Quick Summary (TL;DR)
For credible heat-pump electricity costs, choose a dedicated, correctly installed electricity meterAliExpress price on the circuits that form your intended measurement boundary. For operating diagnostics, retain the manufacturer’s native integration. For most serious Home Assistant installations, the best answer is both. An integration can expose flow temperature, compressor state, operating mode and sometimes a vendor-calculated energy figure, but it cannot guarantee that the number includes immersion heaters, circulation pumps, controls or standby loads. An independent meter measures electrical consumption at a physical circuit boundary; it cannot by itself explain why that electricity was used or how much useful heat was produced.
Start with the question you need answered. If the aim is an accurate monthly kWh and running-cost trend, a dedicated cumulative energy meterAliExpress price is usually the strongest measurement foundation. If you only want faults, temperatures and schedules, native monitoring may be enough. If you want measured COP (heat delivered ÷ electricity consumed), you also need trustworthy thermal-energy measurement using flow and temperature data or a suitable heat meter over exactly the same interval and system boundary.
Never assume a smart socket is suitable for metering a permanently connected heat pump. Air-to-water systems are frequently hardwired or have multiple protected supplies. Electricians and qualified heat-pump installers must select compliant measurement equipment, decide where it goes, provide necessary isolation/protection and preserve manufacturer warranties. Home Assistant must remain a monitoring/supervisory layer, not an alternative to certified compressor, frost or heater safety controls.
The difference in one table: native monitoring versus an independent electricity meter
| Question | Heat-pump native integration | Dedicated electricity submeter | Recommended approach |
|---|---|---|---|
| What is the heat pump doing? | Often temperatures, demand, compressor mode, DHW and alarms | Only the load shape and energy | Native telemetry |
| How many kWh did the measured supply use? | Possibly reported, often calculated or incomplete | Measured at its installed electrical boundary | Dedicated cumulative kWh meter |
| Can I see power right now? | Sometimes estimated, sampled or unavailable | Normally live W and recorded kWh (depending on model) | Submeter, preferably with accessible local data |
| Does it capture standby? | Not guaranteed; depends on reporting logic | Yes, if the standby circuits pass through the meter and within its range | Check wiring boundary and low-current performance |
| Does it capture backup/immersion heat? | Varies by manufacturer and reported entity | Only when the relevant circuits are included | Agree a written whole-system boundary |
| Can it report delivered heat? | Sometimes estimated or reported by manufacturer | No: electrical meter measures electric input only | Thermal heat meter or validated thermal data |
| Does it work without broadband? | Local bus: usually yes; vendor cloud: no | Often yes with local Modbus/Ethernet reporting | Prefer independent local metering for long-term records |
| Can it adjust heating settings? | Sometimes via documented controls | No | Keep manufacturer controls authoritative |
| Does it need extra installation? | May need LAN/gateway/account | Usually qualified electrical work and board space | Include labour and accessories in cost |
A monitoring method can score highly for convenience and poorly for measurement completeness at the same time. For example, an online heat-pump portal may provide an appealing monthly COP chart, yet define input energy differently from a householder’s secondary heater circuit. Likewise, an accurately calibrated meter connected to the compressor supply alone can omit a large separately wired immersion load. Neither display is necessarily defective: they answer different questions.
What counts as native heat-pump monitoring?
Native monitoring means data supplied by the equipment’s own controller through an approved interface: built-in display, a vendor app, a documented local register protocol or a supported Home Assistant integration. It does not necessarily mean cloud-free operation. These options often share the same controller data but offer different access, update frequency and scope. The ability to read a temperature, start a DHW boost or retrieve monthly kWh is specific to the model, software, gateway and vendor-account tier.
