13215d7ddb
Issue #46 (Phase A only — Phase B human rating still pending, issue stays open). Adds the data-collection half of the calibration milestone: - scripts/calibration_runner.sh — runs 20 fixed balanced-depth queries across 4 categories (factual, comparative, contradiction-prone, scope-edge), 5 each, capturing per-run logs to docs/stress-tests/M3.3-runs/. - scripts/calibration_collect.py — loads every persisted ResearchResult under ~/.marchwarden/traces/*.result.json and emits a markdown rating worksheet with one row per run. Recovers question text from each trace's start event and category from the run-log filename. - docs/stress-tests/M3.3-rating-worksheet.md — 22 runs (20 calibration + caffeine smoke + M3.2 multi-axis), with empty actual_rating columns for the human-in-the-loop scoring step. - docs/stress-tests/M3.3-runs/*.log — runtime logs from the calibration runner, kept as provenance. Gitignore updated with an exception carving stress-test logs out of the global *.log ignore. Note: M3.1's 4 runs predate #54 (full result persistence) and so are unrecoverable to the worksheet — only post-#54 runs have a result.json sibling. 22 rateable runs is still within the milestone target of 20–30. Phases B (human rating) and C (analysis + rubric + wiki update) follow in a later session. This issue stays open until both are done. Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
226 lines
27 KiB
Text
226 lines
27 KiB
Text
Researching: Compare the energy density of lithium-ion vs sodium-ion batteries.
|
||
|
||
{"question": "Compare the energy density of lithium-ion vs sodium-ion batteries.", "depth": "balanced", "max_iterations": null, "token_budget": null, "event": "ask_started", "logger": "marchwarden.cli", "level": "info", "timestamp": "2026-04-09T01:54:02.430608Z"}
|
||
{"transport": "stdio", "server": "marchwarden-web-researcher", "event": "mcp_server_starting", "logger": "marchwarden.mcp", "level": "info", "timestamp": "2026-04-09T01:54:03.159945Z"}
|
||
{"event": "Processing request of type CallToolRequest", "logger": "mcp.server.lowlevel.server", "level": "info", "timestamp": "2026-04-09T01:54:03.167971Z"}
|
||
{"question": "Compare the energy density of lithium-ion vs sodium-ion batteries.", "depth": "balanced", "max_iterations": 5, "token_budget": 20000, "model_id": "claude-sonnet-4-6", "event": "research_started", "trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "researcher": "web", "logger": "marchwarden.researcher.web", "level": "info", "timestamp": "2026-04-09T01:54:03.200030Z"}
|
||
{"step": 1, "decision": "Beginning research: depth=balanced", "question": "Compare the energy density of lithium-ion vs sodium-ion batteries.", "context": "", "max_iterations": 5, "token_budget": 20000, "event": "start", "trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "researcher": "web", "logger": "marchwarden.researcher.trace", "level": "info", "timestamp": "2026-04-09T01:54:03.200318Z"}
|
||
{"step": 2, "decision": "Starting iteration 1/5", "tokens_so_far": 0, "event": "iteration_start", "trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "researcher": "web", "logger": "marchwarden.researcher.trace", "level": "info", "timestamp": "2026-04-09T01:54:03.200405Z"}
|
||
{"step": 7, "decision": "Starting iteration 2/5", "tokens_so_far": 1114, "event": "iteration_start", "trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "researcher": "web", "logger": "marchwarden.researcher.trace", "level": "info", "timestamp": "2026-04-09T01:54:14.560598Z"}
|
||
