BESS financing comparison

BESS financing · sixty-second read

Will this floor deal
actually clear?

Merchant vs tolling vs insured revenue floor. Same battery, three doors, per MW. Figures in €k, illustrative. For grid-connected merchant assets.

How this works, in 2 minutes

The problem. A battery earns money, but nobody knows exactly how much each year. Banks hate that, so they lend little and charge a lot.
Door A, merchant: keep everything, accept a bad loan.
Door B, tolling: a trader guarantees fixed payments and takes over the asset. Good loan, but the trader keeps 25–30% of the value forever.
Door C, the floor: a writer guarantees only a minimum. Owner keeps everything above it, pays a premium plus a slice of good years. Good loan, owner stays in control.
The bank pairs: fill BOTH. Never set one pair to zero. Door A always uses the uncertain-revenue pair; doors B and C always use the secured-revenue pair. The tool runs all three doors at once, side by side.
Revenue basis, important: revenue here is AFTER any trading commission. Site costs (rent, O&M, insurance) are opex. Never count the trading commission twice.
Who provides which number: PROJECT project data and bank quotes. FLOOR WRITER the price of the guarantee. MARKET benchmarks.
The purple box shows the viable zone, recomputed live: the lowest floor the bank accepts, and the point beyond which tolling becomes the better deal.

01The project

The battery itself. Enter the project's real numbers.

Expected revenue per MW per year€120kWhat the battery earns from trading, after any trading commission. Use the post-commission figure and never count the commission again.

Capex per MW€440kWhat it costs to build, turnkey. A heavy grid connection can push this higher.

Opex per MW per year€35kSite costs: rent (usually the biggest item), O&M, technical insurance. Excludes any trading commission, which is already off the revenue.

02The floor deal (door C only)

The floor level is the product itself. The writer quotes the premium and upside share.

FLOOR LEVELFloor level, % of expected revenue67% = €80.4kThe height of the safety net, as % of expected revenue. The bank needs it high enough to carry the loan (the purple box shows the minimum). The writer charges more the higher it goes. Each extra hour of duration tends to raise the achievable floor.

FLOOR WRITERFloor writer: premium€6kThe writer's yearly charge for the guarantee. Their quote fills this in.

FLOOR WRITERFloor writer: upside share20%The writer's bonus in good years: a slice of everything above the floor. Premium and share are a seesaw; they quote the package of both.

03Bank terms

Two loan worlds, both always filled, never zero one out. Door A uses the uncertain-revenue pair only. Doors B and C use the secured-revenue pair only.

Uncertain-revenue bank: leverage40%How much of the build cost a bank lends when revenue is uncertain.

Uncertain-revenue bank: interest rate, 7y9.5%Higher when revenue is uncertain, and the bank wants its money back quickly.

Secured-revenue bank: leverage60%At this leverage the bank needs a floor of at least €78.8k = 65.7% of revenueHow much they lend when revenue is guaranteed. The purple box shows the floor this leverage requires.

Secured-revenue bank: interest rate, 10y5.5%Lower where revenue is secured, with a longer tenor available.

04The alternative (door B only)

What door B — a pure tolling contract — actually costs. This is what door C has to beat.

MARKET BENCHMARKTolling take, % of revenue28%What the toller keeps for its fixed payments. Market experience: 25–30%. This is the price door C must beat.

05Does a project qualify for a floor?

Floor-readiness checklist. Every yes makes the project easier to underwrite and prices better.

0 of 8 checked

Real and owned.The developer holds the project, not "hoping to acquire".
Specifics exist.Location, MW, MWh duration, target COD, EPC, dispatcher.
Duration 2h or more.Each extra hour raises the achievable floor.
COD inside 12–18 months.Writers sign up to around 12 months pre-COD.
Adds geographic spread, diversification prices better.
Revenue stated on a post-commission basis, with its source.The forecast source matters to the bank underwriting the loan.
Site costs known, rent especially.Rent is the silent killer of floor economics.
Developer accepts a standard floor contract structure.Standard terms, faster underwriting, cleaner pricing.

Verdict, live

The floor clears.

20.5% ROE▲ 1.8pp vs merchant

20.5% ROE vs 18.7% merchant and 9.3% tolled. Equity per MW: €176k vs €264k — same money builds ~1.5× more MW.

Bank check: at the floor, cash covers loan payments 1.3 times. Banks want 1.25. The loan works.

Owner check: certainty via the floor costs €13.9k per MW-year. Tolling would cost €33.6k. The floor is the cheaper certainty, saving €19.7k per MW every year.

Same battery, three doors

Return on equity, per year

18.7%

9.3%

20.5%

AMerchant

BTolling

CFloor

What's going on

The floor route wins. Owner earns 20.5% on equity, versus 18.7% merchant and 9.3% tolled. Equity needed per MW: €176k with the floor, versus €264k merchant, so the same money builds about 1.5 times more MW. And the owner keeps all upside above the floor minus the 20% share.

The viable zone, live

The bank needs the floor at €78.8k or higher (65.7% of revenue) to lend 60% of capex.
The floor's cost (premium + upside share) stays worthwhile up to €17k per MW-year (staying merchant is the tougher rival here; vs tolling €33.6k, vs merchant €17k). Beyond that, the alternative is the better deal.

Where door C's revenue goes, per MW per year

Revenue in (after trading commission)	+120
Site costs: rent, O&M, insurance	−35
Floor writer: premium + upside share	−13.9
Bank: debt service	−35
Owner keeps	36.1

Comparison detail

	A · Merchant	B · Toll	C · Floor
Owner equity €k/MW	264	176	176
Debt service €k/yr	35.6	35	35
Cost of certainty €k/yr	0	33.6	13.9
Net cash to equity €k/yr	49.4	16.4	36.1

Method, assumptions and sources

Revenue basis: post-trading-commission. Opex: site costs (rent, O&M, insurance), excluding any trading commission. Bank test: (floor − opex) must cover annual debt service 1.25 times. Owner test: the floor's cost of certainty (premium + expected upside share) must be lower than tolling's take. Upside-share cost approximated as share × (expected revenue − floor). Tolling assumed to reach the same leverage as the floored case. Merchant case ignores cash-sweep clauses. Scope: grid-connected merchant assets only.

All figures are illustrative scenario inputs, not quotes or offers. Enter your own project's numbers to see its result.