Quick Summary (TL;DR)
For the electricity cost of an EV charging session, measure energy at a known boundary and match it to the tariff during that session. The wallbox’s reported kWh, a dedicated circuit meter’s kWh, the utility meter’s import and the car’s estimated battery gain are different quantities. They can all be correct while displaying different numbers.
For a house without solar, a professionally installed, suitably specified EV circuit sub-meter is often the clearest way to measure AC energy delivered to the charging equipment. The utility meter remains the reference for what the electricity supplier bills, but its import also includes all other household loads. With solar or a home battery, a wallbox may consume 20 kWh even though grid import rises by only 7 kWh: the remaining energy came from other sources. Never treat the difference as unexplained charging loss.
Best practical approach: record a session start and end, use a cumulative kWh reading from a charger or dedicated meter, reconcile household grid import/export separately, and price the energy against the correct half-hourly or fixed-rate tariff. Use the car’s battery percentage for a plausibility check—not as the authoritative electricity meterAliExpress price. This article provides the equations, worked examples, Home Assistant setup and fault-finding method.
Four different measurements that people call “charging kWh”
An EVSE or wallbox is primarily a safety and control device: it communicates permitted charging current to the car, operates appropriate switching and protective functions, and often measures or estimates electricity. During ordinary AC charging, the vehicle’s onboard charger converts AC electricity into the DC energy used by its battery. That is why a wallbox’s AC reading is normally greater than the change in energy stored in the pack.
| Reading | Typical measurement boundary | What it answers | What it does not answer |
|---|---|---|---|
| Utility import register or supplier interval data | Net flow from the public grid into the property | How much electricity was imported, subject to supplier billing rules | How much the EV alone consumed |
| Dedicated EV circuit meter | AC power entering the EV charging branch | AC electrical energy attributable to that circuit | Exactly how much reached the battery or came from grid versus PV |
| Wallbox charging-session kWh | Metered or calculated AC energy at the chargepoint, depending on product | Approximately how much AC electricity the wallbox delivered in that session | Whether the energy came from grid, solar, or a home battery |
| Car-app battery change | Estimated change in the traction battery’s stored energy, often derived from state of charge | A rough check of the energy gained by the battery | Electrical energy bought from the supplier or precise AC conversion losses |
| House CT or hybrid inverter energy figures | Depends on CT position and direction conventions | Useful context for household, solar and battery flows | An independently calibrated EV-circuit bill |
Before interpreting a discrepancy, find out where each sensor is physically or logically measured. Product names such as “energy consumed”, “charged” and “delivered” are not standardised across all apps. Some systems report energy per session; others report lifetime totals, daily values or nominal battery gains. One figure cannot be reconciled against another unless both cover the same period and boundary.
A home with a 7.4 kW wallbox does not necessarily charge at 7.4 kW continuously. The vehicle may limit current, heating or cooling may run, a tariff schedule may pause the session, or load balancing may reduce available amperes. A constant power rating multiplied by elapsed clock time is therefore at best an approximation.
The simple energy balance before you calculate cost
For a property with grid supply, rooftop PV, a home battery and an EV, keep import/export separate from generation/storage. Instantaneous power is measured in kW; accumulated energy is measured in kWh. Over a selected period, a simplified site balance is:
Grid import + PV generation + battery discharge = house and EV consumption + battery charging + grid export + applicable conversion/loss terms.
This is an accounting identity across appropriately chosen AC and DC measurement boundaries. If your inverter already reports net AC PV output or battery energy after conversion, do not add losses a second time. Put all signals on consistent sides of the same measurement boundary before attempting exact reconciliation.