{"step": 12, "decision": "Starting iteration 3/5", "tokens_so_far": 7183, "event": "iteration_start", "trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "researcher": "web", "logger": "marchwarden.researcher.trace", "level": "info", "timestamp": "2026-04-09T01:54:18.314755Z"}
|
||
{"step": 19, "decision": "Starting iteration 4/5", "tokens_so_far": 13977, "event": "iteration_start", "trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "researcher": "web", "logger": "marchwarden.researcher.trace", "level": "info", "timestamp": "2026-04-09T01:54:28.528912Z"}
|
||
{"step": 24, "decision": "Token budget reached before iteration 5: 28015/20000", "event": "budget_exhausted", "trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "researcher": "web", "logger": "marchwarden.researcher.trace", "level": "info", "timestamp": "2026-04-09T01:54:39.027627Z"}
|
||
{"step": 25, "decision": "Beginning synthesis of gathered evidence", "evidence_count": 24, "iterations_run": 4, "tokens_used": 28015, "event": "synthesis_start", "trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "researcher": "web", "logger": "marchwarden.researcher.trace", "level": "info", "timestamp": "2026-04-09T01:54:39.028531Z"}
|
||
{"step": 26, "decision": "Parsed synthesis JSON successfully", "duration_ms": 50955, "event": "synthesis_complete", "trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "researcher": "web", "logger": "marchwarden.researcher.trace", "level": "info", "timestamp": "2026-04-09T01:55:27.614289Z"}
|
||
{"step": 41, "decision": "Research complete", "confidence": 0.91, "citation_count": 8, "gap_count": 3, "discovery_count": 3, "total_duration_sec": 87.865, "event": "complete", "trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "researcher": "web", "logger": "marchwarden.researcher.trace", "level": "info", "timestamp": "2026-04-09T01:55:27.616834Z"}
|
||
{"confidence": 0.91, "citations": 8, "gaps": 3, "discovery_events": 3, "tokens_used": 48087, "iterations_run": 4, "wall_time_sec": 84.41376757621765, "budget_exhausted": true, "event": "research_completed", "trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "researcher": "web", "logger": "marchwarden.researcher.web", "level": "info", "timestamp": "2026-04-09T01:55:27.617014Z"}
|
||
{"error": "[Errno 13] Permission denied: '/home/micro/.marchwarden/costs.jsonl'", "event": "cost_ledger_write_failed", "trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "researcher": "web", "logger": "marchwarden.researcher.web", "level": "warning", "timestamp": "2026-04-09T01:55:27.617866Z"}
|
||
{"event": "Processing request of type ListToolsRequest", "logger": "mcp.server.lowlevel.server", "level": "info", "timestamp": "2026-04-09T01:55:27.632124Z"}
|
||
{"trace_id": "aaf3b9ef-d91a-4d03-8883-b0a906929cb1", "confidence": 0.91, "citations": 8, "tokens_used": 48087, "wall_time_sec": 84.41376757621765, "event": "ask_completed", "logger": "marchwarden.cli", "level": "info", "timestamp": "2026-04-09T01:55:27.873634Z"}
|
||
╭─────────────────────────────────── Answer ───────────────────────────────────╮
|
||
│ Lithium-ion batteries have significantly higher energy density than │
|
||
│ sodium-ion batteries across all commercial chemistries. Lithium-ion cells │
|
||
│ achieve 150–300 Wh/kg gravimetrically, depending on chemistry: NMC variants │
|
||
│ reach 250–300 Wh/kg in premium automotive applications, while LFP cells │
|
||
│ deliver 150–180 Wh/kg [Source 15]. Volumetrically, lithium-ion batteries │
|
||
│ reach roughly 250–700 Wh/L [Source 16]. Sodium-ion batteries currently │
|
||
│ achieve 90–190 Wh/kg gravimetrically; CATL's first-generation commercial │
|
||
│ cells reached ~160 Wh/kg [Source 15], with newer products like CATL's Naxtra │
|
||
│ reaching ~175 Wh/kg [Source 22], and ScienceDirect prototypes ranging 90–150 │
