Lookup R-value

Retrieve an assembly R-value or U-factor — wall / roof / floor / slab / fenestration — from a source the user authorizes. No internal envelope values; every output row cites a verbatim quote with a SourceURL.

For PASS / FAIL math against IECC prescriptive tables, use /test-energy or /test-glazing. For interpretation use /explain-u-value.

Inputs the command needs

1. Assembly type — name one:
   • Wall — mass, metal-framed, wood-framed, framed + continuous insulation (CI)
   • Roof — insulation entirely above deck, metal building, attic and other
   • Floor — mass, steel-joist, wood-framed
   • Slab on grade — perimeter R-value and depth (F-factor)
   • Below-grade wall — CI rating and depth
   • Fenestration — U-factor / SHGC / VT (use /lookup-fire-rating analog: NFRC-rated whole-window)
2. Climate zone (CZ) — 1A, 2A, 2B, 3A–3C, 4A–4C, 5A–5C, 6A–6B, 7, 8. Optional but narrows the search.
3. Specifics — framing factor (e.g., 2x6 @ 16 in o.c.), insulation type (R-21 mineral wool batt, R-7.5 polyiso CI), framing material (steel, wood).

Step 1 — Source preference (always ask first)

Tell the user, all at once:

"Where should I retrieve the R-value / U-factor from? (a) Standard — https://basc.pnnl.gov/ (DOE Building America Solution Center, free) for whole-assembly performance and best-practice details. I will WebFetch on confirmation and quote the verbatim assembly summary. (b) Your own source — paste from ASHRAE 90.1-2022 App. A (per-assembly U-factor calculations), a manufacturer spec sheet (Owens Corning, Knauf, Rockwool, Dow, Henry), an arch-rag-indexed envelope PDF, or a project basis-of-design. (c) Both — fetch DOE Building America AND check your source; report any mismatch.

Pick (a), (b), or (c)."

The command does not proceed without an explicit pick.

Step 2 — Acquire source

• (a) Standard fetch — WebFetch https://basc.pnnl.gov/ with the user's keywords (assembly + CZ + framing). Quote the verbatim assembly summary including R-value, U-factor, framing factor, and any notes on thermal bridging. Record fetched URL and ISO timestamp.
• (b) Own source — accept paste, run arch-rag query, or WebFetch a user-provided URL on confirmation. Quote verbatim and record source.
• (c) Both — execute (a) and (b); present both quotes; flag any divergence.

If DOE Building America has no entry for the queried assembly + CZ, fall back to: WebSearch <assembly> <CZ> R-value site:basc.pnnl.gov OR site:nrel.gov OR site:energy.gov → ask user to confirm match → ask user to paste from ASHRAE 90.1 App. A or a manufacturer spec.

Step 3 — Output

# Lookup R-value — [Assembly, CZ]

**Assembly:** [2x6 wood-framed wall + R-5 CI / Roof insulation entirely above deck / Slab perimeter]
**Climate zone:** [4A — NYC / 3B — LA / unspecified]
**Specifics:** [framing, insulation type, layers]
**Source preference:** [(a) DOE Building America / (b) own source — name / (c) both]

## Reference area
DOE Building America Solution Center (`https://basc.pnnl.gov/`); `ASHRAE 90.1-2022 App. A` (assembly U-factor calculations); manufacturer spec sheets for product-specific values; `IECC 2024` for prescriptive minimums (use `/test-energy` for PASS / FAIL).

## Source quote
> [Verbatim assembly summary — layer-by-layer description, total R-value, U-factor, framing factor, thermal bridging notes.]

(Source: WebFetched from `https://basc.pnnl.gov/...` on <YYYY-MM-DD> / pasted by user from <doc> / arch-rag query against `<indexed PDF>`)

## Assembly summary table
| Field | Value (per quoted source) |
|---|---|
| Assembly type | "<verbatim>" |
| Total R-value (effective) | "<verbatim>" |
| Total U-factor | "<verbatim>" |
| Framing factor applied | "<verbatim, e.g., 23% for 2x6 @ 16 in o.c.>" |
| Layers (verbatim) | "<...>" |
| Climate zone applicability | "<verbatim, if stated>" |
| Source date | "<...>" |

## Divergence (if (c))
- [Bullet — any R / U / framing-factor difference between DOE Building America and own source.]

## Caveats
- This is a retrieval, not a compliance check. The MEP Engineer / Architect of Record select and stamp the assembly; this command does not verify against `IECC 2024` minimums (use `/test-energy`).
- Effective R-value depends on framing factor and thermal bridging through fasteners. Cavity-only R-value overstates assembly performance — use the source's framing-factor-aware value.
- Manufacturer spec sheets list center-of-cavity R-values that ignore framing — verify the quoted value is whole-assembly.
- For fenestration U-factor / SHGC / VT, NFRC whole-window ratings only; center-of-glass values are not equivalent.
- Binding interpretation is the AOR / MEP Engineer stamp plus AHJ review (`NYC DOB`, `LADBS`, `DSA`, `CA HCAI`, `Toronto Building`).

NA AEC conventions (enforce)

• English only.
• R-values lead, U-factors second. Imperial-first throughout. Never lead with W/m²K.
• Climate zones: ASHRAE 169 / IECC 2024 Ch. 3 — 1A … 8.
• Codes / sources: IECC 2024, ASHRAE 90.1-2022, ASHRAE 90.2-2018, DOE Building America, NFRC, NBC 2020 Sec. 9.36, OBC SB-10, CA Title 24 Part 6. Reject KBC, KS, EN, BS, JIS, DIN.
• AHJs named: NYC DOB, LADBS, DSA, CA HCAI, Toronto Building, Boston ISD, Chicago DOB. Never the AHJ or the city.
• Floors: Level 3 / L3 / 3rd floor. Never 3F.
• Roles: Architect of Record (AOR), MEP Engineer, Energy Modeler, Commissioning Authority (CxA). Never Director as a project role.

Constraints

• Never invent R-values or U-factors from recall.
• Never paraphrase the source summary — verbatim only.
• Always ask the source-preference question at Step 1; do not skip.
• Never WebFetch from manufacturer marketing pages without explicit user authorization — DOE Building America or user-named source only.
• Always include the SourceURL + timestamp.
• Always flag the framing-factor caveat when the user provides cavity-only R-value.
• Caveat block always present.
• Markdown only. No Word / PDF.

Anti-patterns (never do)

• Stating "R-30 batts + R-5 CI = R-25 effective" from recall — never. The effective value comes from ASHRAE 90.1 App. A, DOE BA, or manufacturer assembly testing.
• Adding cavity R + CI R linearly — non-additive due to framing factor.
• Treating center-of-glass U-factor as NFRC whole-window.
• Auto-fetching without explicit user authorization.
• Skipping Step 1 (source preference question) — never.
• Running /test-energy from inside this command — pointer-only.