For an EV circuit measured downstream of the house supply, EV AC kWh is one component of whole-house consumption. It is not an additional purchase from the grid when the utility meter already includes that energy. Home Assistant reflects this distinction: grid sources feed the energy balance, whereas an EV meter should normally appear as an individual device; hierarchy controls are available for nested device meters. See the official Home Assistant individual-device documentation.
| Installation | What the wallbox can show | What the supplier bills | Why the numbers differ |
|---|---|---|---|
| Grid-only, EV plus other loads | EV session kWh | Net household import across all loads | Oven, heat pump, standby loads and time windows |
| Solar-only house | EV AC kWh | Grid import minus separately accounted exports under applicable tariff | Solar supplied some EV or household load |
| Solar plus home battery | EV AC kWh | Grid import plus applicable standing charges and export credits | Battery and PV energy may supply the EV |
| Charger on a detached garage sub-board | EVSE output kWh | Whole-site import | Sub-board energy, cable losses and other garage circuits may be inside or outside the metering boundary |
| Multiple EVs sharing a chargepoint | Session or user-tag kWh | Whole-site import | Session attribution and shared loads require separate records |
Worked example 1: A grid-only overnight charge
Assume a household charges from 00:30 to 04:30. The wallbox reports 24.0 kWh AC and the dedicated EV branch meter records 24.4 kWh. Over exactly the same four hours, the utility meter imports 29.6 kWh. Household non-EV demand accounts for 5.2 kWh in this illustration; 29.6 minus 24.4 gives 5.2.
| Measurement | Example result | Interpretation |
|---|---|---|
| Utility import increment | 29.6 kWh | All electricity imported by the premises |
| EV branch meter increment | 24.4 kWh | Energy at the branch meter boundary |
| Wallbox session estimate | 24.0 kWh | Wallbox’s defined session boundary |
| Remaining non-EV import (simplified) | 5.2 kWh | 29.6 − 24.4 |
| Difference: branch meter minus wallbox | 0.4 kWh | 1.64% of branch reading; may include boundary, measurement and timing differences |
The 0.4 kWh must not automatically be labelled charging loss. It may be cable and charger auxiliary power, differences in the energy-counter definition, timestamp mismatches or accuracy limits. Battery conversion losses mostly occur further downstream in the vehicle, so they would not be isolated by subtracting the wallbox figure from a branch input meter.
At an illustrative import tariff of £0.10/kWh throughout this period, the branch-meter energy costs 24.4 × £0.10 = £2.44. The entire imported household energy costs 29.6 × £0.10 = £2.96, excluding any standing charge. These are example assumptions, not a claim about the current price of any UK tariff.
If you use the wallbox reading instead, 24.0 × £0.10 gives £2.40. That £0.04 difference does not justify assuming the meter is defective. The purpose of reconciliation is to understand the boundary and uncertainty, not force every reading to match to the last watt-hour.
Worked example 2: Why solar makes a charger look “larger” than the bill
Consider a four-hour afternoon session with a measured 20 kWh EV branch consumption. During the same period, PV supplies 12 kWh of usable AC energy and the site imports 13 kWh. Other household loads consume 5 kWh, with no battery charging, discharging or export in this simplified example. The site balance is 12 + 13 = 20 + 5 kWh.
The utility import of 13 kWh is not the EV’s exact grid share. It pays for the combined site demand after solar offset, and there is no physical label on grid and solar electrons in a shared AC installation. A cost calculation requires an allocation method, not just an energy-meter subtraction.
| Item during identical four-hour window | Energy |
|---|---|
| EV branch consumption | 20 kWh |
| Other house consumption | 5 kWh |
| AC solar production used on site | 12 kWh |
| Grid import | 13 kWh |
| Grid export and battery movement | 0 kWh |
| Check | 12 + 13 = 20 + 5 |
You could assign solar production to other house loads first, allocate proportionally, or compare the EV session with a well-matched baseline. Each answers a different economic question. For determining whether the EV displaced valuable export, the foregone export tariff may be more important than the import tariff. If export would otherwise have earned £0.12/kWh, using 12 solar kWh on site can carry an opportunity cost of £1.44—even though the import bill did not increase by that amount. Whether that is the right marginal valuation depends on the actual site controls, export eligibility and what would have happened without EV charging.
