Phase 1 B2 Was Outdoor-Only
Phase 1 Inv 03 evaluated aggressive building electrification (B2, $2B) on outdoor PM2.5 only. It found just 47 deaths avoided (Di CRF), giving a $42.6M cost-per-death that killed the scenario on cost-effectiveness grounds.
But buildings are where Californians live. Gas cooking alone adds ~14 mg PM2.5 per event and 8 ppb NO₂ average daily — effects that evaporate after electrification but were invisible to the outdoor-only health model. Inv 19 couples indoor and outdoor air on a four-level ladder.
The question: once the indoor pathway is modelled, does B2 flip from “not worth $2B” to cost-competitive with Transport T2? Yes.
L1 Outdoor-Only → L4 Personal Exposure
Five California building archetypes from the CEC Residential Appliance Saturation Study (pre-1978 SF, 1978–2005 SF, 2006+ Title 24 SF, pre-2005 multi-family, 2005+ multi-family) span the state’s 14M households. Each archetype carries its own air-exchange rate, penetration factor, and gas-appliance share.
Population-weighted indoor PM2.5 at 9 µg/m³ outdoor baseline. Emission factors: Singer et al. 2017 (gas cooking), Zhu et al. 2020 (NO₂), Urbanski 2014 (outdoor). Time-activity: NHAPS / CHAD.
What Changes When Gas Goes Away
The electrification delta is pre-minus-post indoor PM2.5 at each fidelity level. L1 registers zero because outdoor air is unchanged. L2–L4 add the indoor pathway that Phase 1 omitted.
| Level | Δ PM2.5 (µg/m³) | Deaths (Di) | Deaths (Krewski) | $/Death (Di) |
|---|---|---|---|---|
| L1 outdoor only | +0.00 | 0 | 0 | — |
| L2 single-zone | +1.51 | 338 | 1135 | $5.9M |
| L3 multi-zone | +1.93 | 432 | 1451 | $4.6M |
| L4 personal exposure | +1.32 | 294 | 987 | $6.8M |
Deaths computed over the 23,747,400 Californians in gas-cooking households. Di CRF 9.4 deaths/µg/1M, Krewski 31.6. $/Death = $2B B2 cost divided by level-specific deaths avoided. Phase 1 Inv 03 modelled only the L1 row.
Building Electrification Is Worth $2B
The indoor pathway is additive to the outdoor pathway. Phase 1 found 47 outdoor deaths (Di) for B2. Adding the L4 indoor pathway contributes an additional 294 deaths, bringing the total to 341 deaths. That is $5.9M per death — above Transport T2’s $3.2M but in the same cost-effectiveness band and well below the VSL of $11.6M.
Implication for the portfolio. The Phase 1 combined-portfolio finding (Inv 04) dropped B2 because it couldn’t beat T2 on dollars. With L4 indoor coupling, B2 re-enters the frontier and the combined T2+B2 portfolio is the optimal spend — not T2 alone.
CONTAM-Lite Mass Balance
L3 solves a steady-state 3-zone mass balance per archetype:
A · c = b where A encodes air exchange, deposition,
inter-zone stack flow and exhaust-hood removal; b carries the
outdoor penetration and kitchen cooking source. L4 pulls the L3 field into a
time-activity weighted exposure using the National Human Activity Pattern Survey
(68% home / 16% work / 8% outdoor / 8% transit).
The ladder is deliberately conservative — we under-weight transient cooking spikes that would push the incremental benefit higher. Even with the conservative choice, the verdict flips.
Sources: CEC Residential Appliance Saturation Study 2021 (archetype stock); Singer et al. 2017, Dobbin et al. 2018 (gas cooking PM2.5 emission factors); Zhu et al. 2020 (NO₂ gas-stove epidemiology); Persily & Musser 2017 (CONTAM validation); NHAPS (time-activity); Di et al. 2017 (CRF).