Skip to content

v1.14.0 — COP for everyone, with or without a power meter

Choose a tag to compare

@elRadix elRadix released this 29 Oct 10:27
· 189 commits to main since this release
26b1c89

v1.14.0 — COP for everyone, with or without a power meter

Your heat pump's COP is now calculated even if you don't have a dedicated power meter. ESPAltherma 2.0 users get it automatically. ESPAltherma 1.x users get a built-in estimate. Power-meter users see no change.

📊 Featured dashboard — Daily Summary Tabbed

Overview
Overview
Quality
Quality
Cycles
Cycles
Events
Events

Tap any screenshot → opens the card's README with copy-paste YAML, requirements, and variables.

→ Card documentation & install guide


🌟 What's new in v1.14.0

The headline: COP now works without a power meter

The problem before this release. To show your heat pump's COP, the integration needed to know how much electricity the heat pump is using. Most users solve this by installing a smart power meter (like a Shelly or a P1 meter). But if you don't have one, your cop_today sensor was empty — no data at all.

Meanwhile, ESPAltherma 2.0 (the newer firmware by raomin) already calculates the electrical input internally and publishes it as a Stats: section. We just weren't reading it. That's fixed now.

What this means in practice:

  • You have a power meter → nothing changes. Your meter remains the source of truth.
  • You have ESPAltherma 2.0 (no power meter) → COP appears automatically. No configuration needed.
  • You have ESPAltherma 1.x (no power meter) → COP appears using an estimate. You can fine-tune the estimate in settings.
  • You have neither → no change. You still need one of the two to get COP data.

How the integration picks the right source

Three sources, checked in priority order. First available wins:

  1. Power meter (if you configured sensor.warmtepomp_power or similar) — most accurate, no adjustment applied.
  2. ESPAltherma 2.0 Stats (auto-detected) — read directly from the device, no adjustment applied. ESPAltherma 2.0 handles phases and backup-heater internally, so we don't add anything on top (that would double-count).
  3. ESPAltherma 1.x fallback — calculated from inverter current × 230 V × number of phases + backup-heater consumption + a 100 W offset for pump, fan and electronics that the inverter current doesn't capture.

📖 Full explanation: COP without a power meter →


New settings page for fine-tuning

If you're on ESPAltherma 1.x, you can tell the integration about your installation so the fallback estimate is accurate.

Where to find it: Settings → Devices & Services → Daikin Cycle ML → Configure → COP calculation

Setting Default When to change it
Supply phases 1 You have a three-phase electrical installation → set to 3
BUH step 1 power (kW) 3.0 Your backup heater's step 1 has a different rating
BUH step 2 power (kW) 3.0 Your backup heater's step 2 is 6 kW (common on bigger units)

You don't need to touch this if:

  • You have a power meter configured → these settings are ignored.
  • You have ESPAltherma 2.0 → these settings are ignored (v2 handles it internally).

You only need to touch this if: you're on ESPAltherma 1.x, you don't have a power meter, AND your defaults don't match your installation.

📖 Options reference: 4.10 COP calculation →

New sensor to see what's happening

sensor.daikin_cycle_ml_electrical_power_estimated shows the live electrical input estimate in watts. Its source attribute tells you exactly which path was used:

  • power_meter → your power sensor
  • espaltherma_v2_stats → ESPAltherma 2.0 Stats section
  • inverter_estimated_v1 → ESPAltherma 1.x fallback estimate
  • none → no source available right now

Useful for verifying the cascade is working in real time, especially after changing installation settings.

📖 Entity reference →


⚠️ What you'll see after upgrading

Nothing is rewritten. Historical COP samples keep their old values. Only new samples (taken from the moment of upgrade) use the improved logic. No data loss, no recalculation, no migration step.

Your setup What changes after upgrade
Power meter configured Nothing. COP values unchanged.
ESPAltherma 2.0, no power meter COP now appears where it was empty. Numbers may differ slightly from a manual template — ESPAltherma Stats includes backup-heater consumption, most custom templates don't.
ESPAltherma 1.x, no power meter COP now appears where it was empty. Starting defaults: single-phase, 3 kW + 3 kW backup heater. Adjust in COP calculation if your installation differs.
No power meter, no ESPAltherma Stats Nothing. You still need one of them to get COP.

📖 Upgrade guide: v1.14.0 →


🐛 Full fix list

New features — #206

  • Internal electrical-input cascade — the priority logic described above (raomin Discussion #599 pattern)
  • New OptionsFlow step — "COP calculation" for phases + BUH kW per step
  • New sensor — electrical_power_estimated with full source transparency
  • New internal fallback — COP is now calculated by the integration itself when no external COP sensor is configured

Defensive fixes — #206

These were surfaced by the full test suite while validating the cascade. They protect against edge cases:

  • _errors_total class-level default — 12 partial test fixtures were crashing in the error-handler when the coordinator was instantiated without running __init__ first. Partial fixture setups only; no user impact.
  • _read_power / _read_power_w — now use getattr guards, mirroring the new _power_from_meter pattern. Prevents AttributeError when the power sensor attribute isn't set.
  • cycles_today() — skips cycles with start_ts more than 60 seconds in the future. Protects against clock skew and DB-import corruption.
  • _value_duty_cycle_today_pct — clamped to [0, 100]. Prevents the sensor from showing >100% at the midnight boundary.

Related issues

  • #192 — Consume ESPAltherma 2.0 pre-computed entities. Partially handled by this release: the Stats section is now consumed. Full entity consumption remains blocked on external dependencies.
  • #187 — ESPAltherma collaboration epic. Umbrella tracker for this release and future cross-project work with raomin.

Test coverage

  • Test suite grew from 3703 to 3729 (+26 new tests)
  • 4 new test files:
    • test_205_electrical_power_cascade.py — 14 tests for the cascade logic
    • test_206_cop_fallback.py — 6 tests for internal COP fallback
    • test_g1_defensive_guards.py — 6 tests for the defensive fixes above
    • test_config_flow_cop_calc.py — 2 tests for the new settings page
  • 100.00% branch coverage maintained across the whole codebase
  • mypy strict clean across 38 source files
  • ruff CI-parity clean

📖 Full technical changelog →


🙏 Credits


📚 Documentation