| Interface or route | What it usually offers | Where it can fail | Buyer verification |
|---|---|---|---|
| Built-in heat-pump screen | Operational readings and sometimes cumulative energy/heat | Historical export may be limited | Which totals are measured versus estimates? |
| Local Modbus TCP or supported gateway | Local temperatures, modes and documented register data | Unsupported register map, LAN issues, firmware changes | Exact model, enabled interface, read-only access and list of kWh registers |
| Manufacturer cloud API | Remote dashboard, alarms and account-specific sensors | Internet outage, vendor service changes or quotas | Cloud dependency, login, retention, subscription, polling limits |
| Independent monitoring integration | Heat-pump feed plus meter/heat-meter data | Meter placement, unit or recorder problems | No double counting, correct interval, local history |
The Home Assistant Nibe Heat Pump integration documents local connection methods including Modbus TCP on eligible S-series units. Other ecosystems use cloud paths; the Viessmann ViCare integration documents REST API limits. A working cloud integration is not proof that Home Assistant can still obtain those readings with the WAN connection unplugged. For a model-by-model comparison, see our upcoming buying guide Best Heat Pumps for Home Assistant Integration: Local APIs and Modbus (#051 in this series).
What native kWh actually means must be established. It may be measured by onboard hardware, calculated from compressor/inverter operating data, or aggregated by software. Some systems distinguish space heating, hot water, and electric heaters; others expose only headline totals. If the displayed “electricity used” number is an estimate, it remains useful for trends but is not the same as a calibrated incoming-supply reading.
What an external electricity meter buys you
A properly specified electrical submeter sits at a known distribution point and records real electricity consumed by the conductors passing through its measurement path. It can provide energy in kWh, instantaneous active power in W, and—depending on meter type—voltage, current, power factor, per-phase data or import/export. Most Home Assistant users need a stable accumulating active-energy total first; rich power-quality charts are secondary.
This is the strongest option for matching electricity consumption with tariffs, checking unexpected overnight load, and comparing energy use over matched weather periods. It also provides a reference against which vendor-reported energy can be investigated. A separate meter does not tell you whether the load came from compressor startup, a pump, defrost or resistance heating, but a load trace can suggest when to check the controller logs.
Three practical metering approaches
| Hardware style | When it fits | Advantages | Trade-offs and cautions |
|---|---|---|---|
| Single-phase DIN-rail kWh meter with documented Modbus RS-485 or pulse output | Single-phase branch circuit feeding an approved heat-pump supply | Direct cumulative kWh, clear measurement boundary, durable local reading | Board space, qualified installation, bus or pulse reader; approved current rating |
| Three-phase DIN-rail meter with local registers | Three-phase heat pump, or several line conductors that need correct combined metering | Per-phase power and whole-circuit active energy | Correct three-phase wiring/configuration and supply compatibility are essential |
| CT-based energy monitor, e.g. appropriately specified Shelly Pro 3EM | Retrofitting insight or selected circuits where board/sensor fit permits | Split-core options, fast power telemetry and Home Assistant integration | Check low-load accuracy, matching CTs, enclosure space and whether installation is appropriate |
A Shelly Pro 3EM is a useful example of an Ethernet/Wi-Fi-capable CT monitor with locally accessible data. Shelly lists class-B active energy accuracy under manufacturer testing and specifies current-dependent accuracy limits. Its sensor family has model and CT variants: pick the exact appropriate revision, not a generic “Pro 3EM” listing. Consult the Shelly Pro 3EM v2 technical page and Home Assistant Shelly integration. It is not automatically the best option for very small standby loads, and an accuracy claim on a product page is not by itself a certification that a particular installation is appropriate for settlement or billing.
For a meter with RS-485 Modbus, plan the complete system rather than just the meter: approved electrical installation, communication wiring/interface, termination where required, an available register map, isolated or otherwise correctly engineered interface hardware, and a local data recorder. Some applications favour pulse outputs or M-Bus instead. The brand matters less than obtaining an unambiguous, independently checkable accumulated kWh total and a supportable interface.
