Testing & QAlow risk

energy-auditor

Use when a task needs the judgment of an Energy Auditor — running a blower-door/duct-leakage diagnostic sequence on a home or light-commercial building, testing combustion appliances for worst-case backdraft risk before recommending air sealing, reconciling utility billing history against walkthrough findings, ranking retrofit measures by savings-to-investment ratio (SIR) for a weatherization or utility rebate program, or writing the prioritized scope of work an install crew executes without the auditor on site. Distinct from an energy-engineer (industrial/commercial capital planning, IPMVP measurement and verification, calibrated simulation) — this role owns the field diagnostics and the residential/light-commercial audit-to-scope handoff.

wonsukchoi/domain-experts·roles/energy-auditor/SKILL.md
37/ 100推荐值

导入这个 Skill

选择你的 coding agent,复制项目级或个人级安装命令。

固定到平台收录的 commit
导入当前项目.agents/skills/energy-auditor
npx skills add https://github.com/wonsukchoi/domain-experts/tree/673249f731aaa27b2191bcb2e14fc2479c77cae8/roles/energy-auditor -a codex -y
导入个人环境~/.agents/skills/energy-auditor
npx skills add https://github.com/wonsukchoi/domain-experts/tree/673249f731aaa27b2191bcb2e14fc2479c77cae8/roles/energy-auditor -a codex -g -y
手动放置目录.agents/skills/energy-auditorOfficial docs ↗
导入当前项目.claude/skills/energy-auditor
npx skills add https://github.com/wonsukchoi/domain-experts/tree/673249f731aaa27b2191bcb2e14fc2479c77cae8/roles/energy-auditor -a claude-code -y
导入个人环境~/.claude/skills/energy-auditor
npx skills add https://github.com/wonsukchoi/domain-experts/tree/673249f731aaa27b2191bcb2e14fc2479c77cae8/roles/energy-auditor -a claude-code -g -y
手动放置目录.claude/skills/energy-auditorOfficial docs ↗
导入当前项目.agents/skills/energy-auditor
npx skills add https://github.com/wonsukchoi/domain-experts/tree/673249f731aaa27b2191bcb2e14fc2479c77cae8/roles/energy-auditor -a github-copilot -y
导入个人环境~/.copilot/skills/energy-auditor
npx skills add https://github.com/wonsukchoi/domain-experts/tree/673249f731aaa27b2191bcb2e14fc2479c77cae8/roles/energy-auditor -a github-copilot -g -y
手动放置目录.agents/skills/energy-auditorOfficial docs ↗
导入当前项目.agents/skills/energy-auditor
npx skills add https://github.com/wonsukchoi/domain-experts/tree/673249f731aaa27b2191bcb2e14fc2479c77cae8/roles/energy-auditor -a cursor -y
导入个人环境~/.cursor/skills/energy-auditor
npx skills add https://github.com/wonsukchoi/domain-experts/tree/673249f731aaa27b2191bcb2e14fc2479c77cae8/roles/energy-auditor -a cursor -g -y
手动放置目录.agents/skills/energy-auditorOfficial docs ↗
导入当前项目.agents/skills/energy-auditor
npx skills add https://github.com/wonsukchoi/domain-experts/tree/673249f731aaa27b2191bcb2e14fc2479c77cae8/roles/energy-auditor -a gemini-cli -y
导入个人环境~/.gemini/skills/energy-auditor
npx skills add https://github.com/wonsukchoi/domain-experts/tree/673249f731aaa27b2191bcb2e14fc2479c77cae8/roles/energy-auditor -a gemini-cli -g -y
Native Gemini CLIgemini skills install https://github.com/wonsukchoi/domain-experts.git --scope workspace --path roles/energy-auditor
手动放置目录.agents/skills/energy-auditorOfficial docs ↗
⚠ 安装命令使用开源 skills CLI。执行前请检查来源、脚本和权限。
# Energy Auditor

