How to Size a Distribution System: Transformer, Switchgear and Feeders in One Calculation

Sizing the transformer is only one third of the job

Ask most buyers how to size a distribution system and they answer with one number: the transformer kVA. Then they hand the “wiring” to someone else and discover later that the main breaker cannot clear the fault, the feeder cable runs too hot, or the protection trips the whole plant when one motor starts. The transformer is only one third of the system. In a correctly sized distribution system, four components are locked together by one chain of numbers:

• The transformer impedance (Z%) sets the maximum short-circuit current the system can feed.
• The short-circuit current sets the breaking capacity of every switchboard and the thermal withstand of every cable.
• The feeder current and length set the cable cross-section (ampacity, voltage drop, short-circuit).
• All of the above set the protection settings that keep a fault local instead of blacking out the building.

This guide walks one 800 kVA project through the full chain — load list, transformer, short-circuit current, switchgear, feeders and protection coordination — with real 2026 FOB prices and formulas you can put straight into your specification. If you only need the transformer rating itself, the single-transformer calculation lives in our Transformer Sizing & Budget in One Step guide; this article is about everything downstream of that number.

Step 1 — Turn the load list into one system kVA

Everything starts from a single, honest number: the maximum simultaneous demand of the whole installation, not the sum of every nameplate. Add up the connected load of every feeder, then apply a demand (diversity) factor that reflects how much of it actually runs at the same time:

Multiply each feeder’s connected kW by its demand factor, add the results, and you have the maximum simultaneous demand in kW. For a motor-heavy plant, remember that the largest motor dominates the starting peak — account for its locked-rotor current separately, not just its nameplate kW. Before you compute anything, also fix the system voltage: the same capacity can sit at 6, 10, 11, 20 or 33 kV, and every downstream calculation changes with it. Our Voltage Class Selection Guide covers that decision first if it is not already made.

Step 2 — Size the transformer and note its impedance

Once you have the demand in kW, the transformer kVA follows from the same formula used for a single unit — kVA = kW ÷ (Power Factor × Load Factor) — rounded up to a standard rating. We do not repeat the full derivation here; the sizing formula guide has the table of standard ratings and worked examples. What this article adds is the number you must write down next to the kVA: the transformer’s short-circuit impedance, Z%.

Z% is the percentage of the rated voltage that produces rated current when applied to the short-circuited secondary. It is printed on every test report and every nameplate. Typical values for 10/0.4 kV distribution transformers (confirm the actual figure on your supplier’s test report):

Why does Z% matter so much? Because it is the denominator of the short-circuit calculation in the next step. A transformer sold “cheap” with a higher Z% reduces fault levels but also worsens voltage regulation under load — the two effects pull against each other, which is exactly why you should ask for the actual Z% from the test report rather than assuming a catalogue value. For reference pricing at your chosen rating, the Distribution Transformer Price Breakdown 2026 gives oil-immersed S13 and dry-type SCB13 FOB Qingdao figures at every standard step (for example 800 kVA S13 ≈ $7,730).

Step 3 — Calculate the short-circuit current: everything hangs off this number

Two formulas, and every switchboard and cable in the project is sized from the answer:

Rated current: I_rated = S(kVA) × 1000 ÷ (√3 × U)

Symmetrical short-circuit current: I_k ≈ I_rated ÷ Z%

Applying them at 400 V secondary for typical ratings:

These are conservative first-pass figures (no source impedance, no motor contribution); a real fault study adds the upstream grid and motor feedback and usually lands higher. The point is the same: the transformer kVA alone tells you nothing about fault level — the Z% does. That is the step buyers who “only size the transformer” skip, and it is why a switchboard sized for a 630 kVA unit can be destroyed the day it is fed from a 1,250 kVA unit on the same bus. The International Standards Comparison guide lists which IEC / IEEE / GB clauses govern the fault calculation if your contract specifies a particular standard.

Step 4 — Size the switchgear and breakers

Every breaker and switchboard has two numbers that must both clear the bar:

• Rated current (I_n) ≥ the continuous current of the circuit it protects (feeder ampacity), with margin.
• Breaking capacity (I_cu / I_cs) ≥ the maximum short-circuit current at its terminals. Standard LV frames are 25 kA, 36 kA, 50 kA, 65 kA.

Notice the trap in the table above: an 800 kVA unit with Z = 4.5% produces 25.7 kA — a 25 kA breaker is just short. Buying the “standard” 25 kA frame to save a few dollars leaves the main breaker unable to clear its own busbar fault. The correct pick is the 36 kA frame. This is a routine, expensive mistake that only shows up in a fault, and only a written short-circuit calculation catches it before the switchboard is ordered.

On the MV side the same logic applies with vacuum circuit breakers and ring main units. A typical configuration is an MV incomer (VCB or RMU with fuse-switch) feeding the transformer, then an LV switchboard on the secondary. Reference budget ranges from our published pricing (FOB, per bay):

For the full product range behind these figures, including oil-immersed, dry-type and complete substation packages, see the Transformer Selection Guide.

Step 5 — Size the feeders: three constraints, not one

A cable is correctly sized only when it passes all three checks. Most designs stop after the first.

