Worked Example: Maximum Demand Calculation for a Commercial Building — The 1998 Auckland CBD Power Crisis

In February 1998, Auckland’s central business district lost power for five weeks. Not five hours, not five days — five weeks. Four high-voltage cables feeding the CBD failed in succession over ten days, leaving 6,000 businesses and 4,000 inner-city apartments without electricity. Diesel generators lined the streets. Office workers decamped to suburban locations. The economic cost exceeded NZ$80 million.

The root cause was overloaded cables. Auckland’s CBD had been growing rapidly — new office towers, server rooms, air conditioning loads that didn’t exist when the cables were installed in the 1950s and 1960s. The cables were operating above their rated capacity for years, and the excess heat slowly degraded the oil-paper insulation. When summer temperatures pushed the soil temperature above normal, the cumulative thermal damage became catastrophic. One cable failed, redistributing its load to the remaining three. The increased load accelerated degradation in those cables, and they failed in turn.

The Auckland crisis is the definitive case study for why maximum demand calculations matter. Every office tower, data centre, and commercial building depends on an upstream network designed for a specific capacity. If the building’s actual maximum demand exceeds the designed capacity, the result is thermal overload of cables and transformers — not immediately, but insidiously, over months and years, until the insulation fails.

Calculate the maximum demand for a 20-storey commercial office building to determine transformer sizing.

Connected loads by category:

Apply the demand factors from AS/NZS 3000:2018, Appendix C to each load category:

General lighting:

MD_lighting = 240 kW × 0.90 = 216 kW — (Eq. 1)

From AS/NZS 3000 Table C2: for commercial office buildings, the demand factor for lighting is 0.90 (90% of connected load). This reflects that not every light is on simultaneously — some offices are unoccupied, corridors are on occupancy sensors, etc.

General power (socket outlets):

MD_GPO = 500 kW × 0.40 = 200 kW — (Eq. 2)

The demand factor of 0.40 per Table C3 is critical: it reflects that while a typical floor has 100+ socket outlets, most are unused or lightly loaded at any given time. The 40% figure is empirically validated for commercial offices.

HVAC:

MD_HVAC = 800 kW × 0.80 = 640 kW — (Eq. 3)

HVAC is the largest single load in most commercial buildings. The 0.80 demand factor per Table C4 accounts for the fact that chillers cycle and not all air handling units run at maximum simultaneously.

Lifts:

For 6 lifts per AS/NZS 3000 Table C5: largest motor at 100%, second at 80%, third at 60%, remaining at 50%:

MD_lifts = 35 × (1.00 + 0.80 + 0.60 + 0.50 + 0.50 + 0.50) — (Eq. 4)

MD_lifts = 35 × 3.90 = 136.5 kW

Basement services:

MD_basement = 45 kW × 1.00 = 45 kW

Server room:

MD_IT = 200 kW × 0.85 = 170 kW — (Eq. 5)

The IT load demand factor of 0.85 reflects that servers maintain consistently high utilisation, unlike GPO loads. Modern server rooms often have demand factors approaching 1.0.

The raw maximum demand before applying building diversity factor:

Sum of individual maximum demands = 1,407.5 kW — (Eq. 6)

Fire services (75 kW) are standby loads: they are not added to the maximum demand for transformer sizing because they only operate during a fire event when other loads (HVAC, lifts) are typically shut down. However, the main switchboard must be rated for the fire service load in addition to normal loads for the brief period during fire mode transition.

The building diversity factor accounts for the statistical impossibility of all loads reaching their individual maximum demands simultaneously. Per AS/NZS 3000:2018 Table C1, the diversity factor for a building of this type and size:

Diversity factor = 0.80 (for buildings with > 1,000 kW individual MD sum)

MD_building = MD_sum × DF = 1,407.5 × 0.80 — (Eq. 7)

MD_building = 1,126 kW

At 0.85 average power factor (typical for commercial buildings with mixed loads):

MD_kVA = 1,126 / 0.85 = 1,325 kVA — (Eq. 8)

Standard transformer ratings (oil-type, AN cooling): 500, 750, 1000, 1250, 1500, 2000 kVA.

The maximum demand of 1,325 kVA requires a transformer rated at least 1,500 kVA.

Check the loading factor:

Loading = MD_kVA / S_transformer = 1,325 / 1,500 = 88.3% — (Eq. 9)

The main switchboard (MSB) bus bars must be rated for the transformer’s full rated current (not just the calculated maximum demand), plus an allowance for fire services:

I_MSB = S_transformer / (√3 × V) = 1,500,000 / (√3 × 415) — (Eq. 10)

I_MSB = 2,087 A

Selected: 2,500 A rated main switchboard (next standard bus rating above 2,087 A, providing margin for future load growth).

Main incomer MCCB: 2,000 A frame with adjustable electronic trip (set at 2,000 A).

The diversity factor of 0.80 assumes that different load categories do not all peak simultaneously. But modern commercial buildings challenge this assumption:

On the design day (the hottest day in 100, with full building occupancy), the actual diversity factor is 0.89 — higher than the 0.80 standard value. The 1,500 kVA transformer would see:

MD_design-day = 1,407.5 × 0.89 = 1,253 kVA (at PF 0.85)

This is 83.5% of the 1,500 kVA transformer — still within its continuous rating, but with no margin for any additional loads. The 2,000 kVA option looks much more prudent when you consider design day scenarios.

This is precisely what happened in Auckland in 1998: the cables were sized for a standard diversity factor, but the actual load — driven by increasing air conditioning in old buildings not designed for it — exceeded the assumed diversity, and the cables operated above their rated temperature for years until they failed.

Recommendation: 2,000 kVA transformer at 66% initial loading, providing capacity for 50% growth over the building’s 25-year design life. This matches the lesson of Auckland 1998: infrastructure designed at 85%+ loading has no margin for growth, changing technology, or extreme weather — and the consequences of overloading develop slowly and invisibly until catastrophic failure.

The Auckland CBD power crisis was caused by decades of load growth exceeding the cable system’s design capacity, exacerbated by deferred maintenance and monitoring. The engineering lessons:

• Design for growth, not just current load — transformers and cables should be initially loaded at 60–70% to allow for 30–40% growth over their 25–40 year life
• Use design-day diversity, not average diversity — the standard diversity factors in AS/NZS 3000 represent typical conditions, not peak conditions; design-day analysis may reveal 10–15% higher demands
• Monitor actual loads continuously — smart metering at the transformer and major distribution boards allows early detection of load creep before thermal limits are approached
• Revalidate maximum demand when tenants change — a building designed for general office use may be inadequate when a floor becomes a data centre or dense trading floor
• Separate fire service capacity from diversity calculations — fire pumps and smoke exhaust fans are standby loads that should not be included in normal maximum demand, but the switchboard must handle them