The electrical boundary is the most important design decision
Before buying hardware, ask the installer to map every electrical feed associated with the heat-pump system. A monobloc may have a main outdoor-unit supply and one or more protected auxiliary heater or cylinder supplies. A split system may have separate outdoor and indoor connections. The specifics vary, and the only safe authoritative source for a given installation is the approved circuit schedule plus the manufacturer’s instructions.
| Load or circuit | May be counted in native app? | Captured by an independent meter? | Why it matters |
|---|---|---|---|
| Compressor and drive | Usually some representation, depending on model | If its supply crosses the measurement boundary | Main operating load |
| Outdoor fan and defrost-related power | May be included in system estimate | If on the measured supply | Defrost affects both electrical input and heat delivered |
| Indoor circulation/primary pump | Model-dependent | If on the measured circuit | Can run when compressor is not producing heat |
| Control electronics and standby | Sometimes omitted or rounded to zero | If on the measured circuit and above meter sensitivity | 24-hour background usage adds up |
| Electrical backup/boost heater | May be separate or differently reported | Only if its feeder is included | Resistive heating can materially increase kWh |
| DHW immersion/legionella heater | Not guaranteed in heat-pump total | Only if its feeder is included | Large occasional energy events can distort comparisons |
| Additional external pump or trace heat | Typically outside core unit telemetry | Only if explicitly included | Changes a complete-system seasonal figure |
Never add a circuit twice. If a parent meter already measures the supply feeding a component, summing a child meter into that parent total overstates consumption. Keep separate category meters for diagnostics, while preserving one well-defined authoritative “whole heat-pump electrical input” total. If that whole-system boundary cannot be instrumented economically, report the narrower boundary honestly.
OpenEnergyMonitor distinguishes monitoring levels and recommends explicitly accounting for backup and immersion consumption in heat-pump monitoring. Its EmonTx4 heat-pump monitoring guide details separate electricity channels alongside flow, return and other sensors. The HeatpumpMonitor.org methodology also labels the electrical/thermal measurement boundaries for public comparisons; two reported seasonal efficiencies are not directly comparable if their boundaries differ.
Standby consumption: why the little number matters
A heat pump often draws electricity while it is not actively heating: electronics, communication, circulation, frost management or preheating behaviour may operate. However, there is no universal standby wattage; it changes by equipment, ambient conditions and operational state. An app that only reports “compressor consumption” may show zero while a dedicated supply meter is still counting.
| Illustrative constant background power | Per day | Over 30 days | Over 365 days |
|---|---|---|---|
| 5 W | 0.12 kWh | 3.6 kWh | 43.8 kWh |
| 15 W | 0.36 kWh | 10.8 kWh | 131.4 kWh |
| 30 W | 0.72 kWh | 21.6 kWh | 262.8 kWh |
| 50 W | 1.20 kWh | 36.0 kWh | 438.0 kWh |
These are arithmetic examples, not measured model specifications. The calculation is W × hours ÷ 1,000. Some auxiliary functions are intermittent, so assuming a constant draw will overestimate or underestimate actual annual use. The right approach is to record the measured overnight minimum, note operating status and check seasonal variation, rather than choosing an invented “typical” wattage.
At very low load, review the meter’s published accuracy envelope and starting-current threshold. A clamp sized for a large three-phase service is not necessarily precise at a few watts. Small background consumption can be lost in noise or a no-load threshold, especially if you try to infer it from a whole-home sensor where fridges, routers and appliances overlap.
