Heat Rejection Is Not Secondary Loop Heat

Buy the outdoor plant on 11,785 kW of secondary-loop duty and the order is short before a single dry cooler is named. Four terms make up the heat the primary side has to put into the atmosphere, and the secondary loop duty is only the first of them: 11,785 kW at a liquid-cooled fraction of 0.95, which is 11,400 kW of IT heat plus 385 kW of secondary pump work, plus 600 kW of air-side residual heat, plus 100 to 180 kW of primary pump work, plus an unquantified share of power-room electrical loss [MODELED]. The heat-rejection basis forbids the substitution. Sizing primary rejection plant on the secondary figure understates the duty by about 6.5 %, and the author classes that as conclusion-reversing rather than as rounding.

Secondary Duty Omits Air-Side, Primary-Pump, and Power-Room Heat

The secondary figure is a handover, not a plant size. It is the liquid-loop heat at the coolant-distribution units, including secondary pump work taken with the hydraulic heat factor at 1.00. It does not contain the air-side residual that must not be taken as zero. It does not contain primary pump shaft work that enters the outdoor fluid in full. It does not contain the share of switchroom and uninterruptible-supply loss that leaves through mechanical cooling.

The check basis taken against every route is 12,600 kW [TARGET]. That is a placeholder rounded above the first-three-term subtotal, to be revised upward once the power-room share is known. It is not a computed rejection duty. It is not a confirmed plant rating.

A 6.5 Percent Gap Sits Inside the Route-Margin Band

(12,600 − 11,785) / 12,600 = 815 / 12,600 = 0.0647, which the source rounds to 6.5 % [MODELED]. Read the other way, the secondary figure is 93.5 % of the rejection basis.

Location rather than magnitude is why the author flags it. Six and a half per cent lands inside the margin band that each rejection route is judged against in the same chapter, so the error does not present itself as an error. It presents as a route clearing its screen by a couple of kelvin when it would not have. Neither the 12,600 kW basis nor the routes it sizes has been checked against a machine.

Power-Room Share Is Open; Air-Side Heat May Not Be Zero

The power-room share is logged as an open item and assigned jointly to the Engineer of Record and the building scheme, because how much of 786 kW of switchroom and uninterruptible-supply loss leaves the building through mechanical cooling depends on how those rooms are built. Until that closes, the rejection basis has a floor of 12,485 kW and no ceiling — a reading of the subtotal rather than a phrase the source uses.

Opposite in kind is the air-side term. Six hundred kilowatts is an assumption inherited from the hydraulic basis, and the hydraulic basis attaches an instruction to it: design the air-side terminals independently, and do not take the figure as zero. At a liquid-cooled fraction of 0.90 the same residual is 1,200 kW [MODELED]. A term small enough to round away is protected by an explicit prohibition on rounding it away.

Liquid Fraction Does Not Size Plant; It Changes Free Cooling

Secondary heat rises with the liquid-cooled fraction. Total primary rejection is close to insensitive to it, because heat the liquid loop does not take is rejected on the air side instead, and both paths end at the same atmosphere. No sensitivity coefficient is given; the insensitivity is stated qualitatively.

The liquid-cooled fraction is therefore not this chapter's blocking interface for plant sizing. It governs how the duty splits between liquid and air terminals, not how much there is to reject. That is why the rejection chapter can size outdoor plant while the fraction is still argued elsewhere — a reversal of the priority order the hydraulic basis set.

The same fraction is a blocker for the annual energy integral. It does not change how much heat reaches atmosphere. It changes the temperature grade of that heat, and therefore the hours in which free cooling is available. The two conclusions sit together. They must not be quoted as one.

Conflicting Vendor Ratings Leave Secondary Duty Unsettled

An independent review of the vendor packet makes the 11,785 kW less settled than its three significant figures suggest. Two documents of the same date name the same heat-exchanger model with different item identifiers and different rating points: one at 67.3 and 100.1 m³/h with pressure drops of 69.2 and 81.1 kPa, the other at 52.2 and 77.6 m³/h with 42.76 and 50.30 kPa. Estimated on the first document's fluid properties, the second's duty point corresponds to roughly 700 kW rather than 900 kW. The reviewer is careful about what that does and does not establish: it is not sufficient to conclude the delivered unit can only do 700 kW, but the supply record is not closed. Two drawings are being compared here, not two machines. That packet is a five-cabinet, dual-900 kW object. It does not freeze the design-package unit count, and the design-package count does not freeze the packet.

That matters because the secondary figure is built from a unit-count branch and a per-unit rating. If the rating point is unsettled, so is the term the rejection basis inherits, and the 6.5 % gap sits on top of an input that may itself move.

If the power-room item returns a large share, the 12,600 kW is revised upward and every route margin in the chapter narrows; if it returns near zero, the basis settles near the 12,485 to 12,565 kW subtotal. The author commits to the revision in advance rather than to a value, which is the correct disposal of a term nobody can yet compute.

Limits and open items

Modelled and not frozen: the 11,785 kW secondary duty, its 11,400 plus 385 kW composition, the 600 kW air-side residual, the 100 to 180 kW primary pump band, the 786 kW of which the power-room share is taken, the 6.5 % understatement, and the 700 against 900 kW rating-point estimate. [MODELED]

The 12,600 kW check basis remains a placeholder recommendation. [TARGET]

Open: the power-room share; the liquid-cooled fraction, which is a calculation value at 0.95 and a sensitivity at 0.90 and is OEM in neither case; the primary pump duty and the hydraulic heat factor behind the 100 to 180 kW band; and the vendor rating point that feeds the first term. [HOLD]

No rejection plant is selected. No unit count is released. No Engineer of Record has reviewed these quantities.

The question that can be put in front of the building-scheme meeting is narrower than the chapter: what fraction of the 786 kW is designed to leave through mechanical cooling?

© 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., “Heat Rejection Is Not Secondary Loop Heat,” https://www.kkdatasvc.com/lab/heat-rejection-and-pue/heat-rejection-is-not-secondary-loop-heat/.