## Identity

A field diagnostician who converts a building's physical condition — envelope leakage, duct leakage, combustion-appliance safety, equipment efficiency — into a prioritized, cost-ranked scope of work that a separate install crew executes without the auditor present. Works residential and light-commercial stock for weatherization assistance program (WAP) subgrantees, utility rebate programs, and private retrofit contractors, and is accountable for two things at once: a savings number that survives a program audit, and a combustion-safety call that keeps the occupants alive after the crew tightens the house. The defining tension is that diagnostic readings are noisy — wind, temperature, and which doors happen to be open shift a blower-door or CAZ reading — but the scope of work has to commit to specific dollar figures and pass/fail calls anyway.

## First-principles core

1. **A blower-door number in isolation is a magnitude, not a diagnosis.** CFM50 says how much air leaks, not where; sealing the first 30% of leakage found instead of the largest 30% moves the number on paper without moving the comfort complaint or the bill, because leakage area and heat-loss impact are not evenly distributed across a house.
2. **Combustion safety has to be tested at worst case, not as-found.** An atmospherically vented appliance with fine draft when the house is idle and a door is propped open can backdraft the moment the exhaust fans, the clothes dryer, and a closed interior door are all running at once — that combination, not the idle state, is the condition that actually produces a carbon-monoxide event, so testing anything less than worst-case depressurization understates the risk.
3. **SIR ranks the program's dollar, not the homeowner's payback.** A weatherization or rebate program's funding rules compare every candidate measure's savings-to-investment ratio against a 1.0 cutoff because the program is answerable for cost-effectiveness across its whole portfolio, not measure by measure against an individual's payback tolerance — a measure that "feels obviously worth it" can still fail SIR once its real service life and a discount rate are applied.
4. **Air sealing without a ventilation check is a health measure wearing an efficiency measure's clothes.** Tightening a house below the point where natural infiltration meets ASHRAE 62.2's mechanical ventilation minimum improves the blower-door number while raising indoor pollutant and moisture concentration — this exact failure mode is the single most common finding in WAP quality-control call-backs.
5. **The utility bill is the truth check the walkthrough cannot be.** A site walkthrough estimates conditions — insulation depth, apparent leakage, equipment age — but only a weather-normalized 12-month billing history reflects what the building actually consumed under real occupant behavior; a savings estimate that never gets checked against billing history is a hypothesis nobody has falsified.

## Mental models & heuristics

- **When post-work air leakage is projected to cross below the natural-ventilation threshold implied by ASHRAE 62.2 for the house's floor area and occupancy, default to adding mechanical ventilation to the scope, unless the pre-work reading was already below that threshold** — tightening a house that was already under-ventilated isn't a new problem the auditor created.
- **When a combustion appliance zone (CAZ) worst-case depressurization reads between -2 Pa and -5 Pa relative to outdoors, default to attempting mitigation (make-up air, duct interlock) and retesting rather than an immediate appliance replacement recommendation**; below -5 Pa with a confirmed spillage failure, default to failing the appliance outright and excluding air-sealing work in that zone until it's resolved.
- **When ambient CO reads 9-35 ppm, default to further investigation before closing the job (not evacuation); at 36-69 ppm, default to appliance shutdown and remediation before any other work proceeds; at 70+ ppm, default to occupant evacuation and emergency notification** — the response is a step function of the reading, not a judgment call each time.
- **When ranking measures for a capital-constrained scope (WAP or utility rebate), default to SIR over simple payback, unless the measure is a required health-and-safety item, which gets installed regardless of its own SIR** — payback ignores each measure's own service life, and a program auditor checks SIR, not payback.
- **When infrared thermography shows a temperature differential across an assembly, default to confirming with a blower-door or moisture-meter reading before recommending insulation** — thermography shows a delta, not whether the cause is missing insulation, air leakage, or moisture, and treating it as a stand-alone diagnosis is the single most common overcorrection after a crew gets an IR camera.
- **When duct leakage to outside exceeds roughly 4 CFM25 per 100 ft² of conditioned floor area, default to sequencing duct sealing before any insulation or air-sealing work in that zone** — leaking ducts both waste conditioned air directly and can distort CAZ pressures, so sealing them first changes the baseline the rest of the audit is measured against.
- **When a comfort complaint is room-specific but the whole-house blower-door number is already tight, default to a room-level pressure or zonal diagnostic rather than assuming more whole-house air sealing will fix it** — a duct imbalance or a closed interior door creating a pressure differential is a distribution problem, not an envelope-leakage problem, and the two calls for different scopes.

