Delivered energy | kotona.app

No capital, no time, no intention to execute. Published because the analysis is the product.

Correction, 2026-09-07. The original unit-economics table allocated €0.06 per delivered kWh to pack amortization. That figure divides the pack’s cost by cycle life the pack never gets to use: at five swaps per winter over a ten-year calendar life, the pack delivers about 50 cycles, and €30/kWh of acquisition spread over 50 cycles is €0.60/kWh — ten times the published number. The table and every claim priced from it are rebuilt below. The niche got smaller; the implications at the end got stronger.

A business plan for reviving the Finnish heating-oil delivery model with second-life batteries: a truck brings charged packs, takes empty ones away, and the customer never thinks about energy logistics again. The corrected plan concludes the seasonal-swap service prices like diesel rather than like grid, survives only where the alternative is also priced like diesel, and that its most interesting output is not a company but a set of implications for EVs, home batteries, and off-grid design.

The frame

Until the 2010s, a large share of Finnish detached houses heated with oil. The product was not oil; it was delivered energy — an annual truck visit, a full tank, zero customer involvement in the supply chain. The model died because the energy carrier died, not because the service model was wrong.

The proposal: rebuild the service model on batteries. Monthly or seasonal swap of charged packs, fixed-fee contract, depot charging done where electricity is cheapest — directly behind the meter at wind and solar plants, on curtailed and negative-price hours, using second-life EV packs whose remaining calendar life vastly exceeds their remaining cycle life.

Why now (and why mostly not)

Five environmental curves matter. Three bend toward the idea, one bends away, one cuts both ways.

Pack prices. Stationary-storage packs averaged $70/kWh in 2025, down 45% in one year according to BloombergNEF’s 2025 survey. The cheapest observed LFP packs hit $50/kWh. The model assumes second-life EV packs at €20–40/kWh as Europe’s first major EV retirement wave lands ~2027–2030. Direction: strongly favorable, but the second-life price is still an assumption rather than a quoted supply contract.

Second-life fit. The core economic flaw of swap-based heating is cycle waste: a pack swapped monthly cycles ~12 times a year, throwing away thousands of cycles of paid-for cycle life. Second-life packs are the mirror image — degraded cycle count, abundant calendar life, priced accordingly. A low-cycle seasonal application is the structurally correct home for retired EV batteries. This is the single strongest argument the concept has.

Distribution tariff drift. Finnish DSOs are shifting cost recovery toward fixed and power-based (tehomaksu) charges under cabling-mandate capex pressure. Every euro moved from variable to fixed strengthens the pitch to low-consumption customers, for whom effective all-in distribution cost already reaches €0.30–0.45/kWh. Rural connection fees of €15–45k keep rising. Direction: favorable, and it strengthens before the regulatory countermeasure arrives.

Cheap charging hours — the eroding enabler. Finland logged 725 negative-price hours in 2024 and 465 in 2025, according to Finnish Energy’s 2025 review using ENTSO-E data. Flexible demand — including grid batteries and electric boilers in district heating — is absorbing more of the spread. The arbitrage the depot depends on is being competed away by stationary players with no trucks. Direction: unfavorable; the model assigns a 40–50% probability that the spread stays exploitable through 2030.

Regulatory seams. Every margin component — avoided fixed fees, avoided winter transfer, sähkövero classification of a charged pack as goods rather than network-delivered electricity, customs treatment at the SE1/FI border — is an institutional artifact, not a physical one. Seams pay until a working group convenes. Direction: each seam individually short-lived; the portfolio of seams probably renews.

Net: the environment selects for the boring version (off-grid energy-as-a-service in places where the grid is expensive, provided the packs earn year-round rather than idling between swaps) and against the exciting version (national tariff arbitrage on wheels).

