Costs & Savings¶

This page explains every cost and savings value miniEMS exposes on the dashboard and as a Home Assistant sensor: what it means, how it's calculated, and a worked example. For the plain entity/unit reference see HA Sensors; for the full formulas with code references see the "Costs & Savings" section in Calculations.

How the values come to be¶

All cost and energy values are accumulated by CostOptimizer a little more on every EMS tick (30 seconds by default) — never recalculated once at day's end. Two consequences worth knowing when reading the values:

• Daily values grow over the course of the day and reset to 0 at midnight.
• Restarting the add-on doesn't lose what's already accumulated: everything accumulated so far is restored from the SQLite database before the first new tick is processed. What flowed during the downtime itself, a pure tick accumulation fundamentally cannot make up for – see the next paragraph.

For three kWh quantities (grid import, feed-in, load), miniEMS prefers the inverter's own daily counter when it's configured — more accurate, because a brief measurement gap and add-on restarts don't affect the inverter-internal counter, while a pure tick accumulation is vulnerable to both (every restart leaves a gap that's never made up). If the respective entity isn't set, miniEMS accumulates from instantaneous power instead. Where that applies is noted at each value.

Spike filtering: before any power reading enters a calculation, miniEMS checks it for implausible jumps (more than 500 W and more than 50% change from the last accepted value). Such an outlier — typically a brief communication glitch to the inverter — is rejected and does not affect the daily values.

Daily values¶

Grid cost today¶

sensor.miniems_today_grid_cost_eur — what you actually paid for grid power today (or would have, on a prepay tariff), at whatever dynamic price was in effect at the time.

On every tick with grid import (not export):

cost += (grid_power_W / 1000) × tick_duration_h × current_price_€/kWh

Example: A 30-second tick (= 0.008333 h) with 1200 W of grid import at €0.2744/kWh (the cheap tier) contributes 1.2 kW × 0.008333 h × €0.2744/kWh ≈ €0.0027 to the daily total. Over the whole day, that adds up — depending on how much was drawn at which tier — to your real daily bill.

PV savings today¶

sensor.miniems_today_pv_savings_eur — the amount you saved because PV power covered household consumption instead of buying it from the grid at the current price. Only the share of PV that was directly consumed counts — PV exported to the grid is excluded here (it's compensated separately, see below).

pv_to_load_W = min(pv_power_W, load_power_W)     # never more than is actually needed
savings     += (pv_to_load_W / 1000) × tick_duration_h × current_price_€/kWh

Valued at the current spot price — the same kWh of PV saves more against expensive-tier load than against cheap-tier load. That's intentional: it reflects the actual financial benefit of your system, not just the energy produced.

Total load cost today¶

sensor.miniems_today_load_cost_eur — the hypothetical cost if your entire household consumption today had come from the grid at the current spot price — regardless of whether it was actually covered by PV, battery, or grid:

load_cost += (load_power_W / 1000) × tick_duration_h × current_price_€/kWh

This value is always greater than or equal to the real grid cost, because PV and the battery offset part of it. The gap between the two is a direct measure of the combined value of your PV system and storage.

The cost itself is necessarily tick-based, like grid cost above – it needs the price at every point in time. The underlying kWh figure (today_load_total_kwh) can, since v2.0.3, optionally come from the inverter's own daily counter (load_consumption_entity) instead of being purely extrapolated from ticks – just like grid import and feed-in. Without that entity, the value was structurally fragile: every add-on restart left a gap that was never made up, because there was no hardware counter to anchor it to. Measured live: after a single restart, today_load_total_kwh was already ~1.1 kWh (≈23%) below the inverter's own daily total.

Feed-in revenue today¶

sensor.miniems_today_feed_in_revenue_eur — your income from PV exported to the grid, at the fixed feed-in tariff (feed_in_tariff_eur_kwh, default €0.08) — not the dynamic spot price, since feed-in compensation is contractually fixed.

feed_in_kWh = energy exported today (from the inverter's counter or instantaneous power)
revenue     = feed_in_kWh × feed_in_tariff_eur_kwh

The underlying kWh figure shows on the dashboard card "Feed-in Today" (today_feed_in_kwh); only the euro amount is exposed as its own HA sensor.