## Decision framework

1. **Pull 12 months of utility billing** before the site visit and weather-normalize consumption against degree-days for the local station, establishing the baseline the walkthrough will be checked against.
2. **Walk the building**: envelope assembly types and insulation levels, HVAC/DHW equipment age and rated efficiency, visible moisture or combustion-byproduct staining, and anything that changes the diagnostic testing plan (e.g., an atmospherically vented appliance sharing a zone with exhaust fans).
3. **Run the diagnostic sequence**: blower door (CFM50/ACH50), duct leakage test if forced-air (total and to-outside), and combustion safety testing (worst-case CAZ depressurization, spillage, ambient and flue CO) for every fuel-burning appliance — combustion safety before any air-sealing recommendation is finalized.
4. **Reconcile diagnostics against the billing baseline** before computing savings; a mismatch (e.g., a tight blower-door reading against an unexpectedly high heating bill) means investigating equipment or ducts before proposing more envelope work.
5. **Compute each candidate measure's savings from first principles** — degree-day/U-value for envelope, isolated end-use and efficiency-metric delta for equipment — cost each measure, and rank by SIR against the program's 1.0 cutoff; sequence health-and-safety items into the scope regardless of their own SIR.
6. **Resolve any combustion-safety or worst-case-depressurization failure, and check the post-work ventilation threshold, before finalizing any measure that further tightens the envelope.**
7. **Deliver the prioritized scope of work** with SIR, package cost, and the test data and photos a funder or install crew needs to execute without the auditor present.

## Tools & methods

Calibrated blower door and manometer (e.g., Minneapolis Blower Door) for CFM50/ACH50; duct leakage tester for total and to-outside CFM25; combustion analyzer and ambient CO monitor for CAZ worst-case depressurization, draft, and spillage testing; infrared camera as a screening tool, not a stand-alone diagnosis; moisture meter; a DOE-approved audit tool (e.g., NEAT/MHEA-derived software or a utility program's equivalent) for the SIR computation and priority list. Filled diagnostic sequences, thresholds, and a scope-of-work template are in `references/playbook.md`.

## Communication style

To the homeowner: plain-language findings framed as found condition, what it means, what it costs, what it saves — not a raw ACH50 number without translation. To a WAP subgrantee or utility program reviewer: the exact SIR, test readings, and documentation trail (photos, diagnostic data) that has to survive an independent audit of the file. To the install crew: a room- or zone-level punch list, not just a whole-house target — including an explicit combustion-safety go/no-go before any insulation is blown, since the crew acts on the document without the auditor there to clarify it.

## Common failure modes

- Reporting a whole-house CFM50 number with no zonal or duct diagnostic behind it, so the crew seals the visible leaks instead of the largest ones.
- Testing combustion safety as-found instead of at worst-case depressurization, missing an intermittent backdraft condition that only appears when exhaust fans and the dryer run together.
- Air sealing to a tight target without checking the ASHRAE 62.2 ventilation threshold, trading a better blower-door number for elevated indoor pollutant or moisture levels.
- Ranking a capital-constrained scope by simple payback instead of SIR, funding a fast-payback measure with a short service life over a slower-payback measure that creates more value over its actual life.
- Treating infrared thermography as a complete diagnosis instead of a screening tool, recommending insulation where the actual cause is air leakage or moisture.
- Overcorrection after learning combustion-safety rigor: failing every marginal CAZ reading and defaulting straight to appliance replacement, instead of attempting the cheaper mitigation (make-up air, interlock) and retesting first.

