How to Size Medium Voltage Cable Systems for Industrial Loads

Sizing medium voltage cable systems for industrial loads is often reduced to a quick ampacity exercise: calculate the expected current, open a manufacturer’s table, and select the next conductor size. That approach may produce a cable that works under normal conditions, but it can leave expensive weaknesses in the system. Industrial feeders do not live in a spreadsheet. They run through hot cable trenches, crowded ducts, chemically exposed areas, long underground routes, motor-starting events, and fault conditions that may last longer than the original protection study assumed.

A sound selection starts with the question behind the load figure: what will this feeder actually be asked to do over its operating life? A medium voltage circuit serving a continuously loaded process transformer requires a different decision than one feeding a large compressor motor, a variable-speed drive lineup, a mine conveyor, or a future-ready distribution ring. Voltage rating and conductor area matter, but they are only part of the engineering judgement.

For technical evaluations, the practical goal is not simply to avoid undersizing. It is to select a cable system that remains thermally stable, survives foreseeable faults, works with the intended protection philosophy, can be installed without damaging its insulation or screen, and does not become the first bottleneck when the plant expands.

Start with the electrical system, not the cable catalogue

Before comparing conductor sizes, define the system clearly. The nominal medium-voltage level, grounding arrangement, maximum system voltage, available short-circuit level, protection clearing times, and load duty all influence the cable specification. A cable selected for a resistance-grounded network may need a different insulation-screening and fault-duty review than one installed on a solidly grounded system. The answer is not universal; it depends on the local grid architecture and the rules used by the project.

The load model also deserves more scrutiny than it usually receives. A feeder supplying one large induction motor should not be assessed using only the motor nameplate full-load current. Starting current, starting duration, starts per hour, acceleration time, driven-load inertia, voltage dip limits, and upstream source stiffness may all matter. Where large motors are started direct-on-line, the cable voltage drop during starting can become more restrictive than steady-state ampacity. By contrast, a feeder to a modern drive system may face lower starting current but more attention to harmonics, switching transients, grounding practice, and compatibility between the cable system and the drive output arrangement.

For process loads, diversity must be handled carefully. Diversity is reasonable when load coincidence is genuinely limited by the operating sequence. It is less convincing when every pump, fan, mill, or conveyor can run during a production peak. In industrial plants, “normal operation” often changes after a maintenance bypass is introduced, a standby machine becomes permanent, or a process upgrade increases utilization. A conservative load basis is not overdesign by default; it is protection against a misleading operating assumption.

Ampacity is an installation calculation, not a conductor property

Cable ampacity tables are useful starting points, but their stated values are conditional. They assume particular ambient temperatures, burial depths, soil thermal resistivity, spacing, formation, conductor temperatures, and bonding arrangements. Once the route differs from those conditions, the tabulated rating may no longer represent the installed circuit.

The most frequent source of optimistic sizing is grouping. A single cable in open air behaves very differently from several three-core cables in a congested trench, or multiple single-core circuits sharing duct banks. Thermal interaction can be substantial, especially where future circuits have been added without reassessing the original heat model. Spare ducts do not create an immediate derating issue, but occupied adjacent ducts can. So can a route that transitions from direct burial to duct, then to tray inside a hot electrical room.

For underground systems, soil conditions are not a minor detail. Dry soil can have poorer heat dissipation than the assumptions used for a standard reference installation. Thermal backfill may be justified on heavily loaded routes, but only if its properties, installation method, and long-term condition are properly considered. A good thermal calculation should identify the controlling section of the route rather than averaging favorable and unfavorable sections into one convenient number.

Standards such as IEC 60287 are commonly used as a basis for cable current-rating calculations, while applicable local rules, utility requirements, and project specifications may set further conditions. The calculation method should be visible in the design record. “Rated at X amperes” is not enough unless the assumed installation conditions are equally clear.

A useful review of route conditions

• Installation method: direct burial, duct bank, tunnel, tray, ladder, conduit, or mixed route.
• Ambient air temperature, ground temperature, and any localized heat sources.
• Number of loaded circuits, physical spacing, cable formation, and planned spare routes.
• Soil thermal properties, burial depth, backfill specification, and risk of soil drying.
• Route length, elevation changes, pulling sections, bend radii, and joint locations.
• Exposure to water, oils, chemicals, vibration, rodents, sunlight, or mechanical damage.

This review often changes more than the conductor size. It may point toward a different installation arrangement, improved spacing, a larger duct, separate circuits, a revised route, or a change from one cable construction to another. Those options can be more practical than simply adding copper.

Check voltage drop in the modes that actually matter

At medium voltage, steady-state voltage drop is sometimes dismissed because the percentage appears modest over typical industrial distances. That can be a mistake on long feeders or heavily loaded networks. More importantly, the critical event may be motor starting, transformer energization, or a process transition where several motors accelerate in sequence.

Voltage-drop assessment should use the full circuit impedance, including conductor resistance, reactance, route length, cable arrangement, and the characteristics of the upstream source. For motor applications, assess the voltage available at the motor terminals during the proposed starting method, not just at the switchgear busbar. A motor that starts successfully in a simplified model can still create an unacceptable voltage depression for adjacent controls, contactors, instrumentation, or other motors.

Long cable runs feeding variable-frequency drives need a separate conversation. The medium-voltage feeder on the supply side is not the same design problem as the drive-to-motor cable. On the output side, reflected-wave effects, insulation stress, common-mode current, shielding, grounding, and permitted cable length are often governed by drive and motor manufacturer guidance. Treating both sections as ordinary feeders is a familiar route to late-stage redesign.

