacture of electrical machinery and equipment is listed as one major class, and wire and cable manufacture (C3831) is not listed separately. The census data can therefore give only the order of magnitude of manufacturing and of the electrical machinery major class; they cannot give the number of corporate legal entities or the persons employed in this industry, nor can they be used to check the coverage of this database.
No single authoritative annual figure exists for industry size. The five publicly visible figures belong to four basis systems.
| Basis | Value | Data period | Source and verification level |
|---|---|---|---|
| Association sales | More than RMB 1.8 trillion (expected) | 2025 | Wire and Cable Branch, China Electrical Equipment Industry Association, A |
| Total industry output value | RMB 1.82 trillion (expected) | 2025 | Association cited in a listed company annual report, B |
| Market size | RMB 1.35 trillion, up 4% year on year | 2024 | Consultancy white paper, C |
| Market size | About RMB 1.2 trillion | 2024 | Provincial media citation, B |
| Main-business revenue above designated size | RMB 1,631.103 billion, up 8.1% year on year | 2023 | Second-hand citation only, C |
The table compiles public data. The verification levels follow the definitions in the evidence file: A is a government, association or standard text; B is a listed company announcement, a government website reprint or a provincial party newspaper; C is industry media, a consultancy or a company release. The level describes the type of source, not the accuracy of the figure; two sources at the same level may still give different values.
The five rows belong to four bases — sales, total output value, market size and main-business revenue above designated size — and cannot be added, averaged or compared across years. The size of the differences must be stated: RMB 1.8 trillion and RMB 1.35 trillion differ by RMB 450 billion; RMB 1.82 trillion and RMB 1.35 trillion differ by RMB 470 billion; RMB 1.35 trillion and RMB 1.2 trillion differ by RMB 150 billion. The differences come from the statistical object (all enterprises or enterprises above designated size), the statistical stage (ex-works or end-user) and whether copper contained in the price is included. The 2023 figure of RMB 1,631.103 billion cannot be used as a substitute for 2024: the years differ, the copper prices differ, and main-business revenue above designated size is not the same indicator as sales or total output value. This report cites it only to indicate the order of magnitude under the National Bureau of Statistics basis.
The number of enterprises has four values: association-cited figures put production enterprises in 2025 at more than 7,500, of which more than 4,500 are above designated size; a consultancy basis puts enterprises at more than 12,000, of which more than 4,000 are above designated size; other media say 4,000 above designated size. The above-designated-size counts differ by 500, and the all-enterprise counts by 4,500. Concentration likewise has two sets: about 12% for the top ten on the association-cited basis, and more than 20% on the consultancy basis. Both sets of conflicts are listed side by side in the text, and neither is chosen.
The original page of the General Administration of Customs' Table of Quantities and Values of Major Export Commodities was blocked in this exercise, and the original table was not obtained; the import and export values and quantities in the text all come from compiled customs data and are marked according to the compiled version. The original announcement page of the national supervision sampling inspection was likewise not obtained: the 2025 notice is used through a reprint on the Hunan provincial government website, and the 2024 notice through a reprint on an industry website; the notice title, issue number and date can all be verified on the reprint pages.
The text of GB/T 19666-2019 was not obtained in this exercise, and the clause numbers and limit values cited in the text are treated as pending verification.
For all four designations, no awarding body or year was found in this exercise: "Home of Chinese Wire and Cable" (Yixing), "Home of Chinese Cable" (Wuwei), "China Cable City" (Guanlin, Yixing) and "National Wire and Cable Industry Base" (Guanlin, Yixing). The text writes only "designations used by local governments and media". The document recognising the "four major cable production bases" was likewise not obtained, and this report gives neither the year nor the body of recognition.
No credible basis exists for the sub-market size of special-purpose cable. The retrieved figures for LSZH, high-temperature-resistant and flame-retardant categories contradict one another and all come from document resale sites; this report uses none of them and gives only qualitative description. The mandatory aluminium-for-copper substitution ratio has three versions — 30%, 35% and 90% — none of which is found in the original Ministry of Industry and Information Technology text; this report does not state one.
The time boundary for policies and standards is 18 October 2026. On standards, GB/T 13849-2025 takes effect on 1 May 2026, replacing GB/T 13849.1-2013 National Standards Full-Text Public System; GB/T 33594-2025 takes effect on the same day National Standards Full-Text Public System; GB/T 7424.3-2025 has already taken effect on 1 February 2026 National Standards Full-Text Public System. Readers verifying the text should note the version dates.
This report's approach is fixed at three rules: where an indicator has several bases, set them side by side rather than averaging; mark the citation level of figures taken at second hand; do not use figures whose original text was not obtained.
Chapter 36 Questions This Report Does Not Resolve
Ten questions were confirmed unanswerable in the course of writing, ordered from greatest to least impact. Each states what data are missing and which class of conclusion the gap prevents from holding.
