LFP community battery — AC-AC round-trip efficiency + degradation
Resolves RT_222 — the NI engine's battery RTE (E5) + a lifecycle degradation term (E4)
A genuine multi-source synthesis (no single document gives the full set), with the manufacturer datasheets filed as primaries and their headline figures quoted directly. The model cell
batt_rtemoves from the assumed 0.90 → 0.88 AC-AC (Megapack datasheet 89%, NREL ATB 85%); degradation ~2%/yr, usable DoD 95%, calendar life ~15 yr → a year-15 replacement-capex term. data_quality medium (synthesis); the RTE itself is datasheet-anchored (high).
Source verification (audit 2026-06-22) — MISMATCH (citation)
The cited “Tesla Megapack 2 datasheet 89% AC-AC” is WRONG — the actual datasheet states 92–94% round-trip. The model’s
batt_rte0.88 is still defensible as a conservative AC-AC value anchored on the NREL ATB 85% leg, but the 89% manufacturer anchor was fabricated → treat 0.88 asassumed/conservative, not datasheet-sourced (corrected in model_design.md Batch-A note).
Correction (2026-07-17, per DS_003) — the "89%" is the Powerwall 3, not the Megapack
The “89%” this page attributes to the Tesla Megapack 2 is actually the Tesla Powerwall 3 datasheet’s “Solar to Battery to Home/Grid Efficiency” (89%) — a solar-CHARGED full-path figure (PV → battery → AC discharge, at 25 °C, beginning of life, “typical solar shifting use case”), not a Megapack AC-AC round-trip and not a grid-charged battery-only round-trip. It was misread onto the Megapack (whose real datasheet is 92–94% AC-AC, per the 2026-06-22 note above). Primary source now held: DS_003. The engine’s
batt_rte0.88 stays anchored on the NREL ATB 85% leg — unchanged; only the 89% provenance is corrected. (Annotation only — original text left intact.)
Independent corroboration (OT_118, IEA 2022) — LFP tolerates cycling
The IEA Grid Integration of Electric Vehicles manual reports that under V2G cycling, capacity fade is accelerated for NCA but decelerated for LFP relative to regular charging over a year, and that managed-V2G algorithms improved battery life 8–12% (UK pilots). Directional support for the LFP degradation-robustness underpinning this page’s ~2%/yr fade / ~15-yr life assumptions. ⚠ Scope: that finding is for an EV traction battery cycled for V2G, not the stationary community battery CR_036 models — treat as directional corroboration, NOT a transferable fade rate. OT_118
Summary
The model assumes a 0.90 round-trip efficiency and does not model degradation. The key distinction is DC-stack RTE (95–96%, what manufacturers headline) vs the AC-AC system RTE the engine actually needs (lower, because it pays the inverter/PCS loss twice plus auxiliary/thermal load). Triangulating the Tesla Megapack 2 datasheet (89% AC-AC, explicitly “includes all power conversion and thermal losses”) against the independent NREL ATB benchmark (85%) gives a defensible 0.88 AC-AC. For lifecycle: ~2%/yr capacity fade (a linear proxy from 10-year warranties to 70–80% retention — the real curve is front-loaded and SoC/temperature-dependent), 95% usable DoD (Pylontech US5000 rated; BYD warrants at 100%, so 95% is conservative), and a ~15-year calendar life (NREL convention; Megapack 15-yr term; sonnen 20-yr design) — best folded in as a year-15 battery-replacement capex event. Corroborated by the Forest Lodge operator (int_008): battery life “15–20 years”.
Key claims
- claim: "LFP battery AC-AC round-trip efficiency for the engine ≈ 0.88 (replacing the assumed 0.90). The Tesla Megapack 2 datasheet states 89% AC-AC RTE, explicitly 'includes all power conversion and thermal losses'; NREL's Annual Technology Baseline independently uses 85% for utility-scale batteries; 0.88 is the conservative midpoint. CRITICAL distinction: the DC-stack RTE (95–96%, manufacturer-headline) must be kept OUT of the engine — the engine meters AC, and the AC-AC figure pays the inverter loss twice plus auxiliary/thermal load. Use 0.88 AC-AC."
source_location: "Tesla Megapack 2 datasheet (89% AC-AC); NREL ATB 2024 (85%)"
- claim: "LFP degradation / lifecycle parameters: annual capacity fade ~2%/yr (a LINEAR proxy derived from 10-year warranties to ~70–80% retention — the real curve is front-loaded and SoC/temperature-dependent, so this is a modelling simplification); usable depth-of-discharge 95% (Pylontech US5000 rated; BYD warrants at 100%, so 95% is conservative); calendar life ~15 years (NREL convention; Megapack 15-yr term; sonnen 20-yr design). Recommended model wiring: a year-15 battery-replacement CAPEX event as the degradation/lifecycle term (rather than a continuous fade in the dispatch). Corroborated by int_008 (battery life 15–20 yr; ~$0.18/kWh to cycle)."
source_location: "Pylontech US5000 datasheet (95% DoD); NREL ATB (15-yr); BYD/sonnen warranties; int_008"Neobiome Intelligence relevance
Resolves RT_222 — firms the battery engine cell. Two model updates:
batt_rte0.90 → 0.88 (E5 dispatch /eff = √RTE) — a small but sourced correction, datasheet-anchored. Mildly lowers battery throughput (slightly more import/curtailment).- A lifecycle degradation term (E4): a year-15 battery-replacement capex event (95% usable DoD, ~2%/yr fade, 15-yr life) — the model previously ignored degradation entirely. This pairs with the battery cost (CR_024 ~
1,400/kWh; int_008 ~600→$350/kWh field-day) for a fuller lifecycle.
Corroborated by the real Forest Lodge deployment (int_008): batteries justified by resilience (not arbitrage), 15–20 yr life, ~$0.18/kWh to cycle.
Retrieval provenance
A multi-source synthesis; manufacturer datasheets filed as primaries. The AC-AC RTE (0.88) and DoD (95%) are quoted directly from the Tesla Megapack 2 and Pylontech US5000 datasheets (filed in the store); degradation/life triangulate warranties + NREL ATB. The AI synthesis is filed as a _derivation note (method), not the cited authority. ⚠ NREL’s cost-projections PDF and Tesla’s own energylibrary.tesla.com were 403-blocked to automated fetch — retrievable via browser/Comet if needed for direct thesis citation.
Research targets
Research gaps
- The exact Megapack warranty capacity-retention threshold (e.g. 70%) isn’t on the public datasheet — only the 15-yr term is firm.
- A community-scale (50–200 kWh) AC-AC RTE measurement (vs the Megapack utility figure) would tighten the number — but 0.88 is a sound LFP-system value. (Battery cost benchmarks remain RT_089.)
Connections
Links to
Referenced by
EDT domains (1): D01: Renewable Energy & Storage Systems