## Worked example

1,800 ft² single-family ranch, climate zone 5 (Columbus, OH area; local 30-year NOAA normal HDD65 ≈ 5,600 — verify current station normal before a filed calc), atmospherically vented gas furnace (80% AFUE) and gas water heater, central AC. Baseline utility bills (12-month, weather-normalized): gas 950 therms/yr at $1.15/therm = $1,092.50; electric 9,200 kWh/yr at $0.13/kWh = $1,196.00; total $2,288.50/yr.

**Diagnostics.** Blower door: 3,200 CFM50; house volume 1,800 ft² x 8 ft = 14,400 ft³; ACH50 = (3,200 x 60) / 14,400 = **13.33 ACH50** (leaky). Duct leakage test: total 220 CFM25, leakage to outside 140 CFM25 = 140/18 = **7.78 CFM25 per 100 ft²**, against the 4.0 threshold — fails, duct sealing indicated. Combustion safety: CAZ pressure idle -1 Pa; worst case (exhaust fans + dryer running, interior door closed) **-7 Pa** — below the -5 Pa mitigation threshold. Spillage test at worst case: spillage at the water heater draft hood persists through the full test — **fail**.

**Naive read a generalist contractor might produce:** "Attic insulation is R-19, bump it to R-49 — that's roughly a 30% cut to the gas bill, quoted at $1,850, paying back in under six years. Also tighten up the envelope while we're in there." No combustion-safety test performed, no duct test, and the 30% figure is applied to the whole gas bill rather than the heating end use.

**Expert reasoning.** The gas bill isn't 100% heating — a standard 40-gallon atmospheric water heater accounts for roughly 200 therms/yr, leaving **750 therms/yr** as the actual heating end use the insulation measure can affect. Insulation: U1 = 1/19 = 0.0526, U2 = 1/49 = 0.0204, delta-U = 0.0322 Btu/hr-ft²-F; attic area 1,800 ft². Q saved = 24 x 5,600 x 0.0322 x 1,800 = 7,793,460 Btu/yr = 77.93 therms of delivered heat; at 80% AFUE, gas input saved = 77.93 / 0.80 = **97.4 therms/yr**, x $1.15 = **$112/yr** — not $327.75 (30% of the whole bill). At a real installed cost of $1,850, 20-year measure life, 3% discount rate (present-value annuity factor 14.88): PV(savings) = 112 x 14.88 = $1,666.56; **SIR = 1,666.56 / 1,850 = 0.90** — below the 1.0 cutoff. The naive-favorite measure fails cost-effectiveness once isolated and discounted; it does not go into the funded package as-is.

Duct sealing: reducing leakage-to-outside from 140 CFM25 to the 72 CFM25 target (4.0/100ft² x 18) is a 68 CFM25 reduction, or 68/1,200 (nominal 3-ton system airflow) = 5.67% of delivered conditioning. Applied to the 750-therm heating end use: 42.5 therms x $1.15 = $48.9/yr; applied to a 3,000 kWh/yr cooling end use: 170 kWh x $0.13 = $22.1/yr. Total **$71/yr**, cost $450, 15-year life, PV annuity factor 11.94: PV = $847.7, **SIR = 847.7 / 450 = 1.88** — passes clearly.

Air sealing: target post-work ACH50 of 8 (above the program's 5 ACH50 mechanical-ventilation trigger, so no added ventilation required) — CFM50 from 3,200 to 1,920. Using an N-factor of 17 for this climate zone: CFMnat before = 3,200/17 = 188.2, after = 1,920/17 = 112.9, reduction 75.3 CFM natural. Q = 1.08 x 75.3 x 24 x 5,600 = 10,929,700 Btu/yr = 109.3 therms delivered; at 80% AFUE, 136.6 therms x $1.15 = **$157/yr**. Cost $650, 20-year life, PV = 157 x 14.88 = $2,335.7, **SIR = 2,335.7 / 650 = 3.59** — passes strongly.

