Heat-Rejection Weather File Was Built for the Wrong State

The heat-rejection chapter issued a change-candidate requiring the cooling-hydraulics chapter to stop naming the correct state. A later revision reversed the arrow. The heat-rejection chapter was the one that had taken a utility-substation designation as a town in another state, built the weather file there, and then instructed the file that had the geography right to conform. The revision now carries every weather-derived quantity as void for this site: design dry bulb, design wet bulb, approach, annual temperature frequency, the heat-rejection capability checks that used them, and the annual energy figures built on top. [FACT]

Heat Rejection, Not Cooling, Named the Wrong State

The cooling chapter had written the correct state's name against the summer extreme. The heat-rejection chapter had written the substation town as the site, and from that town it took climate, code path, structural loads, and serving-utility territory. Its first issue did not treat this as its own error. It registered a change-candidate requiring the cooling chapter to amend the place name. The later revision is explicit: the direction was opposite, and the heat-rejection chapter was the one that was wrong.

The geography finding and the weather recomputation are not the same close. The incremental-closure record that followed the red-team review split the item. The site-fact segment is recorded as closed. The weather and energy recomputation remains open. Closing the geography line does not publish an energy number.

The Wrong Place Still Returned Four Valid Answers

Nothing in the merge steps objected. A place name is a valid string wherever it appears, so the weather tool returned a well-formed file, the structural parameters came back as plausible numbers, the code path resolved to a real authority, and the interconnection discussion had a real counterparty. Four wrong answers, all of them well formed. The error becomes visible only where a person compares the site against the document — at Engineer of Record review, or at submission to the authority having jurisdiction — by which point the derived values have already been cited elsewhere as established.

Struck through, not deleted

The disposition chosen here is more instructive than the error. The revision marks the whole chain dead in place instead of recomputing it. Every affected value stays visible, struck through and annotated as belonging to the wrong chain and retained for traceability. Deleting the values would be tidier and worse: a downstream document that had cited a design wet bulb would no longer be able to find what it cited, and the citation would survive as an orphan with nothing left to test it against. Retention with annotation makes each downstream reference fail loudly rather than quietly.

Splitting the item keeps the record honest about who is able to close it. Declaring the chain invalid sits with the author of the design document and can be closed now. Recomputation does not sit there. It needs the site coordinates and elevation from the owner and the surveyor, a named weather station or climate dataset entry with its version, the selected design percentile together with its coincident quantities, and a confirmed list of every derived result inside the scope. None of those four inputs has been obtained, so the item is recorded as partially closed and may not be written up as closed.

Five named gates sit in front of the recomputation. Until they are all present and accepted by the Engineer of Record, inventing replacement dry-bulb and wet-bulb figures is forbidden. Translating the wrong-state file because the climates are "close" is forbidden in the same breath. Translating it because both sites sit in the same climate zone is forbidden with it. A residual paragraph in the heat-rejection chapter still names the wrong-state climate after the geography correction. That leftover is a reminder that a stamp on the site line does not rewrite every sentence that used it.

The Formulae Survive; No Value Has Site Validity

The formula structure survives. Band identifiers for annual, summer-design, and extreme conditions survive with it. The relationships between rejection duty, dry bulb, wet bulb and approach remain checkable, and the arithmetic inside them is still reproducible. No value substituted into them carries site validity. On the current record the project has no confirmed heat-rejection design condition at all. [FACT]

A conclusion drawn from those formulae means only that the arithmetic is reproducible under placeholder inputs. It does not mean the conclusion holds for this project. Capability checks, route vetoes, and annual energy figures that used the dead chain keep their algebra and lose their site.

Limits and open items

Modelled inputs and everything derived from them are void for this site. The formula structure survives. Open: the site coordinates and elevation; the named station or dataset entry and its version; the design percentile; the scope list of affected derived results; and the recomputation itself, which belongs to the Engineer of Record. No temperature, percentile, station identifier or place name appears in this note, because republishing the invalidated values would propagate the same error this note describes. [HOLD]

No recomputation has been performed, no corrected design condition exists, no heat-rejection equipment is selected, and no Engineer of Record has reviewed any of it. The geography correction is not a licence to publish an energy number.

Which Proper Nouns Carry No Type, Asset, or Place

The naming lesson is narrower than "check your inputs". What failed was not a weather file. It was a proper noun carrying an asset meaning in one discipline and a geographic meaning in another, crossing the boundary between them without ever being marked as either. A data register that records what kind of thing each proper noun is, asset or jurisdiction or place, would have caught this at the first merge instead of at review. The question worth putting to the next project's register is which of its proper nouns currently carry no type at all.

© 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 Weather File Was Built for the Wrong State,” https://www.kkdatasvc.com/lab/heat-rejection-and-pue/we-built-the-weather-file-for-the-wrong-state/.