The first is the capacity utilisation and output series. Historical series for national wire and cable output, capacity and operating rates are missing, as is output by province. Without them, "about 2 times for high voltage and about 3 times for mid- and low-end" can only be cited as an industry estimate, cannot be verified, and cannot show whether the degree of overcapacity is widening or narrowing. The more direct consequence is that revenue cannot be split into price and volume: revenue contains copper, a rising copper price raises revenue, and a decline in output would be concealed by revenue. If this series were published first, the first thing to revise would be Chapter 34.
The second is the number of C3831 enterprises, persons employed, and new registrations and cancellations. The census mid-class table, annual data for enterprises above designated size, and annual counts of newly registered and cancelled enterprises from the market regulation authorities are missing. Without them, it is impossible to calculate the coverage of this database's 103,600 entities against the industry, to calculate average plant size and size distribution, or to judge whether "a large number of enterprises" is an industry characteristic or an effect of the statistical boundary. Chapter 34 uses the rising enterprise count in the Ningjin cluster to question the exit judgement; what is missing is precisely this layer of data.
The third is output and prices by voltage class. Output, capacity and winning bid prices for the 1kV, 10kV, 35kV, 66–110kV and 220kV-and-above classes are missing. Without them, the structural judgement that "overcapacity at the mid- and low-end is higher than at high voltage" cannot be verified, nor can it be judged whether the product mix is upgrading. The 3.22-fold price gap between average export and import prices shows only that the products at the two ends differ; it cannot show supply and demand in each class.
The fourth is the conversion fee series. Pricing formulas are public; conversion fees are not. Time series of conversion fees by specification and by process (yuan per tonne or yuan per kilometre) are missing. Without them, the intensity of price competition can be observed only indirectly through gross and net margins, and both indicators are affected at the same time by the copper price, product mix and capacity utilisation. The margin decline given in Chapter 34 cannot be decomposed to show the share taken by lower conversion fees.
The fifth is the lag in copper price pass-through and inventories. The proportions of pricing against a quoted date and price locking, the order cycle, and the turnover of raw material and finished goods inventories are missing. Without them, it is impossible to measure the lag from spot copper price to financial statements, or to judge how much of the cost after September 2026, when the copper price reached RMB 100,000 per tonne, will be passed out through contracts and how much will remain in the plant. This item determines the profit judgement for 2026 to 2027, and also determines whether the conclusion on contract risk in Chapter 20 can be quantified.
The sixth is tariff subheadings for exports. The eight-digit and ten-digit subheadings of HS 8544, the country structure and the types of operating entity are missing. Without them, the rise in the average export price cannot be decomposed: a higher copper price, an upgraded product mix and exchange-rate movements all push up the dollar per-tonne price, and the available data give only one aggregate average. Nor can it be judged which category of product and which market export growth comes from.
The seventh is the volume structure by downstream segment and grid procurement. Cable volumes for the three downstream segments of construction, industry and communications are missing; so are the tender and award prices and award shares of State Grid and China Southern Power Grid, and the conversion relationship between installed capacity and cable procurement. Without them, the RMB 650 billion investment cannot be converted into cable demand, and the statistical boundaries of the three figures — RMB 80 billion for UHV DC cable, RMB 17.5 billion for submarine cable and RMB 150 billion for special-purpose cable for renewable energy — cannot be verified. The demand-side chapter can therefore go only as far as direction.
The eighth is sub-segment volumes of special-purpose cable and aluminium conductors. The output and output value of flame-retardant, fire-resistant, LSZH, photovoltaic and new-energy-vehicle cable are missing, as is the share of aluminium conductor in power cable. Without them, it is impossible to answer whether the share of special-purpose cable is rising, to quantify the claim of product upgrading, or to judge whether "saving copper with aluminium" has already occurred at the manufacturing end. Wuwei's published statement that "the share of special-purpose cable output value has risen to 40%" is on a county-level basis Wuwei Municipal Government, and cannot be extrapolated nationally.
The ninth is capacity under construction at cluster level. Projects under construction, commissioning dates and designed capacity for each cluster are missing. Without them, the supply increment of the next two to three years cannot be judged. Humen town plans to exceed RMB 30 billion in output value by the end of 2028 and to reach more than RMB 50 billion in 2030 Southern Daily, and Ningjin's cluster revenue in 2025 was RMB 83.96 billion Xinhua; the growth targets of both places are written in government documents and reports, but no public list of the corresponding new capacity exists.
The tenth is national data for the first half of 2026. Output, export value and main-business revenue above designated size for January to June 2026 are missing. Without them, it cannot be judged whether the 22.8% export growth of 2025 continued. Wuwei's output value of RMB 25.83 billion in the first half of 2026, up 42.5% year on year Wuwei Municipal Government, is a single-point figure and cannot represent the country; nor can the growth rates of the two years be linked directly, since the city's growth rate in 2025 was 19.2%. The verification date of this report is 18 October 2026; full-year 2026 data have not yet been released, and until then any judgement involving 2026 rests only on first-half and single-point data.