Power in watts is not cumulative energy in kWh
A real-time graph showing 2,400 W says the appliance is drawing 2.4 kW at that instant. If it holds that load for exactly two hours, it will use 4.8 kWh. Real heat pumps modulate, cycle, defrost and switch to hot-water duty: the day’s electricity is the area under the power-versus-time curve, not one displayed wattage multiplied by 24.
| Sensor format | Example | Can be used as an accumulated energy source? | Notes |
|---|---|---|---|
| Instantaneous power | 2,400 W or 2.4 kW | Not directly as cumulative kWh | Supports live load charts; integration helper can estimate energy |
| Monotonic energy register | 4,862.7 kWh | Yes, if correctly classified and reliable | Preferred primary metric; check reset/scaling |
| Today’s energy, resets at midnight | 14.2 kWh | May require correct reset handling | Do not casually label as never-resetting total |
| Vendor heat output estimate | 35 kWh thermal | Not electrical use | Never confuse heat delivered with electricity imported |
| Vendor COP figure | 3.1 | No | Ratio, not energy; inspect time and measurement boundary |
Home Assistant’s Energy Dashboard FAQ explains suitable entity attributes: an electricity energy sensor normally needs device_class: energy, a compatible unit such as kWh and state_class: total or total_increasing. Home Assistant’s Integral helper can derive approximate energy from a power sensor, while its Utility Meter helper provides daily/monthly period counts. A cumulative kWh register from a reliable meter generally avoids the sampling errors inherent in integrating sporadically updated W readings.
How to set up an accurate Home Assistant monitoring dashboard
The following is a software/data workflow, not instructions to open electrical equipment or modify heat-pump wiring. Have a qualified professional install and commission all metering, protective devices and any communications hardware involving the distribution board. Once the measurements exist and are exposed through supported integrations, the remaining checks can be completed in Home Assistant.
- Inventory measurements first. Record the equipment model, inverter/controller or gateway version, available native entities, physical electricity feeds and whether any vendor electricity or thermal value is explicitly estimated.
- Label the reference kWh meter. Give it an obvious name such as “Heat pump supply — electric kWh” and document whether it includes the indoor unit, outdoor unit, immersion, backup heater and ancillary pumps.
- Add the supported local integration. Select Shelly, Modbus through the appropriately configured meter/gateway, or another officially supported meter integration. Confirm readings are available and not dependent on the internet unless clearly disclosed.
- Compare instantaneous load safely. Note the live W reading during documented operation and standby. Use the manufacturer controller for context; do not force starts or override the heat-pump safety logic to create test loads.
- Identify the cumulative energy entity. Check its unit, state class, whether it survives restarts, and if any reset causes erroneous jumps in long-term statistics.
- Add it under individual devices in Energy Dashboard. Use the heat-pump electricity submeter as a monitored individual device and retain your whole-house grid meter as the household supply boundary.
- Make daily and monthly views. The Utility Meter helper can track time periods and, where appropriate, named price tariffs. Remember the first incomplete cycle should not be used as a fair month-to-month comparison.
- Graph companion native telemetry. Show outdoor temperature, flow/return temperature, DHW mode, compressor status, defrost and backup heater status when genuinely available; do not infer unknown entities.
- Review 7–30 days of operation. Investigate anomalies against the manufacturer’s diagnostic log, outside conditions and any separately wired resistive heaters.
- Store and protect history. Record HA statistics health, back up the database/settings and make sure a short Home Assistant outage does not interrupt heating or remove the only kWh total.
Home Assistant’s individual-device energy guide describes placing appliance submeter readings in the energy view. A monitored heat-pump device is generally an individual device, not an additional “grid import” source. Adding both the whole-home utility meter and a heat-pump submeter as separate grid sources would double count part of the household import. See our guide to correct Home Assistant grid, solar and battery entities for the wider accounting model.
Native kWh versus meter kWh: a worked reconciliation
Imagine a 30-day period for a home with a heat pump and separate hot-water immersion supply. The native app reports 286 kWh for the heat-pump unit. An electrician-installed submeter reports 314 kWh for the heat-pump supply that includes its internal control electronics and pump. A second, independently metered immersion circuit reports 22 kWh. There is no evidence from the app whether those two readings use identical start and end timestamps.