|
||
│ Wh/kg [Source 7]. The volumetric energy density of sodium-ion is │
|
||
│ approximately 20–40% lower than lithium-ion equivalents [Source 8]. This gap │
|
||
│ exists fundamentally because sodium ions are heavier and larger than lithium │
|
||
│ ions, reducing the energy stored per unit mass or volume [Source 3, Source │
|
||
│ 20]. A notable exception is a late-2025 announcement by ZN Energy of an │
|
||
│ anode-free solid-state sodium-ion pouch cell achieving 348.5 Wh/kg, verified │
|
||
│ by CATARC, using a high-energy layered oxide cathode and anode-free │
|
||
│ solid-state architecture—though this is a laboratory/prototype result, not │
|
||
│ yet commercial [Source 10]. In practical terms, sodium-ion batteries are │
|
||
│ best suited for stationary storage and cost-sensitive low-performance EVs │
|
||
│ where energy density is less critical, while lithium-ion dominates portable │
|
||
│ electronics, robotics, and long-range EVs [Source 1, Source 8]. │
|
||
╰──────────────────────────────────────────────────────────────────────────────╯
|
||
Citations
|
||
┏━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━┓
|
||
┃ # ┃ Title / Locator ┃ Excerpt ┃ Conf ┃
|
||
┡━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━┩
|
||
│ 1 │ Battery Energy Density 2025: │ Nickel Manganese Cobalt (NMC) │ 0.95 │
|
||
│ │ State of the Art & Next-Gen │ variants deliver the highest │ │
|
||
│ │ Tech │ energy densities at the cell │ │
|
||
│ │ https://timharper.net/fieldno │ level, reaching 250-300 Wh/kg │ │
|
||
│ │ tes/battery-energy-density-20 │ in premium automotive │ │
|
||
│ │ 25/ │ applications... Sodium-ion │ │
|
||
│ │ │ batteries have emerged from │ │
|
||
│ │ │ laboratory curiosity to │ │
|
||
│ │ │ commercial reality, with │ │
|
||
│ │ │ CATL's first-generation cells │ │
|
||
│ │ │ achieving 160 Wh/kg energy │ │
|
||
│ │ │ density. │ │
|
||
├─────┼───────────────────────────────┼────────────────────────────────┼───────┤
|
||
│ 2 │ Sodium ion batteries: A │ Current prototypes of SIBs │ 0.95 │
|
||
│ │ sustainable alternative to │ have energy densities of │ │
|
||
│ │ lithium-ion ... │ 90–150 Wh/kg, which remain │ │
|
||
│ │ https://www.sciencedirect.com │ lower than the 130–285 Wh/kg │ │
|
||
│ │ /science/article/pii/S2949821 │ typically achieved │ │
|
||
│ │ X25002418 │ │ │
|
||
├─────┼───────────────────────────────┼────────────────────────────────┼───────┤
|
||
│ 3 │ Sodium-ion batteries: Should │ Sodium is heavier than │ 0.97 │
|
||
│ │ we believe the hype? │ lithium, and its ions are │ │
|
||
│ │ https://cen.acs.org/energy/en │ larger, resulting in a │ │
|
||
│ │ ergy-storage-/Sodium-ion-batt │ volumetric energy density that │ │
|
||
│ │ eries-Should-believe/103/web/ │ is 20–40% less than that of │ │
|
||
│ │ 2025/11 │ lithium ion. Consequently, a │ │
|
||
│ │ │ sodium-ion battery is bigger │ │
|
||
│ │ │ and heavier than an equivalent │ │
|
||
│ │ │ one made with lithium. │ │
|
||
├─────┼───────────────────────────────┼────────────────────────────────┼───────┤
|
||
│ 4 │ Energy Density of Lithium-Ion │ Modern lithium-ion batteries │ 0.90 │
|
||
│ │ Batteries Explained: Wh/kg vs │ achieve 150-300 Wh/kg and │ │
|
||
│ │ Wh/L │ 250-700 Wh/L, depending on │ │
|
||
│ │ https://www.longsingtech.com/ │ chemistry and design. │ │
|
||
│ │ energy-density-of-lithium-ion │ │ │
|
||
│ │ -batteries/ │ │ │
|
||
├─────┼───────────────────────────────┼────────────────────────────────┼───────┤
|
||
│ 5 │ Sodium Ion vs Lithium Ion │ Energy Density (Gravimetric): │ 0.88 │
|
||
│ │ Batteries: 2026 Comparison & │ Sodium-ion typically ranges │ │
|
||
│ │ Key Advantages │ from 100–175 Wh/kg (e.g., │ │