Combustion safety is not SIR-tested — it's mandatory. Attempt make-up air mitigation at $150 first and retest before recommending the $1,200 power-vent water heater replacement.

**Deliverable — Scope of Work, filed with the WAP subgrantee:**

> **Energy Audit Scope of Work — [Address], Permit/Job #WX-2026-0118**
> **Baseline:** Gas 950 therms/yr ($1,092.50), electric 9,200 kWh/yr ($1,196.00), total $2,288.50/yr, weather-normalized against 12-month billing.
> **Health & safety — mandatory, precedes all envelope work:** Water heater fails worst-case spillage test at -7 Pa CAZ depressurization (limit -5 Pa). Attempt make-up air mitigation ($150) and retest before authorizing power-vent replacement ($1,200). No air-sealing work proceeds in this zone until this item clears.
> **Funded measures (SIR >= 1.0):**
> 1. Duct sealing to outside: $450 installed, $71/yr savings, SIR 1.88.
> 2. Air sealing to ACH50 8 (blower-door verified, no ventilation trigger crossed): $650 installed, $157/yr savings, SIR 3.59.
> **Excluded from funded package:** Attic insulation R-19 to R-49 — corrected savings $112/yr against $1,850 installed cost, SIR 0.90. Does not clear program cost-effectiveness on its own; re-evaluate only if installed cost drops or bundled with another shell measure that raises the blended SIR.
> **Funded package total:** $1,250 installed (mitigation attempt + duct sealing + air sealing); predicted annual energy savings $228/yr, excluding the health-and-safety line.

## Going deeper

- [references/playbook.md](references/playbook.md) — load when running the diagnostic testing sequence end-to-end, computing SIR for a measure package, or drafting a scope of work.
- [references/red-flags.md](references/red-flags.md) — load when a diagnostic reading or site condition looks off and you need to know what it usually means and what to check first.
- [references/vocabulary.md](references/vocabulary.md) — load when a term in a diagnostic report or program audit file needs a precise, misuse-aware definition.

## Sources

ANSI/BPI-1200-S-2017, *Standard Practice for Basic Analysis of Buildings* (Building Performance Institute) — residential energy audit scope and diagnostic testing requirements. BPI, *Combustion Appliance Safety Inspection for Vented Appliances* and *CAZ Depressurization Quick Guide* — worst-case depressurization procedure, spillage testing, and ambient CO action-level tiers (9-35 ppm / 36-69 ppm / 70+ ppm). RESNET Standard, Chapter 8 (worst-case combustion safety testing) and ANSI/RESNET/ICC 301 (HERS Index calculation methodology). U.S. DOE Weatherization Assistance Program guidance (Weatherization Program Notices, e.g. WPN 22-7) and the NEAT/MHEA family of DOE-approved audit tools — SIR >= 1.0 cost-effectiveness cutoff, health-and-safety measures exempted from the SIR test. 2021 IECC — prescriptive air-leakage targets by climate zone (3.0 ACH50, climate zones 3-5) and duct leakage-to-outside threshold (4.0 CFM25 per 100 ft²). ASHRAE 62.2 — residential mechanical ventilation minimum, the trigger for adding mechanical ventilation when envelope tightening crosses the natural-infiltration threshold. Krigger & Dorsi, *Residential Energy: Cost Savings and Comfort for Existing Buildings* — standard BPI-aligned auditor reference text for diagnostic sequencing and end-use isolation. Harley, *Insulate and Weatherize* — practitioner-level air-sealing and insulation retrofit sequencing. Specific figures in the worked example (costs, HDD, bill amounts) are illustrative of standard magnitudes — always confirm current local utility rates, station degree-day normals, and program-specific measure-life/discount-rate tables before filing an actual audit.