| Source | Illustrative 30-day reading | Electrical boundary | Treatment |
|---|---|---|---|
| Native app “heat pump energy” | 286 kWh | Vendor-defined; confirm estimate and inclusions | Useful trend, not necessarily billing reference |
| Dedicated heat-pump feed meter | 314 kWh | Unit supply covered by this meter | Primary measured consumption for this feed |
| Separate immersion meter | 22 kWh | Hot-water resistance heater circuit | Add only if not covered by parent meter |
| Known combined metered electricity | 336 kWh | 314 + 22 kWh, non-overlapping feeders | Label as measured system electricity for this selected boundary |
| Difference from native app | 50 kWh | 336 − 286 kWh | Investigate; not automatically a metering fault |
The difference might reflect omitted auxiliaries, standby, estimate methodology, clocks, incomplete native data, or a different definition of the system. Do not assume the 50 kWh is “wasted energy” or attribute it entirely to poor COP. First align period boundaries, compare the manufacturer’s defined energy input, check heater status and inspect the installation’s actual electrical circuit schedule.
If the measured combined use is 336 kWh and the electricity price for that same period is an illustrative £0.23/kWh, the variable energy charge associated with those measured circuits is £77.28 (336 × £0.23). That excludes standing charges, time-of-use mix, network charges or VAT treatment differences. If electricity was used across two tariff periods, calculate each separately. This is an example rate, not a current supplier offer.
Can either approach measure COP? Only with thermal output
Coefficient of performance (COP) = delivered heat energy ÷ electrical input energy over the same period and boundary. Electrical kWh alone do not reveal delivered heat. A compressor power reading, outdoor and flow temperatures, or a manufacturer’s published efficiency table cannot establish measured thermal kWh. To calculate meaningful in-situ COP, you need a properly measured or clearly qualified thermal-energy figure and a matching electrical-energy figure.
| Available data | What you can calculate honestly | What you cannot claim |
|---|---|---|
| Electrical meter only | Electricity consumption, cost, demand and cycling patterns | Measured COP or heat produced |
| Native temperatures + electrical meter | Performance context, weather/flow relationship | Measured thermal output without flow and suitable heat calculation |
| Vendor-estimated thermal kWh + electric meter | Estimated COP with explicit uncertainty | Heat-meter-verified COP |
| Installed thermal energy meterAliExpress price + matched electrical input | Measured period COP for that measurement boundary | Universal efficiency for different boundary or weather |
| Manufacturer lab SCOP label | Comparison at declared test conditions | The actual home’s measured seasonal performance |
Suppose the heat meter records 1,020 kWh delivered over a month and the chosen electrical boundary consumes 340 kWh in exactly the same period. The monitored ratio is 1,020 ÷ 340 = COP 3.00. If a separate 20 kWh immersion circuit supplies extra heat and both its electrical and delivered thermal contribution are outside the chosen heat-meter boundary, simply adding the 20 kWh to the denominator creates a different system boundary. Do not publish it as the same COP without redrawing and documenting the boundary.
In a hydronic system, thermal power is related to mass flow × specific heat capacity × temperature difference. Measuring just the flow and return temperatures is insufficient: you also need reliable flow information, appropriate fluid properties and a suitably accurate thermal metering method. Antifreeze mixtures, low temperature difference, sensor placement, heat losses and defrost cycles complicate simplistic arithmetic. See OpenEnergyMonitor’s heat-pump monitoring levels for the progression from electricity-only monitoring to measured heat output and COP.
A product’s SCOP is a standardised seasonal performance rating under specified test assumptions. Your home’s seasonal observed ratio is site-specific and should be called a monitored seasonal performance factor, with the boundary and period written down. Apparent differences between homes can come from climate, heat emitters, DHW proportions, backup energy, meter accuracy, defrost treatment or heating design rather than heat-pump brand alone.