Unit economics

The model that matters is lifetime delivered energy, not nameplate capacity. Every assumption is on the table so it can be replaced:

Assumption	Value
Pack, nominal	400 kWh second-life, €30/kWh
Pack acquisition	€12,000
Usable energy per swap (80% depth)	320 kWh
Seasonal cottage service	5 swaps/winter
Calendar life in service	10 years → ~50 swaps
Lifetime delivered energy	50 × 320 = 16,000 kWh
Retained value after service (stationary use)	€2,000
Net capital per delivered kWh	€10,000 / 16,000 = €0.63

Cost stack per delivered kWh on those assumptions:

Component	€/kWh
Pack amortization (calendar-limited, ~50 lifetime swaps)	0.63
Depot charging (behind-the-meter PPA, curtailment hours)	0.00–0.02
Logistics (routed swap, ~8 stops/day)	0.03–0.05
Depot, handling, overhead	0.02
Delivered cost	0.68–0.72

That is before losses, finance, and everything not listed. Against effective grid cost of €0.30–0.45/kWh at cottage consumption levels, the seasonal swap loses outright — the spread the original table claimed does not exist. What the service actually competes with is generation at a site with no wire at all: a diesel generator’s all-in cost of roughly €0.50–1.00/kWh, and the €15–45k connection fee a low-consumption site never wants to pay. The service is priced like diesel, and it survives only against alternatives priced like diesel.

The way back toward the original number is utilization, and it inverts the company. A pack that does five cottage swaps a year and otherwise sits idle carries the full €12,000 on 16,000 lifetime kWh. A pack that works at the depot between swaps — spot arbitrage, reserve markets, the stationary business already named as the fallback — can see hundreds of equivalent full cycles, and amortization per kWh falls toward €0.06–0.12 only when that stationary revenue carries the capital. Then delivery is a marginal side-use of an asset that already pays for itself, which means the fallback is the business and the truck is a feature.

The monthly full-house heating variant is worse than previously stated, not marginal: twelve swaps a year gives ~120 lifetime cycles, so amortization is roughly €0.26/kWh and delivered cost lands near €0.31–0.35 against €0.15–0.20 all-in grid. It loses by two to one, and home batteries squeeze it from below by doing the same arbitrage through the existing wire.

Market sizing, Finland

Segment	Size	Annual value/customer	Realistic SAM
Electrified cottages, low consumption, high fixed cost	495k registered cottages; est. 50–100k in viable density clusters	€600–1,200	€30–120M/yr
New rural builds avoiding connection	~2–4k/yr in expensive-connection zones	€800–1,500	€2–6M/yr
Remaining oil-heated houses (full replacement)	shrinking; register data materially overstates active oil heating	n/a	excluded — heat pumps win
Construction / events / grid-queue temporary power	existing market, diesel-priced	€0.30–0.60/kWh	adjacent, already contested

The corrected cost stack cuts this table down further. A 2,000 kWh winter at €0.68–0.72 delivered cost is roughly €1,400 before margin, which is above the €600–1,200 annual value the electrified-cottage segment was sized at — that segment only exists for an operator whose packs earn at the depot year-round. The segments that survive at seasonal-only utilization are the ones already paying diesel prices: no wire, or a €15–45k connection for trivial consumption.

SOM for a pilot operator: hundreds of customers, low single-digit €M revenue, one depot, one truck route geometry. This is a lifestyle-business ceiling unless the model exports beyond Finland’s cottage belt.

Plan, 2026–2030

Phase 0, 2026 — paper. Vero advance ruling on sähkövero treatment of swapped packs. ADR assessment: lithium packs are UN 3480 Class 9 dangerous goods; routed multi-tonne residential delivery of used packs is a real compliance and insurance problem, possibly the plan’s hardest practical constraint. DSO tariff-reform watch. One spreadsheet model with contact-tested logistics quotes.

Phase 1, 2027 — pilot. One depot behind the meter at a wind park with curtailment history. 10–20 cottage customers in one route cluster. Second-life packs from the first meaningful EU supply. Success metric: delivered cost under diesel parity (~€0.50/kWh) with real trucks and real winters, zero thermal events, and measured depot dwell revenue — the model only closes if the packs earn between swaps.

Phase 2, 2028 — product split. Seasonal off-grid service plus the hybrid product: fuse-size arbitrage for grid-connected customers, where a swapped pack caps peak draw and the customer downsizes their main fuse against rising tehomaksu. Reserve-market stacking (FCR-D) on depot dwell capacity.