Grid charging today¶

sensor.miniems_today_grid_charge_kwh (energy) and the cost derived from it show how much your battery charged from the grid today — as opposed to from PV surplus. This is derived from the power balance independently of the current EMS mode:

battery_charge_W = max(0, −battery_power_W)         # negative = charging
pv_surplus_W      = max(0, pv_power_W − load_power_W)
grid_charge_W     = max(0, battery_charge_W − pv_surplus_W)

In other words: whatever the battery charges beyond what the PV surplus can supply must have come from the grid. The cost of that share is accumulated at the current price, just like grid cost above.

Cost without grid charging¶

sensor.miniems_today_cost_without_grid_charge — what your grid bill today would have been if the battery had not charged from the grid:

cost without grid charging = max(0, grid_cost_today − grid_charge_cost_today)

Immediately shows whether the grid-charging strategy increased or reduced your total cost today, once you compare it against the real grid cost.

Cost at fixed price¶

sensor.miniems_today_cost_fix_price_tariff — for comparison: what the same day's load would have cost on a classic fixed-price tariff, instead of your dynamic one:

fixed-price cost = total_load_kWh_today × fix_price + daily_base_price_eur

fix_price (default €0.30/kWh) and the optional daily standing charge daily_base_price_eur (default 0, appears as its own sensor sensor.miniems_today_base_price_eur once set) are configurable in Settings. This answers at a glance whether the dynamic tariff pays off for you.

The four cost scenarios compared¶

Four sensors together answer the same question from four angles — handy for a shared dashboard card or an automation comparison:

Weekly, monthly and yearly totals¶

The same grid cost and PV savings amounts are additionally available rolling and calendar-aggregated:

The week value is rolling (always the last 7 days), month and year are calendar based (each starts on the 1^st). All three build on the same daily values described above — nothing is recalculated, only summed.

Advanced metrics (optional)¶

Three further sensors become available once the additional inverter sensors for balance-based calculation are configured (see the "Advanced Sensors for Balance-Based Cost Calculation" section in Configuration). They supplement, not replace, the values above.

Inverter efficiency¶

sensor.miniems_today_efficiency_pct — how much of today's PV production is actually usable (the rest is lost as conversion loss in the inverter):

η = (pv_production_today − losses_today) / pv_production_today × 100

Example: 32.5 kWh production, 2.6 kWh losses → η = 92.0%. Typical values range from 88% to 95%.

Grid charging (energy balance)¶

sensor.miniems_today_grid_charge_kwh_bilanz — the same quantity as "grid charging today" above, but calculated from the inverter's daily total counters instead of instantaneous power — more robust against brief measurement gaps:

grid_charge_energy = grid_import_today − household_consumption_today + battery_discharge_today

Why not just take grid import? Part of it directly covers household consumption. The formula subtracts that share, leaving only the actual battery-charging portion.

Grid-charging strategy ROI¶

sensor.miniems_today_grid_charge_roi_eur — the key sensor for judging whether grid charging paid off financially today:

usable_energy = grid_charge_kWh_bilanz × η
savings       = usable_energy × avg_discharge_tariff_eur_kwh
roi           = savings − grid_charge_cost_today

avg_discharge_tariff_eur_kwh is the tariff the charged energy avoided when discharged (typically evening/night at the high tier) — configurable in Settings, default 0.0 = sensor disabled.

Example: 5.5 kWh grid charging, η = 92%, discharge tariff €0.394/kWh, charging cost €1.51:

usable energy = 5.5 × 0.92        = 5.06 kWh
savings        = 5.06 × 0.394      = €1.99
roi            = 1.99 − 1.51       = +€0.48  ✅

• Positive ROI → grid charging paid off today.
• Negative ROI → grid charging cost more than the benefit it later provides.

Where the values show up¶

• Dashboard (/) — the key daily values as cards, see Dashboard & UI.
• Home Assistant — every value described here is available as its own sensor, with long-term statistics support. Full entity list in HA Sensors.
• Database tab (/database) — all daily values searchable historically, see Data Storage.