These ten gaps fall into three classes, differing in how hard they are to fill. The first is stratification within the statistical system: census bulletins go down to major class, industry statistics to mid-class, enterprise reports to the legal person. Items one to three belong here; the data exist but are not published at this level, and will become available as yearbooks and departmental reports appear. The second is commercial and industry sub-segment information: conversion fees, the share of pricing against a quoted date, inventory turnover, tender award prices, export tariff subheadings, special-purpose cable volumes. Items four to eight belong here; the data sit with enterprises, tender issuers and customs, do not enter public statistics, and can only be assembled through continuous survey and industry aggregation. The third is local bases and timing: capacity under construction in clusters, county-level output value structure, and current-year data. Items nine and ten belong here; the figures for the first two sit in local ledgers, and the published part is often a target value; current-year data must await release.
Among the ten, items one to three determine whether industry totals can be stated clearly, items four and five whether price competition can be stated clearly, items six to eight whether demand structure can be stated clearly, and items nine and ten whether near-term trends can be stated clearly. The order of data collection follows this arrangement.
If capacity utilisation and output series were published first, the first thing to revise in this report would be the statement on the degree of overcapacity in Chapter 34; if HS subheadings were published first, the first thing to revise would be the attribution of the rise in the average export price in Chapter 33; if the conversion fee series were published first, the first thing to revise would be the judgement of competitive position in Chapter 34. If any one of the three arrives, the conclusions of this report advance one level from directional judgement; the rest remain where they are.
The cost of obtaining the three differs as well. Statistical reports are a public good, available on publication; conversion fees and award prices require series, and a single data point is meaningless, so continuous tracking is needed; capacity under construction in clusters requires verification place by place and is prone to omission. The most costly is the second class of commercial information, which is precisely the key to judging the intensity of price competition.
Until the three data sets — capacity utilisation, output by class and conversion fees — are published, the judgements in this report on China's wire and cable industry reach only two levels: the boundaries of product, process and access can be described factually, while the magnitude of demand and competition can be stated only as direction.
Data Sources and References
The data in this report fall into three categories. The first is the supply-side statistics of the Tianxia Gongchang enterprise database, compiled by matching keywords against company names, products and business scope rather than by taking figures by industry code under the Industrial Classification for National Economic Activities; the basis is explained in each chapter, and these data are used only for distribution and structural comparison, not as national totals. The second is government documents, national standards, association statistics and census bulletins. The third is public company disclosures and industry reporting, whose nature is indicated in the text where they are cited. The last verification date for all links is 18 October 2026.
- Tianxia Gongchang industrial platform. The counts of manufacturing entities, insured persons and regional distribution referred to in this report are all taken from the Tianxia Gongchang enterprise database.
The other sources cited in the text are as follows.
- Industry report
- Industry report
- National Standards Full-Text Public System
- CCGF 708.1—2015
- National Standards Full-Text Public System
- Fangchenggang Municipal Administration for Market Regulation
- National Standards Full-Text Public System
- Draft standard for 1kV to 35kV aluminium alloy cable
- Draft standard for 220kV
- National Standards Full-Text Public System
- Guofa [2024] No. 11
- Wuwei Municipal Government
- Search results
- SAMR Announcement No. 18 of 2020
- SAMR Announcement No. 9 of 2024
- National Standards Full-Text Public System
- std.samr.gov.cn: old standard
- National Standards Full-Text Public System
- Standard information
- TÜV testing and certification material
- Local market regulation announcement
- National Standards Full-Text Public System
- Shanghai Securities News report
- xinhuanet.com: Xinhua
- Huaon Industry Research Institute compilation
- Import data compilation
- Association release
- Second-hand citation
- Industry body citation
- Yixing Municipal Government
- Jiangxi Copper 2025 annual report summary
- 2024 notice reprint
- 2025 notice
- China Electrical Equipment Industry Association
- Yixing Municipal Government interview
- news.gmw.cn: Guangming Online
- Wuwei Municipal Government
- Wuwei Municipal Government
- www2.xinhuanet.com: Xinhua
- Hebei Provincial Department of Industry and Information Technology
- Great Wall Network Cangzhou channel
- Huzhou Municipal Administration for Market Regulation
- Southern Daily
- Eastmoney investor relations record
- China Cable Network industry report
- Ministry of Industry and Information Technology release page
- Sohu industry report
- Yanshi Municipal People's Government website EIA disclosure
- Industry article reprint
- Catalogue note
- Administrative reconsideration document
- Announcement No. 49 of 2025
- GB/T 12706.1-2020
- Yueqing 2025 implementation rules
- 2024 online-sold product sampling notice
- 2024 data
- Stockstar
Cover image: Win Henderson (FEMA) / Public domain / original file page. Used under the original licence without alteration to the content; only resized proportionally to 2,000 px wide and stripped of metadata.