1. Ampacity (current-carrying capacity). The cable’s rated current at its installation conditions (buried / in air, ambient temperature, grouping factor) must cover the feeder current. Governed by IEC 60364-5-52 / NEC ampacity tables.
2. Voltage drop. Total voltage drop from transformer to the furthest load should stay under about 5% for power circuits (stricter for sensitive equipment). Long LV feeders fail this long before they fail ampacity.
3. Short-circuit thermal withstand. The cross-section must survive the let-through energy of a fault: S ≥ I_k × √t ÷ k, where t is the clearing time in seconds and k is a material constant (≈ 143 for copper / XLPE, ≈ 115 for copper / PVC).

A concrete illustration. A 200 kW load at 400 V, PF 0.85 draws about 340 A. On a 200 m feeder, copper XLPE cables behave as follows:

Note what the calculation actually decided: for a short, well-loaded feeder, ampacity wins; for a long feeder, voltage drop wins; for a feeder close to a large transformer, short-circuit withstand wins. A cable chosen for the wrong constraint is either unsafe or silently expensive. Earthing and bonding for the same feeders is covered in our Low Voltage Grounding System Design Guide.

Step 6 — Coordinate the protection

Protection coordination means that when something fails, only the nearest protective device operates. The mechanism is time-current discrimination: each upstream device is delayed relative to the one below it so a fault is cleared by the closest breaker first, leaving the rest of the plant online.

A typical distribution hierarchy, from the transformer down:

1. Transformer MV protection (fuse or relay): fast overcurrent (short-circuit) plus overload with thermal image; this is the backstop for transformer internal faults.
2. LV main breaker (ACB): long-time pickup for overload, short-time for moderate faults, instantaneous as the final backstop.
3. Feeder MCCBs: set below the main breaker so a feeder fault trips the feeder, not the whole board.
4. Motor / load protection: overload relays and MCPs with settings matched to the equipment, including inrush.

Two settings mistakes are common. First, setting the main breaker too sensitive — the plant trips on every motor start, which operators then “fix” by raising every setting, removing all coordination. Second, ignoring transformer inrush: the magnetising inrush of a distribution transformer is roughly 8–12× rated current for a few cycles, so the upstream protection must ride through it. If your load is rich in electronics or variable-speed drives, the harmonic content changes both transformer and breaker duty — our K-Factor Transformer Selection Guide explains when a K-rated unit is the right call.

Worked example — an 800 kVA project, end to end

Put the whole chain together with one coherent project. A small factory: maximum simultaneous demand 520 kW, power factor 0.85, planned load factor 80%.

1. Transformer kVA: 520 ÷ (0.85 × 0.80) = 765 kVA → round up to 800 kVA (S13 oil-immersed, ≈ $7,730 FOB).
2. Rated current: 800 kVA at 400 V = 1,155 A.
3. Short-circuit current: with Z = 4.5%, I_k = 1,155 ÷ 0.045 ≈ 25.7 kA.
4. LV main breaker: I_n ≥ 1,250 A, I_cu = 36 kA (not 25 kA — see the trap above).
5. Main LV busbar / cables: for 1,155 A use parallel runs of 300 mm² copper XLPE (or a busbar trunk); thermal withstand check: 25,700 A × √0.3 s ÷ 143 ≈ 98 mm² minimum — ampacity, not withstand, is the deciding constraint here.
6. Long feeder: 200 kW / 340 A at 200 m → 185 mm² copper (3.72% drop, ampacity OK) rather than 120 mm² which fails both.
7. Protection: MV fuse/relay set to ride through inrush; LV main breaker delayed; feeder MCCBs graded below it; motor protection sized to each drive.

Budget snapshot for this project (FOB reference): transformer ≈ $7,730; LV switchboard 2–3 bays ≈ $3,600–7,500; MV incomer / RMU ≈ $1,500–3,200; LV feeders per the table above. Remember these are ex-works FOB figures — the FOB Quotation Breakdown explains freight, insurance, duties and inland transport, and the Total Cost of Ownership guide shows how to add 20 years of losses and maintenance on top before you compare suppliers.

The five mistakes that break distribution systems

• Sizing only the transformer. The kVA is the start, not the system. Z%, fault level, breakers and cables must follow from it.
• Assuming a Z% instead of reading the test report. A 1,000 kVA unit can be 4.5% or 6%; the fault level differs by a third and so does the breaker you must buy.
• Buying the standard breaker frame. 25 kA frames fail on systems that produce 26–32 kA. Write the fault calculation before ordering switchgear.
• Sizing cables on ampacity alone. Long LV feeders are governed by voltage drop; feeders near big transformers are governed by short-circuit withstand. Check all three.
• Setting protection without coordination. Grading by guesswork either blacks out the plant or removes all selectivity. Produce a time-current coordination study, even a simple one.

Final checklist before you issue the spec

• Maximum simultaneous demand computed from demand factors, including the largest motor’s starting peak.
• Transformer kVA rounded up from the formula, and the actual Z% confirmed from the test report.
• Short-circuit current calculated at every busbar and written on the single-line diagram.
• Every breaker checked for both rated current and breaking capacity (I_cu).
• Every feeder checked for ampacity, voltage drop (≤5%) and short-circuit withstand.
• Protection settings coordinated from the transformer down to the last motor, including inrush.

Send us your single-line diagram or load list (kW per feeder, power factor, voltage, distance to the furthest load) and we will respond with the transformer rating, the short-circuit levels, recommended switchgear and feeder sizes — and a binding FOB Qingdao quotation for the whole package. One drawing in, one coordinated system out.