Worked Example: Medium Voltage Cable Sizing for a Hospital Feeder — The Gorakhpur Tragedy

In August 2017, at BRD Medical College in Gorakhpur, Uttar Pradesh, India, over 63 children died in the paediatric and neonatal intensive care wards over a five-day period. The immediate cause was the failure of the hospital’s liquid oxygen supply, but the chain of events that led to the catastrophe revealed a deeper infrastructure crisis — including the degradation of the hospital’s 11 kV medium voltage feeder cable.

The MV cable supplying BRD Medical College had been repeatedly repaired with substandard joints over the years. As the hospital expanded — adding ICU beds, NICU facilities, and additional wards — the electrical demand grew well beyond what the original cable had been designed to carry. Monsoon conditions caused waterlogging of the direct-buried cable trenches, dramatically increasing the soil thermal resistivity around the cable and reducing its effective current rating. When the cable finally failed under sustained overload during the monsoon, the backup diesel generator could not start due to a separate maintenance failure, and the oxygen concentrator plant lost power entirely.

This tragedy underscores a critical but often overlooked aspect of hospital electrical design: the medium voltage feeder cable is the single point of failure for the entire facility. IEC 60287 provides the methodology for calculating continuous current ratings of MV cables under real-world thermal conditions — not just the idealised conditions printed in manufacturer catalogues. This worked example demonstrates how to properly size an 11 kV hospital feeder cable, accounting for tropical soil conditions, burial depth, and the thermal environment that ultimately determines whether the cable survives or fails.

Size an 11 kV XLPE medium voltage feeder cable for a 500-bed hospital with expanding demand.

For a three-phase system, the design current from the hospital demand:

I_b = S / (√3 × V) — (Eq. 1)

I_b = 2,500,000 / (√3 × 11,000)

I_b = 2,500,000 / 19,053

I_b = 131.2 A

Per IEC 60502-2:2014, select a 3-core 11 kV XLPE cable with copper conductors. The candidate sizes and their base current ratings (in air at 30°C, from manufacturer data per IEC 60287 reference conditions) are:

The base ratings above assume standard conditions: soil thermal resistivity of 1.0 K·m/W, ground temperature of 20°C, and burial depth of 0.8 m. Our actual conditions differ significantly — we must apply correction factors per the IEC 60287 thermal circuit model.

The soil thermal resistivity is the single most important environmental factor for buried MV cables. Per IEC 60287-3-1:1999, Clause 4, the correction factor for soil thermal resistivity different from the standard 1.0 K·m/W:

The thermal resistance of the surrounding soil in the IEC 60287 thermal circuit:

T₄ = (ρ_soil / 2π) × ln(2L / D_e) — (Eq. 2)

Where ρ_soil = soil thermal resistivity (K·m/W), L = burial depth (m), D_e = external diameter of cable (m).

For a 95 mm² 3-core 11 kV XLPE cable, D_e ≈ 0.055 m. At standard conditions (ρ = 1.0, L = 0.8 m):

T_4,std = (1.0 / 2π) × ln(2 × 0.8 / 0.055) = 0.159 × ln(29.1) = 0.159 × 3.37 = 0.536 K·m/W

At our actual conditions (ρ = 0.7, L = 1.0 m):

T_4,actual = (0.7 / 2π) × ln(2 × 1.0 / 0.055) = 0.111 × ln(36.4) = 0.111 × 3.59 = 0.399 K·m/W

The waterlogged soil (ρ = 0.7) actually improves heat dissipation compared to standard dry soil (ρ = 1.0). However, the critical danger is when the water table drops — dry alluvial soil can reach ρ = 2.0–3.0 K·m/W, dramatically reducing the cable’s rating.

T_4,design = (2.5 / 2π) × ln(36.4) = 0.398 × 3.59 = 1.429 K·m/W

The standard ground temperature for IEC 60287 base ratings is 20°C. At 35°C ground temperature (tropical India), the temperature correction factor per IEC 60287-3-1, Table 2:

k_θ = √((θ_max − θ_ground) / (θ_max − θ_ref)) — (Eq. 3)

Where θ_max = 90°C (XLPE maximum operating temperature), θ_ground = 35°C, θ_ref = 20°C:

k_θ = √((90 − 35) / (90 − 20))

k_θ = √(55 / 70)

k_θ = √(0.786)

k_θ = 0.886

The high ground temperature alone reduces the cable’s current rating by over 11% compared to temperate conditions.

Deeper burial increases the thermal resistance between the cable and the ground surface, reducing the cable’s ability to dissipate heat. The correction factor for 1.0 m depth versus the standard 0.8 m per IEC 60287-3-1, Clause 4.2:

k_d = ln(2 × L_ref / D_e) / ln(2 × L_actual / D_e) — (Eq. 4, simplified)

k_d = ln(2 × 0.8 / 0.055) / ln(2 × 1.0 / 0.055)

k_d = ln(29.1) / ln(36.4) = 3.37 / 3.59

k_d = 0.939

The deeper burial reduces the rating by approximately 6%.