Phase 3, 2029–2030 — scale or fold. Decision gates: second-life pack price under €30/kWh at volume; charging spread surviving; depot dwell revenue actually covering most of pack amortization; no Energiavirasto bypass-fee response. Any gate fails, fold the operating business and keep the depot as a stationary storage asset — the fallback is itself a viable business, which is the plan’s main risk hedge.

Risk register

The fatal ones first. ADR and fire liability: transporting and residentially siting used lithium packs at scale may simply be unpermittable at acceptable insurance cost; this kills the plan before economics matter. Spread erosion: the depot’s charging advantage is being eaten by stationary flexibility; the 2024→2025 negative-hour decline is the market announcing it. Regulatory reflex: the margin is a tariff-structure short position; the moat is measured in legislative sessions. Home battery cannibalization: at $50/kWh packs, every customer is one wall-box purchase from self-supply, except where there is no wire at all — which is why the only defensible segment is the one without the wire.

The actual conclusions

The plan’s real value is what falls out of it when you remove the company.

The EV is the truck. A cottage-owning household driving a 75–100 kWh EV to the mökki every weekend already performs this exact logistics, for free, with hardware they own. The price of joining depends on how much of the house the car has to feed: a V2L adapter at €300–500 runs appliances and small loads directly from the vehicle, while a complete V2H installation — bidirectional charger, installation, grid paperwork — costs several thousand euros. Either undercuts a delivery contract priced at €0.68/kWh and up. The family car is the delivery fleet, the home wallbox charged on night spot prices is the depot, and the delivered-energy company is out-competed by its customers’ own vehicles — this is the strongest single finding, and the corrected arithmetic only widens its margin.

Off-grid design point moved. At second-life pack prices, the rational new-build mökki skips the €25k connection: solar covers April–September, and a 20–40 kWh pack plus EV top-ups covers the rest. A commercial winter-swap service can still fill that gap, but at the corrected cost it is priced like diesel — a convenience purchase, not the cheap default the original table implied. The connection-fee comparison still loses in a growing share of rural cases, because the owned pack and the owned car do the work.

Home batteries are the same trade through the wire. Everything the depot does — buy curtailment hours, sell against peak tariffs and fixed fees — a wall pack does without trucks, wherever a wire exists. The delivered-energy model is a proof that the arbitrage exists and a demonstration that stationary capture of it dominates mobile capture.

Delivered energy survives only where delivery is the product and the alternative is priced like diesel. No wire, no road to a wire, or no time to wait for one: construction sites, events, grid-queue-delayed industry, and the deepest end of the cottage belt. At €0.68–0.72 delivered, the service does not compete with a grid at €0.30–0.45; it competes with a generator at €0.50–1.00. Everywhere else, the 1960s oil-truck nostalgia loses to the boring fact that the pipe was already built.

Idle capital was the flaw hiding in the arithmetic. The tenfold error was not a typo; it was the model silently assuming the pack’s whole cycle book was available to the delivery business when the seasonal service only ever uses fifty of them. Any swap business built on capital-intensive carriers has to answer where the asset earns when it is not being delivered — and once the depot’s stationary revenue has to carry the pack, the stationary business is the company.

Sources and model boundary

The external anchors are deliberately few:

BloombergNEF’s 2025 battery survey for new stationary-pack pricing
Finnish Energy’s Electricity Year 2025 for negative-price hours
Statistics Finland’s 2025 overview for 495,145 registered free-time residences in 2024
Statistics Finland’s building-stock documentation for the warning that register data overstates active oil heating

Everything else in the cost stack and market sizing is an order-of-magnitude model assumption: second-life pack price and life, route density, depot cost, eligible-customer share, connection cost, and future tariff response. None is a supplier quote, regulatory ruling, or investment recommendation. The analysis is useful only if those inputs remain visible enough to replace.

The idea survived a long attempt to kill it in one room of the house, then lost a factor of ten to an arithmetic audit and kept only that room. It has no electricity, which is the point.