A home battery adds further ambiguity: energy used by the EV may have been purchased overnight, stored with losses and discharged later. Charging the EV at noon does not automatically mean the energy was free. The correct accounting approach records battery charging and discharge, any tariffs that funded the stored energy, and the cost of cycle losses, without pretending that an instantaneous grid import sensor can uniquely trace energy origin.
Worked example 3: Charger kWh versus the car’s battery percentage
Suppose a vehicle has a nominal 75 kWh battery and the app reports its state of charge rising from 30% to 60%. A naive calculation gives 75 × 0.30 = 22.5 kWh. But nominal capacity is not necessarily usable capacity, the BMS SOC estimate can be rounded or adjusted, and the car may simultaneously consume power for thermal management. It is not a calibrated revenue meter.
If the dedicated AC circuit meter recorded 25.2 kWh in that interval, then 22.5 ÷ 25.2 = 89.3% is only an illustrative apparent stored-energy ratio under the nominal-capacity assumption. It is not a measured charging efficiency. Real efficiency estimation requires a credible usable-capacity estimate or directly measured battery-side energy, matched start/end conditions, and allowance for auxiliary consumption. Do not use the SOC arithmetic to argue that the wallbox is overbilling.
| Common shortcut | Why it misleads | Better test |
|---|---|---|
| Battery capacity × change in displayed SOC | Capacity, rounding and BMS estimation uncertainty | Use consistent conditions and battery-side telemetry if properly available |
| Charger power × wall-clock duration | Charging is rarely at full power for the entire window | Use the cumulative session or branch energy reading |
| Grid import during session = EV demand | Includes other house loads; excludes solar/battery contribution | Use a dedicated EV branch meter for EV AC demand |
| Car trip efficiency × distance added | Driving efficiency is variable and separate from charging energy | Keep trip-consumption estimates out of billing reconciliation |
Which meter should you buy? Four sensible monitoring architectures
For household cost estimation, it may be sufficient to use the wallbox’s reliable cumulative energy sensor and the right tariff. For independent reconciliation, a professionally installed EV branch energy meterAliExpress price usually provides a clearer boundary. For formal reimbursement, tenancy allocation or resale, ask for the particular metering standard and evidence your agreement requires. “Has a kWh counter” is not the same claim as “approved for this billing application”.
| Monitoring choice | Strengths | Weaknesses | Best use |
|---|---|---|---|
| Wallbox built-in cumulative meter | No additional installation; session attribution; often integrates with app or Home Assistant | Some figures calculated rather than independently metered; cloud dependence and counter resets vary | Household charging cost estimates and convenience |
| Manufacturer-supported EV branch sub-meter | Independent electrical boundary and often reliable kWh registers | Professional fitting, board space, gateway and procurement checks | Whole-EV-circuit reconciliation and monitoring |
| DIN-rail meter with supported RS485 Modbus | Potentially local long-term counters and Home Assistant integration | Electrical design, communications isolation, register mapping and model-specific accuracy classes | Technical installations with installer support |
| Whole-house CT energy monitor | Useful site balancing and simultaneous household power | Accuracy varies with CT placement and phase setup; EV demand may not be separately attributable | Context for site-level import/export and load balancing |
| Utility smart meter / supplier account | Primary import/export billing reference | Delays, netting, tariff settlement intervals and no direct EV attribution | Reconcile supplier invoices and overall imports |
If choosing a sub-meter, confirm single- or three-phase support, rated and maximum current, accuracy class, import/export direction, whether current transformers are integral or matched, pulse/Modbus interface, meter certification, cumulative total behaviour after outages and a compatible read-only data connection. For billing or reimbursement, ask what verification, sealing or MID approval the particular context requires rather than assuming a generic marketplace meter qualifies.