|
||
│ │ https://chargeprotexas.com/so │ CATL's Naxtra at ~175 Wh/kg). │ │
|
||
│ │ dium-ion-vs-lithium-ion-batte │ Lithium-ion hits 150–250+ │ │
|
||
│ │ ries-2026-comparison/ │ Wh/kg (LFP: 150–210; NMC: │ │
|
||
│ │ │ 240–350). │ │
|
||
├─────┼───────────────────────────────┼────────────────────────────────┼───────┤
|
||
│ 6 │ ZN Energy Breaks Sodium-Ion │ Its >25Ah large-format AFSSSIB │ 0.78 │
|
||
│ │ Battery Density Record at │ pouch cell achieved a │ │
|
||
│ │ 348.5Wh/kg │ gravimetric energy density of │ │
|
||
│ │ https://www.linkedin.com/post │ 348.5Wh/kg, verified by CATARC │ │
|
||
│ │ s/jerry-wan-069b41105_breakin │ (China Automotive Technology & │ │
|
||
│ │ g-the-sodium-ceiling-zhaona-e │ Research Center, Tianjin). │ │
|
||
│ │ nergy-activity-74134108276403 │ This is not an incremental │ │
|
||
│ │ 20000-NHd_ │ improvement—it directly │ │
|
||
│ │ │ challenges the long-held │ │
|
||
│ │ │ assumption that sodium │ │
|
||
│ │ │ chemistry is structurally │ │
|
||
│ │ │ capped at 'low energy │ │
|
||
│ │ │ density.' │ │
|
||
├─────┼───────────────────────────────┼────────────────────────────────┼───────┤
|
||
│ 7 │ Sodium as a Green Substitute │ But there are also downsides │ 0.93 │
|
||
│ │ for Lithium in Batteries │ to sodium-ion batteries, the │ │
|
||
│ │ https://physics.aps.org/artic │ top one being a lower energy │ │
|
||
│ │ les/v17/73 │ density than their lithium-ion │ │
|
||
│ │ │ counterparts. Energy density │ │
|
||
│ │ │ has a direct bearing on the │ │
|
||
│ │ │ driving range of an electric │ │
|
||
│ │ │ vehicle. │ │
|
||
├─────┼───────────────────────────────┼────────────────────────────────┼───────┤
|
||
│ 8 │ Sodium-Ion vs Lithium-Ion │ lithium-ion batteries dominate │ 0.85 │
|
||
│ │ Batteries Differences and │ high-performance applications │ │
|
||
│ │ Applications in 2025 │ like consumer electronics and │ │
|
||
│ │ https://www.large-battery.com │ robotics, owing to their │ │
|
||
│ │ /blog/na-ion-vs-li-ion-batter │ superior energy density of │ │
|
||
│ │ ies-2025/ │ 100–270 Wh/kg. │ │
|
||
└─────┴───────────────────────────────┴────────────────────────────────┴───────┘
|
||
Gaps
|
||
┏━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
|
||
┃ Category ┃ Topic ┃ Detail ┃
|
||
┡━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
|
||
│ source_not_found │ Volumetric energy │ Most sources provide │
|
||
│ │ density figures for │ gravimetric (Wh/kg) data │
|
||
│ │ sodium-ion batteries │ for sodium-ion; specific │
|
||
│ │ │ Wh/L volumetric figures │
|
||
│ │ │ for sodium-ion cells at │
|
||
│ │ │ the commercial pack level │
|
||
│ │ │ were not found in │
|
||
│ │ │ evidence. │
|
||
├───────────────────────┼──────────────────────────┼───────────────────────────┤
|
||
│ contradictory_sources │ Independent verification │ The 348.5 Wh/kg result │
|
||
│ │ of ZN Energy 348.5 Wh/kg │ for sodium-ion is from a │
|
||
│ │ claim │ LinkedIn post summarizing │
|
||
│ │ │ a company announcement. │
|
||
│ │ │ No peer-reviewed or │
|
||
│ │ │ independent third-party │
|
||
│ │ │ publication was found to │
|
||
│ │ │ corroborate this figure. │
|
||
├───────────────────────┼──────────────────────────┼───────────────────────────┤
|
||
│ scope_exceeded │ Cycle life vs energy │ While cycle life is │
|
||
│ │ density trade-offs in │ mentioned in some │
|
||
│ │ sodium-ion │ sources, a detailed │
|
||
│ │ │ quantitative comparison │
|
||
│ │ │ of how energy density │
|
||
│ │ │ degrades over cycle life │
|
||
│ │ │ compared to lithium-ion │
|
||
│ │ │ was not covered in the │
|
||
│ │ │ evidence. │
|
||
└───────────────────────┴──────────────────────────┴───────────────────────────┘