Combining the soil resistivity correction (using design worst-case ρ = 2.5), ground temperature correction, and depth correction. The overall derating relative to the standard direct-buried rating:

For the soil resistivity effect, we apply the ratio of thermal resistances. The combined correction factor for soil resistivity versus the standard 1.0 K·m/W condition:

k_ρ ≈ √(T_4,std / T_4,design) = √(0.536 / 1.429) = √(0.375) = 0.613 — (Eq. 5, simplified)

Combined derating factor:

k_total = k_θ × k_d × k_ρ = 0.886 × 0.939 × 0.613 — (Eq. 6)

k_total = 0.510

This is a dramatic derating — the cable retains only 51% of its catalogue rating under these conditions. Now determine the required cable size:

I_z,required = I_b / k_total = 131.2 / 0.510 = 257.3 A — (Eq. 7)

At 11 kV, voltage drop is calculated using the impedance method per IEC 60364-5-52, Clause 525. For 150 mm² 3-core XLPE copper cable at 50 Hz:

From cable data (per IEC 60228): R_ac = 0.124 Ω/km, X_L = 0.078 Ω/km.

ΔV = √3 × I_b × L × (R_ac cosφ + X_L sinφ) — (Eq. 8)

ΔV = √3 × 131.2 × 0.8 × (0.124 × 0.85 + 0.078 × 0.527)

ΔV = 1.732 × 131.2 × 0.8 × (0.1054 + 0.0411)

ΔV = 181.7 × 0.1465

ΔV = 26.6 V

ΔV% = 26.6 / 11,000 × 100 = 0.24%

The allowable voltage drop for an MV feeder is typically 5% per IEC 60364-5-52, Annex G. At 0.24%, voltage drop is negligible for this MV cable. This is typical for MV systems — current capacity and thermal rating almost always govern the cable size, not voltage drop.

The cable conductor must withstand the prospective fault current for the protection clearing time without exceeding the short-circuit temperature limit. Using the adiabatic equation per IEC 60949:1988:

S_min = I_sc × √t / k — (Eq. 9)

Where I_sc = 12,500 A (prospective fault), t = 0.5 s (clearing time), k = 143 (copper conductor, XLPE insulation, initial 90°C, final 250°C per IEC 60364-5-54, Table 54.4).

S_min = 12,500 × √0.5 / 143

S_min = 12,500 × 0.707 / 143

S_min = 8,838 / 143

S_min = 61.8 mm²

Our selected 150 mm² cable far exceeds the minimum 61.8 mm² required for short circuit withstand. ✓ PASS

The cable screen (copper wire screen) must carry earth fault current until the protection clears. For a 150 mm² cable, a typical screen cross-section is 16 mm² copper wire screen. Using the same adiabatic equation for the screen:

I_screen,max = k_s × S_screen / √t — (Eq. 10)

Where k_s = 143 (copper screen, initial 90°C, final 250°C), S_screen = 16 mm², t = 0.5 s:

I_screen,max = 143 × 16 / √0.5

I_screen,max = 2,288 / 0.707

I_screen,max = 3,236 A

The earth fault current through the screen depends on the earthing system. For a solidly earthed 11 kV system, earth fault current can approach 60–70% of the three-phase fault level. At 70%:

I_ef = 0.7 × 12,500 = 8,750 A

8,750 A > 3,236 A — the 16 mm² screen is insufficient. Solutions: specify a cable with a larger screen (e.g., 35 mm²), or use resistance earthing to limit earth fault current to below 3,236 A.

Selected cable: 150 mm² 3-core 11 kV XLPE, copper conductor, with upgraded 35 mm² copper wire screen. The governing factor is thermal rating under worst-case dry soil conditions (ρ = 2.5 K·m/W). The dramatic 49% derating means a cable that appears generously oversized on paper (catalogue 270 A vs 131 A load) is in reality only marginally adequate.

The Gorakhpur tragedy was caused by a chain of failures — oxygen supply management, payment disputes, generator maintenance, and infrastructure neglect. For the MV cable component specifically:

• Size MV cables for worst-case soil thermal resistivity, not average conditions — in regions with seasonal water table variation, the dry-season soil resistivity can be 3–4× the wet-season value; using catalogue ratings or monsoon-condition ratings leads to chronic overheating during summer
• Include a minimum 20% growth factor for hospital feeders — hospitals expand incrementally (a new ICU here, additional wards there), and each expansion adds load to the same feeder; the cable must be sized for the anticipated demand, not just today’s load
• Never use ad-hoc cable joints on MV circuits — MV cable joints must be factory-made heat-shrink or cold-shrink types installed by certified jointers; field-improvised joints are the leading cause of MV cable failure worldwide
• Implement redundant feeders for critical facilities — hospitals, data centres, and water treatment plants should have two independent MV feeders with automatic changeover, so that a single cable failure does not result in total loss of supply
• Conduct periodic thermal imaging and load monitoring — infrared surveys at MV cable terminations and joints can detect hot spots before they become faults; continuous load monitoring reveals when a cable is approaching its thermal limit