A 32 A single-phase EV charging circuit is not an argument for fitting a loose 32 A gadget inside a distribution board. A qualified installer must consider design current, overload protection, conductor ratings, installation environment, isolation and relevant regulations. EV charging equipment has additional protection requirements that a general-purpose smart plugAliExpress price cannot replace. Do not use a consumer smart plugAliExpress price as a metering adaptor for a high-current EV charger.
For a side-by-side introduction to energy-meter protocols, see the existing ESP32 SDM120/SDM630 Modbus energy-meter guide. Installation of mains devices must be designed and performed by a competent professional.
Are smart wallbox energy readings legally “accurate” in Great Britain?
The UK Electric Vehicles (Smart Charge Points) Regulations 2021 impose requirements on chargepoints within their scope. The regulations require those products to measure or calculate energy use and power, and specify a within-10% accuracy requirement for defined readings with inaccuracies not systematic. They also require relevant energy-usage information to be available to the owner. These are product compliance obligations within the scope of that legislation, not a promise that every model, installation or individual session reading will match the electricity supplier’s settlement meter within 10%.
Crucially, the regulation’s permitted measurement or calculation and its tolerance should not be confused with the specifications of a certified meter used for an invoice, employee reimbursement or subletting. For a cost claim involving third parties, check the contract and relevant metrology rules with the responsible party. A tariff-sensitive estimate can be useful even when it is not suitable for legal billing.
| Metering claim | What to ask for |
|---|---|
| “The charger shows session kWh” | Does the value come from an embedded meter, power integration or car-reported data? |
| “Meets smart-chargepoint rules” | Which product model, jurisdiction and version of the relevant regulations? |
| “MID compliant” | The specific model, accuracy class, certificate, measurement boundary and installation conditions |
| “Suitable for business reimbursement” | Employer policy, required evidence, treatment of solar, standing charges and taxes |
| “Accurate for load balancing” | Response latency, phase measurement, CT pairing and safe control behaviour on sensor loss |
No single label substitutes for reading the manufacturer’s datasheet and the contractual requirements. A meter used for load limiting and a meter used to settle money have different priorities, even if both display kWh.
How to calculate a charging session accurately on a flat tariff
Begin with an energy counter that accumulates rather than an instantaneous kW value, and record its value before and after a full charging session. For a counter reading 1,208.4 kWh at the beginning and 1,232.9 kWh at the end, the measured increment is 24.5 kWh. At a flat, fully loaded marginal tariff of £0.24/kWh, the estimated variable cost is £5.88. That is simply 24.5 × 0.24.
Use the price component that changes with the charging session. A household standing charge that would be payable whether or not the EV was charged is generally not incremental to the session, though it belongs in total household-cost reporting. If the tariff rate quoted excludes VAT, convert it to the actual billed marginal rate before comparing against a supplier invoice.
When the supplier offers discounted rates for all household consumption during an EV window, remember that the EV is not responsible for other devices using the cheap period. Price EV branch kWh at the corresponding times, and separately record any fixed subscription or tariff premium if comparing annual total cost against another tariff.
Example spreadsheet columns
| Field | Example | Why keep it |
|---|---|---|
| Session start and end | 00:30–04:30 local | Align tariff and meter windows |
| Start cumulative EV kWh | 1,208.4 | Avoid relying on resettable session counters |
| End cumulative EV kWh | 1,232.9 | Difference is 24.5 kWh |
| Tariff applied | £0.24 per kWh | Include correct tax treatment |
| EV AC energy cost | £5.88 | 24.5 × £0.24 |
| Utility import for same time | Separately recorded | Reconcile, not substitute for EV energy |
| PV/battery movement | Separately recorded | Explain energy-source effects |
Variable tariffs: price each interval, not the entire session at one rate
Dynamic pricing and schedules spanning tariff boundaries require more care. For each interval, multiply the EV meter’s incremental kWh in that interval by the applicable import rate. Add the interval amounts. Do not multiply the total session kWh by the cheapest hour’s rate or by a non-weighted average of the tariff’s published prices.