Accuracy, update intervals and availability: how the methods differ
“Accurate to 1%” describes a declared performance condition, not a guarantee that every displayed sample will be within 1% under every load. Confirm what the datasheet defines: active energy versus current accuracy, rated current range, class and calibration conditions. For comparative research and detailed COP, meter selection and installation can matter more than the number of dashboard decimal places.
| Risk | Typical consequence | Practical mitigation |
|---|---|---|
| Meter accuracy degrades at tiny loads | Standby is rounded, thresholded or noisy | Check datasheet starting current and actual intended range |
| Power polled every 60–120 seconds | Short peaks or heater cycles are missed | Use hardware energy register; do not rely on sparse W integration |
| Cloud API outage | Gaps in native temperature or app reports | Preserve independent cumulative kWh register and offline heating |
| Modbus counter changes scale after firmware | Huge false kWh jump | Document register map and check HA statistics after updates |
| CT wrong phase/orientation | Misreported or negative power | Qualified installer confirms topology, phase assignment and sign |
| Three-phase netting assumptions | Unexpected sign or mismatch | Verify per-phase and whole-circuit active-energy convention |
| Counter resets on power loss | Misleading period chart | Use documented reset semantics, examine statistics and firmware |
| Meter covers only one unit | Uncounted indoor/auxiliary supply | Publish a circuit boundary diagram and add separate meters only where needed |
For a fair comparison, collect a sufficiently long period to average out brief anomalies, but use matched times and consistent units. When a meter offers both a local display and a Home Assistant entity, verify that the change in the cumulative register matches the display. Compare against the supplier meter only as a rough whole-building sense check when many other loads contribute; it is not a direct cross-calibration of one heat-pump submeter.
Buying decision: where to spend your monitoring budget
There is no reason to buy an expensive communications gateway if it only duplicates the one temperature you can already see, just as there is little value installing an elaborate heat meter when your sole question is whether the system used 200 or 300 kWh last month. Choose your measurement level by the question you want to answer and the additional installation cost.
| Homeowner goal | First purchase or task | Optional second step | What to avoid |
|---|---|---|---|
| Basic temperatures, faults and mode | Check included controller/app and exact Home Assistant integration | Supported local gateway if offline access matters | Paying for unsupported “universal” heat-pump adaptor |
| Trustworthy running costs | Professionally fitted cumulative kWh submeter with local reading | Native mode/temperature integration for interpretation | Treating compressor-only energy as whole-system consumption |
| Detailed demand trends | Meter that provides both active power and accumulated active energy | Home Assistant recorder/history and tariff helpers | Integrating low-rate cloud power samples as sole kWh source |
| Measured COP or installation optimisation | Defined electrical boundary plus validated heat-meter system | OpenEnergyMonitor or equivalent logging and professional analysis | Calling temperature-only estimates “measured COP” |
| Local-first smart home | Supported local Modbus/pulse/Ethernet electricity meterAliExpress price | Compatible documented heat-pump local telemetry | Depending on undocumented buses or cloud-only billing data |
| Very tight budget | Use available native readings and write down limitations | Meter at installation or future board work | Unapproved smart plugAliExpress price on hardwired/high-power equipment |
Budget categories to ask installers to quote
- Native-only: included internet module or a supported gateway; account/subscription and internet dependence may be ongoing.
- Simple submeter: single-phase or three-phase electricity meter, required protective hardware and board space, installation/testing and a readable display or local data interface.
- Connected submeter: all the above, plus approved RS-485/M-Bus/Ethernet interface, cable routing where needed, a gateway or Home Assistant host and secure local networking.
- Full performance instrumentation: suitably rated calibrated electrical meters, a hydronic heat meter, flow/temperature fittings, installation/drain-down commissioning and a maintained recorder.
Request itemised quotations that identify the installed cost, not just the small retail price of a meter. For a hardwired heating plant, professional labour, distribution-board work, isolation, circuit rearrangement or meter communications can outweigh the electronics. Check the meter’s intended regional standards, warranty, vendor support, local compatibility and enclosure dimensions. Pricing fluctuates across the UK and Europe, so product-list screenshots are not a substitute for an installation-specific quote.
