Even Facility-Load Split Exceeds the 7 MW Feeder Allocation

In normal operation — not after a fault — an even split of the 14.391 MW [MODELED] facility demand puts 7.196 MW [MODELED] on each of the two utility feeders. The smaller feeder is allocated 7 MW [TARGET], so it is already 0.196 MW [MODELED] over its allocation while the plant is healthy. The larger feeder is allocated 8 MW [TARGET]. Symmetric 50/50 loading is therefore unavailable, and mechanical and house load has to be biased toward the 8 MW side.

That result is an allocation exceedance. It is not a demonstrated thermal overload of the smaller feeder.

Halving Facility Demand Exceeds the 7 MW Allocation

Facility demand here is a summation, not a reading. Critical medium-voltage demand of 13.317 MW [MODELED], mechanical and house demand of 1.031 MW [MODELED], and 0.043 MW [MODELED] of medium-voltage switchgear and cable loss close at 14.391 MW [MODELED]. Halving that total gives 7.1955 MW, carried as 7.196 MW [MODELED]. The difference against the 7 MW allocation is 0.196 MW [MODELED].

The Gate 1 closure record accepts that a symmetric allocation is unavailable and defers the allocation itself to a load-flow study [FACT]. The later conditional-freeze list writes the same rule in operating language: bias the normal state toward the 8 MW side; do not run 50/50.

The Even Split Stops Exceeding at 14 MW

One threshold is exact rather than approximate. An even split stops exceeding the smaller allocation at a facility demand of 14.000 MW [MODELED], because half of that figure is precisely the 7 MW contract value.

At the current 14.391 MW basis the exceedance is 0.196 MW. At the 14.125 MW [MODELED] steady-state figure from the later heat-rejection recomputation it falls to 0.0625 MW [MODELED] — one third of the magnitude, same direction. The gap between the finding and its own expiry is therefore 391 kW [MODELED] of facility demand, which is smaller than several single line items in the load tree.

If the mechanical reconciliation lands at the low end and verified facility demand settles at or below 14.000 MW, the even split stops exceeding the allocation and this finding expires on its own.

Mechanical Load Is Still Moving in Both Directions

A cooling-hydraulic cross-check states directly that the arithmetic premise behind 0.196 MW is no longer complete, because the mechanical load's absolute value and its distribution are both moving.

Reverse-calculating secondary pumping from the bottom-up pressure-drop ledger yields 46 kW [MODELED] with twelve pumps in parallel and 105 kW [MODELED] with eight pumps at full load under N−1. The protected-load tree still carries 385 kW [MODELED] for cooling-distribution pumping, drives and controls. That is 280 to 339 kW lower [MODELED]. The difference would push the required bias down. The same record forbids deducting it on its own.

Two items push the other way. Air-side terminal heat rejection, at 600 kW [SENSITIVITY] where the liquid fraction is 0.95 and 1,200 kW [SENSITIVITY] where it is 0.90, is not explicitly identified in the facility figure. It may or may not already sit inside the 585 kW [MODELED] auxiliary allowance; the documents do not answer that. If a thermal ride-through requirement materialises, UPS-backed secondary pumping adds a further 46 to 105 kW [SENSITIVITY]. The net effect can run either way. The cooling basis prohibits claiming any margin improvement until the reconciliation closes.

The Bias Binds Assignment, Alarms, Studies, and Procedures

Biasing load toward the larger feeder is a written engineering setting, not an operations preference. It binds four specific things: which feeder the mechanical and house transformers are assigned to, the alarm setpoint the power monitoring system uses on the smaller feeder, the load-flow study scope, and the wording of the normal-operation section of the operating procedures.

None of those four can be written from an A/B one-line. The one-line shows two sides of equal weight. The operating point does not.

Whether 7.196 MW would thermally overload anything on the smaller feeder is a separate question, and it is not answered here. Allocation is a commercial and operational limit set by the service arrangement. Thermal loading depends on the conductor and protection ratings behind that arrangement, which are not on this record. Collapsing the two is the error that makes a 0.196 MW figure look either alarming or trivial, depending on which reading the reader happens to bring.

Limits and open items

Confirmed on the current record: the two allocations are 8 MW and 7 MW; the closure record accepts in principle that symmetric allocation is unavailable and biases load toward the larger feeder. The finding concerns the normal operating point rather than a fault case, and it is an allocation exceedance, not a demonstrated thermal overload. [FACT]

Modelled and unfrozen: 14.391 MW facility demand, 7.196 MW per side, the 0.196 MW exceedance, and the 14.125 MW recomputation with its 0.0625 MW exceedance. Every one of these inherits an unverified item-by-item load schedule, and the auxiliary allowance, end-of-path loss and UPS efficiency inputs behind it are assumptions. The 14.125 MW figure is a change candidate from a later heat-rejection itemisation; it does not rewrite the shared facility-demand basis. [MODELED]

Requiring resolution before the allocation can be set: the mechanical reconciliation named above, the load-flow study the closure record defers to, the item-by-item controlled load schedule, and the serving utility's rating basis for the two allocations. [HOLD]

Nothing above has been reviewed or sealed by an Engineer of Record, and no equipment quantity is released.

At What Facility Demand Does the 7 MW Feeder Stop Binding

What belongs in the load-flow scope now is not "is a 50/50 split acceptable", which the next load-schedule revision can silently invalidate. It is "at what facility demand does the smaller feeder stop being the binding side". Which of those two questions is currently written into your study scope?

© 2026 K&K Data Service Inc. All rights reserved. Reproduction or republication is permitted only with clear attribution to K&K Data Service Inc. and a working hyperlink to the canonical URL of this article. Excerpts must preserve the technical context, maturity labels, assumptions, and limitations. No excerpt may imply project approval, field validation, certification, or endorsement that the original article does not state.

Source: K&K Data Service Inc., “Even Facility-Load Split Exceeds the 7 MW Feeder Allocation,” https://www.kkdatasvc.com/lab/electrical-architecture-and-failure-domains/why-a-50-50-split-overloads-the-weak-feeder/.