| Half-hour slot | EV energy (kWh) | Illustrative rate | Cost |
|---|---|---|---|
| 01:00–01:30 | 3.2 | £0.08/kWh | £0.256 |
| 01:30–02:00 | 3.4 | £0.12/kWh | £0.408 |
| 02:00–02:30 | 3.0 | £0.20/kWh | £0.600 |
| 02:30–03:00 | 2.8 | £0.06/kWh | £0.168 |
| Total | 12.4 | Weighted £0.1155/kWh | £1.432 |
All prices are illustrative. The weighted price is £1.432 / 12.4 kWh = £0.11548/kWh, or about 11.55 p/kWh. If your supplier publishes local-time settlement periods, handle daylight-saving changes and duplicated/missing clock hours using the supplier’s settlement timestamps, ideally with time-zone-aware timestamps. A one-hour misalignment can create a much larger error than the metering tolerance.
Cost of grid imports is not identical to marginal cost of solar charging. If a PV system could otherwise export at a specified rate, home consumption displaces potential export income. If solar would otherwise be curtailed, the opportunity cost can be different. State which comparison you’re making: cash paid to the supplier, total economic cost, or incremental cost versus charging at a different hour.
Home Assistant: Set up energy sources and the EV as a device
The Home Assistant Energy dashboard supports site grid import/export, solar production, home battery movement and individual devices. An EV charger measured as part of the household electrical system belongs in individual device consumption, not as a second grid supply. Adding the same branch kWh as grid import on top of your actual utility meter will double count demand.
A suitable energy sensor should have device_class: energy, units such as kWh, and the correct state class—commonly total_increasing for a cumulative register that only grows apart from known resets. A live kW power sensor is not a cumulative kWh meter. Home Assistant’s energy FAQ explains eligible sensor attributes and statistics troubleshooting is useful when an entity does not appear.
The Utility Meter helper can track daily or monthly consumption from a cumulative source, including tariff-specific buckets. It does not improve the original meter’s accuracy. In the example below, the entity ID is a placeholder; replace it only with a verified, monotonically accumulating EV AC energy sensor. Check that a source does not reset with each session or unexpectedly jump after reconnecting.
# Example configuration.yaml; adapt the entity to your actual meter.
utility_meter:
ev_ac_energy_daily:
source: sensor.ev_circuit_energy_total
cycle: daily
ev_ac_energy_monthly:
source: sensor.ev_circuit_energy_total
cycle: monthlyIf you only have a continuous power sensor in W, Home Assistant can integrate power over time using the Riemann-sum integral helper. Sampling frequency, data gaps and integration method affect the result. Prefer a reliable hardware cumulative kWh counter when it exists; calculation from instantaneous power is a fallback, not automatically a better meter.
Do not put an already cumulative kWh sensor through an integration-of-power helper. That would integrate the wrong quantity. Likewise, never silently convert an instantaneous power sensor to kWh just by changing the unit label.
Build a useful side-by-side dashboard
Before changing automations, add a small diagnostic view with EV circuit power, EV cumulative kWh, utility net import/export power, utility import kWh, solar AC power, home-battery charge/discharge power and tariff. When inspecting the differences, use the same time span for every line and note data-source latency.
At a minimum, examine three cases: charging with no PV, charging while PV is producing, and a session that crosses a tariff change. A fourth useful case is after a Home Assistant restart: confirm cumulative meter values are preserved and long-term statistics do not invent a spike. Maintain a separate log of charger-session start/end times, as the charger may report its session data after a cloud delay.
If you have sub-meters upstream of other individual device meters, Home Assistant allows parent-child relationships between individual devices to help avoid hierarchical duplication. This is separate from setting the whole-house utility meter as a grid source. Read the current individual-device hierarchy guidance before restructuring the dashboard.
