The CDU RFQ Freezes a Four-Tuple Rated Capacity It Never Defines

The procurement package names a four-tuple rated capacity as a defining parameter of the cooling-distribution package, then uses the same undefined noun as a blocker on that package's long-lead row. A full-text search of the file returns those two appearances and no expansion of the four members. The technical card for the same equipment answers the same question with six numbered rows, not four [TARGET]. Neither passage reconciles with the other. So the package's own instruction, that every defining parameter be frozen and ruled in writing before a technical enquiry is issued, currently has no written object on the one package its long-lead register treats as interface-blocked.

The RFQ Names Four Members; the Card Lists Six Parameters

Defining has a written test here, and the test is not about precision. A parameter is defining if changing it would force a supplier to reorganise any one of five things [TARGET]:

• the product family;
• the production line;
• the redundancy configuration;
• the listing scope;
• the transport arrangement.

Applying that test, the package states the answer for two items and states it as a pair each time. For the cooling distribution units the defining parameters are unit count and the four-tuple rated capacity, expressly not pressure-drop precision. For the generator package they are nameplate power and unit count, expressly not fuel consumption rate. One conclusion names the failure in both directions: treating a precision question as a defining one delays issue for nothing, while treating a defining question as a precision one voids every quotation after issue.

Six defining parameters are then listed on the equipment card: rated heat transfer with its secondary flow and temperature difference; the working fluid; design pressure and temperature on each side; redundancy configuration; the wetted-material and fluid compatibility family; and the controller protocol family with its fail position. Four of the inputs those rows depend on are recorded in the same card as having no obtained value, which is why the card forbids any single design point appearing in its performance row at all. The long-lead register uses the term as a bare noun beside the interface specification and the owner's redundancy branch, which is its second load-bearing use and its second appearance without a definition.

Two Documents Under One Model Name Carry Different Ratings

A supplier review dated 15 September 2026 shows why the missing definition is not a drafting complaint. Two documents for one model designation, both dated the same day, carry different item identifiers and different rating points: one at 67.3 and 100.1 cubic metres per hour with pressure drops of 69.2 and 81.1 kilopascals, the other at 52.2 and 77.6 cubic metres per hour with 42.76 and 50.30 kilopascals [TARGET]. Working from the first document's fluid properties, the reviewer computes that the second set of temperatures and flows corresponds to roughly 700 kilowatts against the 900 kilowatts the first was selected for [MODELED]. The stronger inference is then refused in the same paragraph: that is not sufficient to establish that the physical unit can only do 700 kilowatts, because the two may be selection points of one plate type. The supply record is closed in neither direction.

Holding a model name constant across two documents did not hold the rating constant, which is the shape of the failure the four-tuple exists to prevent. The two documents differ in flow and in pressure drop at once, one of which the package calls defining and the other of which it tells everybody not to freeze. A bidder comparing them would be comparing scopes, not prices.

The Procurement Evidence Register Has Columns and No Rows

Nothing has been recorded against any of it. Ten columns make up the procurement evidence register for this project, running from evidence identifier and object through purchase order, document digest, approval, approver role, release date and release quantity to status and source reference, and the register carries no rows [TARGET]. The freeze the package demands has therefore never been performed on any item, so no quotation has yet been received against an undefined parameter. Drafting the freeze object is not sending the enquiry. The send-out column on every package remains forbidden.

Naming the four quantities in the card's first row would give the freeze a written object, and that row already carries candidates for them. Rows five and six of the same card, the wetted-material compatibility family and the controller protocol family, have no named adjudicator anywhere in the package, while the quantity table two chapters earlier names the owner and the main design as the parties that rule the same package's unit count.

Limits and open items

The four-tuple is named twice and defined in neither place. The card that answers the same question is numbered one to six. Neither count is a freeze. [TARGET]

Study values, not a rating: the 67.3 / 100.1 and 52.2 / 77.6 cubic-metre-per-hour points; the 69.2 / 81.1 and 42.76 / 50.30 kilopascal drops; and the reviewer's 700 kilowatt figure against a 900 kilowatt selection. The two documents may be two selection points of one plate type. [TARGET] [MODELED]

Open: the four members of the tuple; the four unobtained card inputs; an adjudicator for the parameter list; manufacturer reconciliation of the two documents; any row in the procurement evidence register. No enquiry has been issued. Nothing here is a purchase quantity. Nothing is a listing conclusion. [HOLD]

Name the Four Quantities the Long-Lead Row Is Waiting On

The useful test is not whether the card is complete. Write down the four quantities the long-lead row is waiting on. If the list has six entries, or none, the freeze still has no object.

© 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., “The CDU RFQ Freezes a Four-Tuple Rated Capacity It Never Defines,” https://www.kkdatasvc.com/lab/rfq-vendor-review-and-commissioning/the-rfq-freezes-a-four-tuple-it-never-defines/.