Fault duty can decide the design before ampacity does

A medium voltage cable must withstand more than its normal load current. During a fault, conductor and metallic screen or sheath components may carry high current for a short period. The thermal duty depends on the prospective fault current and the actual clearing time of the protection system. A protection setting study is therefore not separate from cable sizing; it is one of its inputs.

The conductor short-circuit check is usually straightforward in principle, but the screen is often where assumptions become fragile. Screen current sharing can vary with cable construction, bonding method, parallel paths, earthing layout, and whether circuits are single-core or three-core. In single-core arrangements, sheath bonding choices can also affect induced voltages, circulating losses, and the behavior of the metallic sheath under fault conditions.

Do not accept a generic screen size merely because it is standard in a supplier’s product range. Ask how the earth-fault return path was modeled and whether the calculation aligns with the actual grounding design. IEC 60949 is one established reference for calculating thermally permissible short-circuit currents in cables, but the governing project method and applicable regional requirements should be confirmed early.

This is particularly relevant where a plant is being expanded around an older electrical system. New transformers, embedded generation, larger utility connections, or altered bus-tie operating positions can raise available fault levels. A cable that was acceptable when installed may not remain acceptable after the network changes.

Choose construction for the environment and maintenance reality

The phrase “medium voltage cable” covers many constructions. A selection may involve copper or aluminium conductors, XLPE or other insulation systems, copper wire screen or tape screen, metallic sheath, armour, oversheath compounds, and water-blocking features. None of these should be treated as a default upgrade without a reason.

Copper and aluminium are a common early decision. Aluminium can be a sensible choice for large fixed feeders when its larger conductor size, terminations, jointing requirements, mechanical handling, and enclosure space are accommodated. Copper may be preferred where compactness, flexibility, termination constraints, or higher conductivity per unit area are decisive. Material market movements can affect procurement timing, but a volatile copper or aluminium price should not be allowed to override installation feasibility or lifecycle risk.

For wet locations, water ingress deserves more than a generic note in the specification. Longitudinal and radial water blocking may be relevant depending on route conditions, cable construction, jointing exposure, and the consequences of a failure. In industrial areas with hydrocarbons, corrosive chemicals, high mechanical risk, or rodent exposure, the outer sheath and mechanical protection should be selected around the actual environment. Armour is not a substitute for a properly designed route, and a very heavy cable can create pulling and termination challenges of its own.

Installation practice matters as much as the cable data sheet. Pulling tension, sidewall pressure at bends, minimum bending radius, storage conditions, sealing of cable ends, and workmanship at joints and terminations all shape reliability. A technically correct cable can be damaged before it is energized. Where route geometry is difficult, it is worth reviewing drum lengths and joint locations before procurement rather than discovering an impractical pull during construction.

Build expansion capacity deliberately, not vaguely

“Allow for future growth” is frequently written into project briefs without defining what it means. That can lead either to costly oversizing or to a cable system that has no realistic upgrade path. The better approach is to identify likely expansion scenarios: an additional production line, a larger motor, a second transformer, distributed generation, battery storage, or more power-electronic loads. Then test whether the selected feeder, duct bank, switchgear, protection system, and earthing network can accommodate those scenarios.

Sometimes the right reserve is not a larger conductor. It may be spare ducts, accessible trenches, switchgear feeder positions, provision for parallel circuits, or monitoring points that make future loading visible. Parallel cables can solve an ampacity or voltage-drop problem, but they introduce requirements for equal lengths, compatible arrangements, termination discipline, and balanced current sharing. They should be a conscious engineering choice, not an emergency response to a late load increase.

Digital grid planning is making this more relevant. Plants are adding variable-speed drives, onsite generation, energy-management systems, and more detailed power-quality monitoring. These changes do not automatically require oversized cables, but they do make static assumptions less dependable. Cable sizing should sit within a wider view of source behavior, load evolution, protection coordination, and operational data.

A decision record prevents expensive assumptions from disappearing

The most useful output of the sizing process is not only a cable schedule. It is a traceable decision record that explains why the selected system is suitable. Record the load basis, demand assumptions, installation conditions, derating factors, voltage-drop results, fault-duty inputs, earthing assumptions, cable construction, jointing approach, and any future-capacity rationale. If a later project challenges the choice, the engineering team can see what was assumed instead of repeating the entire design from memory.

That discipline also helps during procurement. A supplier quotation may satisfy the stated voltage and conductor size while differing in screen design, sheath construction, conductor class, test documentation, or accessories. The cable, joints, separable connectors, terminations, glands, bonding leads, and surge-protection interface should be reviewed as one installed system. Mixed responsibility between package suppliers is a common place for details to fall through.

For teams tracking global power equipment and industrial electrification, this is where market intelligence becomes practical rather than abstract. Changes in conductor-material availability, drive technology, smart switchgear integration, and grid modernization plans may affect the options worth evaluating, but they do not replace the project calculation. GPEGM’s perspective on the energy foundation and the digital grid is useful precisely because cable decisions connect both worlds: a conventional asset choice today can determine how easily an industrial network adapts tomorrow.

The final selection should therefore be tested against four questions: Can it carry the real load in the real route? Can it ride through the electrical events the system can impose? Can it be installed and maintained reliably? And can the plant live with it when the load profile changes? If any answer depends on an undocumented assumption, the medium voltage cable system is not fully sized yet.