Troubleshooting: Eight common reasons the numbers do not match
| Symptom | Likely explanation | What to check safely |
|---|---|---|
| Utility kWh much higher than EV kWh | Other domestic appliances used energy in the period | Observe base load before/after charging; align timestamps |
| Utility import much lower than EV kWh | Solar or home battery supplied some site demand | Compare PV, battery and import/export data over exactly the same interval |
| Wallbox shows more than car SOC implies | AC-DC conversion, vehicle auxiliaries, approximate SOC/capacity | Treat car SOC as a plausibility check, not a calibrated meter |
| EV circuit meter modestly exceeds wallbox | Different measurement boundary, standby losses or counter accuracy | Compare full sessions on identical timestamps; obtain meter specifications |
| Charging power is half what you expect | Onboard vehicle limit, load balancing or reduced phases/current | Check car and EVSE charging-current logs without bypassing protection |
| Session kWh jumps after app reconnect | App delay, backfilled data, duplicate session entry | Compare cumulative local meter and event timestamps |
| Home Assistant energy jumps on reboot | Counter rollover/reset or incorrect state class | Inspect statistics and integration documentation before repairing data |
| Three-phase CT monitor reads strange values | Phase mapping, direction or incompatible CT/meter configuration | Have the installer verify CT placement and mapping; no live-board DIY adjustments |
Some gaps are explainable; others merit testing. A steady 20–30% disagreement between two meters that supposedly measure the same branch and period deserves closer investigation. A 20% difference between household import and the EV’s session is not inherently evidence of a fault, because the boundaries differ.
A seven-step reconciliation test you can repeat monthly
- Identify every meter boundary. Draw a simple one-line diagram showing the utility meter, PV inverter, battery, EV branch and wallbox. Label which quantities are AC and which are DC.
- Choose a low-complexity test period. Prefer a normal grid-only charging interval, with solar generation and home-battery output absent or measured separately. Do not disconnect safety systems merely to simplify the test.
- Capture timestamped starting readings. Record cumulative EV branch kWh, utility import and the wallbox session counter before charging. Synchronise clocks where possible.
- Log actual power if available. A graph often reveals charger pauses, reduced current, load balancing or delayed reporting that a pair of kWh figures would conceal.
- Capture ending readings for the same boundary. Use cumulative counter differences and record each sensor’s last update time. Avoid mixing a 03:00 charger ending timestamp with a 03:30 supplier settlement window.
- Build separate site and EV balances. Explain net grid import with PV and battery data; compare EV branch energy with wallbox energy independently. Do not confuse site balance error with battery charge efficiency.
- Price only matched interval energy. Use the tariff actually applied, then compare the result against the household bill at the appropriate scope. Record your measurement uncertainty and any known gaps.
Run the same test on at least two sessions, including one with solar production or mixed rates, before changing your regular reporting. Persistent unexplained discrepancies should go to the charger manufacturer, metering installer or electricity supplier depending on which boundary is affected.
Costs and the buying decision
The cheapest approach is usually to exploit a wallbox’s existing useful energy data. Buy additional hardware only when it answers a distinct question: verified EV circuit consumption, better local reporting, a contractual reimbursement requirement, or safe dynamic load management supported by the charger. A sub-meter installed solely to make a dashboard look more precise may offer little value if the charger already supplies a credible energy register.
| Requirement | Minimum sensible solution | Worth paying extra for? |
|---|---|---|
| Personal nightly cost estimate | Wallbox kWh + applicable tariff | Usually not, unless the reported figure proves unreliable |
| Reconcile EV draw against whole-house energy | Dedicated, correctly specified EV branch meter | Often yes for data-focused owners |
| Home Assistant operation without cloud access | Verified local charger API or locally readable meter | Yes if internet independence matters |
| EV and household dynamic load limiting | Charger-approved meter and safety-rated local controller | Yes; not equivalent to a generic dashboard meter |
| Reimbursement or sale of energy | Metering evidence specified by employer/contract and law | Potentially essential; verify before buying |
| Solar and battery cost allocation | Import/export + PV/battery readings + clear economic assumptions | Useful for optimisation, but not a promise of exact physical energy tracing |
Budget for the complete installed solution: the meter, compatible CTs if required, appropriately specified enclosure/space, electrician labour, configuration, local communications, and any software or cloud fees. Do not invent a universal UK installed cost: variation in board layout, supply type and installer requirements overwhelms a simplistic parts-only estimate.