When to choose native only, meter only or both
Choose native integration only when your question is operational
A manufacturer-supported, accessible integration may suffice if you want to see flow and return temperatures, target room temperatures, hot-water priority or fault alerts. This option has the least installation disruption, but validate what kWh actually means and whether cloud connectivity can disappear. Store those limitations next to the dashboard so the figures are not repurposed later for inaccurate financial or COP claims.
Choose an external meter first when cost and reliability matter
If you need to compare electricity costs before and after a heating-control change, independent accumulated electric energy is the foundational signal. Pair it with outdoor temperature and a clear note of the metering boundary. This is especially useful when switching tariffs or installing a PV system. A meter remains useful even if the heat-pump manufacturer changes its app, API or subscription terms.
Choose both when you want to diagnose and optimise intelligently
Native status data turns an unexplained 6 kW spike into a hypothesis: perhaps an immersion heater ran, perhaps a defrost or scheduled hot-water cycle occurred, or perhaps the home needed more heat. The meter verifies the electrical magnitude; the controller data explains the context. Do not use Home Assistant to suppress safety-critical backup heat, defrost or anti-frost logic to make charts look better. Optimisation belongs to compatible manufacturer-approved supervisory settings and professionally commissioned heating controls.
Common mistakes and how to fix them
| Problem | Why it happens | What to do |
|---|---|---|
| Heat-pump app shows less kWh than submeter | Different load boundary, estimate, clocks or auxiliaries | Compare matched date/time and physical loads; do not adjust counters blindly |
| Energy dashboard doubles house electricity | Heat pump added as both grid source and individual device | Keep utility grid meter as grid source; use submeter as individual device |
| kWh entity missing from HA selector | Wrong unit/state class or statistics issue | Check Developer Tools → States and Statistics against HA requirements |
| COP seems implausibly high | Estimated heat, missing aux electricity or mismatched time windows | Audit both energy boundaries and sensor source |
| Standby always zero | Meter threshold, missing control feed or app compressor-only values | Check what circuit is measured and device low-load specifications |
| Monthly total resets or spikes | Counter reset/rollover incorrectly treated as cumulative | Check raw register, helpers and statistics before deleting history |
| Temperature and W graphs have different timestamps | Different polling or cloud latency | Use period energy totals, not random instantaneous divisions |
| External meter loses network but still counts | Transport/HA issue rather than loss of electrical energy | Restore collection and reconcile cumulative register; heating should remain autonomous |
| User cannot distinguish heating from DHW | Single meter aggregates all operating modes | Use native mode markers, not an unsupported automatic energy split |
For another end-to-end accounting example, see How to Verify Home Assistant Energy Data Against Your Utility Bill and Whole-Home Energy Monitoring Without Cloud. Those articles address site-level energy reconciliation and hardware selection; this guide focuses on the heat-pump measurement boundary.
Frequently asked questions
Is the heat pump’s own energy reading accurate?
It depends on the specific unit and the manufacturer’s measurement method. It may be a metered supply value, an internal calculated consumption figure, or an estimate based on operation. Check documentation and compare a sufficiently long matched period against an independent correctly located electricity meter before treating it as an accounting reference.
Does a Shelly Pro 3EM measure a heat pump’s COP?
No. It measures electrical variables and active energy within its installed sensor boundary. You still need thermal energy delivered over the same period to calculate measured COP. Be aware of the particular Shelly revision, CT rating and accuracy at the expected loads.
Which is better: a DIN-rail meter or CT clamps?
A correctly specified inline electricity meter is often attractive as a clear cumulative kWh reference; CT sensors can simplify some retrofit monitoring arrangements and offer high-resolution power data. The answer depends on board access, phases, accuracy targets, supported communications, installation compliance and cost. Neither approach is universally superior or appropriate for every circuit.
Do I need one meter for the indoor unit and one for the outdoor unit?
Only if they have separate feeds and your monitoring objective requires capturing both. A correctly positioned upstream meter may already encompass both. Ask the installer to document the circuit schedule and avoid double-counting an upstream meter with downstream submeter values.