|
||
Discovery Events
|
||
┏━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━┓
|
||
┃ ┃ Suggested ┃ ┃ ┃
|
||
┃ Type ┃ Researcher ┃ Query ┃ Reason ┃
|
||
┡━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━┩
|
||
│ new_source │ arxiv │ anode-free │ ZN Energy's 348.5 │
|
||
│ │ │ solid-state │ Wh/kg claim would │
|
||
│ │ │ sodium-ion │ benefit from │
|
||
│ │ │ battery energy │ peer-reviewed │
|
||
│ │ │ density 2025 │ validation on │
|
||
│ │ │ │ arXiv or similar │
|
||
│ │ │ │ preprint server. │
|
||
├──────────────────┼───────────────────┼───────────────────┼───────────────────┤
|
||
│ related_research │ database │ sodium-ion │ Volumetric energy │
|
||
│ │ │ battery │ density for │
|
||
│ │ │ volumetric energy │ sodium-ion at the │
|
||
│ │ │ density Wh/L │ cell and pack │
|
||
│ │ │ commercial cells │ level is │
|
||
│ │ │ 2025 │ underrepresented │
|
||
│ │ │ │ in current │
|
||
│ │ │ │ evidence. │
|
||
├──────────────────┼───────────────────┼───────────────────┼───────────────────┤
|
||
│ related_research │ arxiv │ layered oxide │ Multiple sources │
|
||
│ │ │ cathode │ mention cathode │
|
||
│ │ │ sodium-ion │ engineering as │
|
||
│ │ │ specific capacity │ the key │
|
||
│ │ │ cycle stability │ bottleneck for │
|
||
│ │ │ 2025 │ sodium-ion energy │
|
||
│ │ │ │ density │
|
||
│ │ │ │ improvement. │
|
||
└──────────────────┴───────────────────┴───────────────────┴───────────────────┘
|
||
Open Questions
|
||
┏━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┓
|
||
┃ Priority ┃ Question ┃ Context ┃
|
||
┡━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┩
|
||
│ high │ Will sodium-ion batteries ever │ ZN Energy's prototype achieved │
|
||
│ │ match or exceed LFP lithium-ion │ 348.5 Wh/kg, but commercial │
|
||
│ │ in gravimetric energy density │ CATL sodium-ion cells are at │
|
||
│ │ at the commercial pack level? │ ~160–175 Wh/kg while LFP cells │
|
||
│ │ │ are 150–180 Wh/kg. The gap is │
|
||
│ │ │ closing in prototypes but not │
|
||
│ │ │ yet in commercial products. │
|
||
├──────────┼─────────────────────────────────┼─────────────────────────────────┤
|
||
│ medium │ How does energy density change │ Sources mention sodium-ion's │
|
||
│ │ over the cycle life of │ lower risk of thermal runaway │
|
||
│ │ sodium-ion vs lithium-ion │ and good low-temperature │
|
||
│ │ batteries under real-world │ performance, but long-term │
|
||
│ │ conditions? │ energy density retention data │
|
||
│ │ │ was not found. │
|
||
├──────────┼─────────────────────────────────┼─────────────────────────────────┤
|
||
│ medium │ What is the volumetric energy │ C&EN states volumetric density │
|
||
│ │ density (Wh/L) of current │ is 20–40% lower than │
|
||
│ │ commercial sodium-ion battery │ lithium-ion but provides no │
|
||
│ │ packs? │ absolute Wh/L figures for │
|
||
│ │ │ sodium-ion. │
|
||
└──────────┴─────────────────────────────────┴─────────────────────────────────┘
|
||
╭───────────────────────────────── Confidence ─────────────────────────────────╮
|
||
│ Overall: 0.91 │
|
||
│ Corroborating sources: 8 │
|
||
│ Source authority: high │
|
||
│ Contradiction detected: False │
|
||
│ Query specificity match: 0.97 │
|
||
│ Budget status: spent │
|
||
│ Recency: current │
|
||
╰──────────────────────────────────────────────────────────────────────────────╯
|
||
╭──────────────────────────────────── Cost ────────────────────────────────────╮
|
||
│ Tokens: 48087 │
|
||
│ Iterations: 4 │
|
||
│ Wall time: 84.41s │
|
||
│ Model: claude-sonnet-4-6 │
|
||
╰──────────────────────────────────────────────────────────────────────────────╯
|
||
|
||
trace_id: aaf3b9ef-d91a-4d03-8883-b0a906929cb1
|