Before buying a DIN-rail or CT meter, ask the installer whether your selected charger or inverter supports that exact model, what measurement data will remain accessible locally, and whether exported kWh can be read as a separate register. If the purpose is cost reimbursement, ask for evidence of suitability rather than assuming a consumer API metric is acceptable.
Frequently asked questions
Why does my smart charger show 30 kWh when the utility meter rose by only 18 kWh?
First check that both readings cover the same clock interval. With solar or a home battery, some EV AC energy may have been supplied without net grid import. The utility register also contains other loads and may be subject to delayed settlement updates. You cannot calculate EV charging loss by subtracting these figures until you reconcile the rest of the site.
Does the difference between car battery kWh and wallbox kWh equal charging losses?
Not reliably. A car’s battery-energy figure may be based on estimated state of charge and nominal capacity; its auxiliaries may consume energy. The wallbox measures or estimates AC energy at another boundary. Without a credible battery-side measurement and matched conditions, their difference is not a precision efficiency test.
Is a MID-certified meter needed for Home Assistant?
Not for an ordinary personal energy dashboard as such. For charging reimbursement, selling energy or splitting bills with another party, the agreement and jurisdiction may demand particular metering evidence or approval. Check that separately; the presence of an RS485 port or “kWh” field does not prove legal suitability.
Can I get exact EV cost from the Home Assistant Energy dashboard?
It can provide useful estimates where the inputs and tariff are reliable. Dynamic tariffs need accurately time-aligned readings; energy sourced from PV or a home battery requires an allocation method for economic cost. The Energy dashboard is a monitoring tool, not automatically your electricity supplier’s billing engine.
Which reading should control load balancing?
Use the metering and controller architecture supported and approved by the charging-equipment manufacturer and installer. Home Assistant can supervise and visualise but should not replace mandatory overload protection or equipment-specific fail-safe controls.
Bottom line
Use three distinct ledgers: the utility meter for imported/exported energy and the supplier bill, the EV circuit or wallbox meter for AC charging consumption, and vehicle battery data for approximate stored-energy context. Keeping those boundaries separate makes disagreements intelligible and prevents bad automation, incorrect tariff calculations and double-counted Home Assistant charts.
If the goal is a more accurate estimate of charging cost, start with a reliable EV kWh counter and interval-matched tariffs. Add a dedicated meter if the current wallbox data cannot answer your question. Where solar and storage are involved, record the full site energy balance and make the opportunity-cost assumption explicit. A meter cannot solve an accounting question until you decide what it is measuring.
Sources and further reading
- Home Assistant: Energy dashboard — grid, solar, battery and individual device accounting.
- Home Assistant: Integrating individual devices — submeter hierarchies and preventing double counting.
- Home Assistant: Energy FAQ — energy sensor attributes and troubleshooting.
- Home Assistant: Utility Meter — usage cycles and tariff-specific consumption helpers.
- Home Assistant: Sensor entity guidance — state class and reset semantics.
- UK Electric Vehicles (Smart Charge Points) Regulations 2021 — smart chargepoint energy information and accuracy rules within scope.
- Existing ESP32 guide: Verify Home Assistant Energy Data Against Your Utility Bill — reconciliation workflow.
- Existing ESP32 guide: ESP32 Modbus Energy Meter with Home Assistant — meter monitoring fundamentals.