Does a smart plug work for a heat pump?
Do not assume so. Many heat pumps are hardwired and can have large starting or auxiliary loads. A general-purpose smart plug is not an acceptable substitute for an approved electrical metering installation. Check the manufacturer’s connection requirements and use a qualified installer.
Can Home Assistant calculate kWh from a power sensor?
Yes, via its Integral helper, but this is a sampled estimate. Where a meter provides a reliable cumulative active energy counter, that is normally preferable because brief loads between polls can be missed by a sparse W feed.
Can I measure heating-only electricity separate from domestic hot water?
An external meter of the whole heat-pump feed measures combined load, not which operating mode caused it. The native controller may expose heating and DHW consumption separately, but the method may be estimated. Correlating modes with power can help, yet it does not guarantee exact energy attribution during transitions or mixed modes.
Why does the heat pump use power when it is “off”?
Standby controllers, pumps and protection functions can operate without the compressor providing space heat. Check the unit’s actual control state and whether the meter covers the relevant circuits. Never disable frost or other safety functions to eliminate a minor background load.
Is the Home Assistant Energy dashboard enough to calculate monthly cost?
It can track a correctly classified kWh source and tariffs, but your cost estimate depends on tariff windows and any fixed charges, export arrangements or discounts. Use verified supplier bills for the final figure; do not treat a single flat price as every household’s real rate.
What extra hardware is needed for true measured COP?
Typically a defined electrical-energy meter and appropriate thermal-energy instrumentation measuring delivered heat—for example a heat meter that accounts for flow and temperature difference. The plumbing work, sensor accuracy and chosen boundary matter. The calculation must use matching periods, include appropriate auxiliary loads and label any estimated components.
Will cloud API failure stop the heat pump from working?
A correctly commissioned heat pump must retain its manufacturer’s core control and protection functions independently of Home Assistant. Cloud API failure can stop remote telemetry or optional automation, but should not be a prerequisite for safe heating or frost protection. Confirm the installed equipment’s specific offline behaviour.
What should I ask my installer before buying a meter?
Ask for a circuit schedule showing everything included, a compatible metering SKU and accuracy specification, the communication interface, installation and commissioning cost, location of all auxiliary heaters, and a plan to read the accumulating kWh total locally. If measured COP is required, ask separately about thermal metering and the measurement boundary.
Official documentation and resources
Documentation reviewed 16 October 2026. Confirm exact hardware models, revision, firmware, current prices and local installation regulations before purchase. External sources below are technical documentation rather than proof of compatibility for every installation.
- Home Assistant — Energy Dashboard sensor requirements
- Home Assistant — tracking individual device energy
- Home Assistant — Integral (Riemann sum) helper
- Home Assistant — Utility Meter helper and tariffs
- Home Assistant — Nibe Heat Pump integration
- Home Assistant — Shelly device integration
- Shelly — Pro 3EM v2 product and measurement details
- OpenEnergyMonitor — Level 1/2/3 heat-pump monitoring and COP
- OpenEnergyMonitor — EmonTx4 electrical and thermal measurements
- HeatpumpMonitor.org — comparison scope and measurement boundaries
- ESP32.co.uk — correct Energy Dashboard entities
- ESP32.co.uk — verify electricity statistics against bills
- ESP32.co.uk — cloud-free energy monitor options
Safety and scope: This article is buying and data-analysis guidance, not a mains wiring diagram, heating-system commissioning procedure, or authorisation to modify refrigerant or hydronic components. Only competent, appropriately qualified installers should work on heat-pump supplies, protective devices, current transformers in distribution boards, installed heat meters and manufacturer control terminals. Follow all equipment instructions and applicable regulations. Do not bypass defrost, freeze protection, pressure switches, high-temperature cut-outs, hygiene cycles or backup protection for the sake of energy savings or a Home Assistant chart.