Underwriting a captive solar+BESS asset for a financier

A financier evaluating debt against a captive solar+BESS project was handed the usual artefact: a developer's pro-forma showing healthy coverage ratios across the loan tenor. The question a credit committee actually has to answer is different — not whether the project works if the assumptions hold, but what happens to debt service when they don't.

For a captive renewable asset, the assumptions that matter are specific: how much the plant generates in a poor year, how much of that generation is curtailed or lost to banking rules, what the avoided grid tariff is actually worth block by block, and how the state's regulatory settings move over a loan tenor measured in decades. A lender that cannot see those risks quantified either prices them with a blunt haircut — or walks away.

The standard pro-forma is built on annual arithmetic: expected annual generation, multiplied by an average avoided tariff, escalated at a fixed rate, degraded by a flat percentage. Every one of those averages conceals the thing a lender needs to see.

• The tariff is not one number: a captive project's savings depend on which 15-minute blocks its energy displaces. Midday solar displaces the cheapest grid power; the battery's evening discharge displaces the most expensive. An average tariff blends the two and mis-states both
• Generation is a distribution, not a point: a single expected-generation number says nothing about the spread between a typical year and a bad one — which is precisely the spread debt service has to survive
• Surplus is not savings: energy the load cannot absorb in the block it arrives is banked at a haircut, compensated at a low rate, or lost, depending on the state's rules. Annual netting silently counts it at full value
• The rules move: banking windows, time-of-day slabs, and settlement provisions are amended repeatedly over a 25-year asset life. A pro-forma frozen at today's rules carries regulatory risk it never prices

None of these is a modelling nicety. Each one moves cash flow in the direction a lender cares about — downward — and each is invisible at annual resolution.

The Asset at 15-Minute Resolution

EarthSync's planning engine simulated the project the way the market settles it: generation, the consumer's load, and the battery's dispatch at 15-minute resolution across the full year, evaluated against the applicable market rules — time-of-day tariff slabs, banking provisions, open access charges, and deviation settlement. The battery's charge and discharge in every block is decided by the optimiser against those rules, not assumed as a fixed daily cycle.

Curtailment was treated as what it actually is: a physical and regulatory constraint, not a percentage plucked from precedent. Physically, generation beyond what the load and the battery can absorb in a given block has nowhere to go. Contractually and by regulation, the treatment of that surplus — banked under a shrinking window, compensated at an administratively set rate, or lapsed — is defined by the state's rules, and the simulation applies those rules block by block. The result is a curtailment and surplus-loss profile derived from the asset's own shape against its own market, rather than asserted.

Uncertainty, Propagated Honestly

On top of the block-level simulation, Monte Carlo analysis propagated uncertainty through the variables a committee will probe: generation between P50 and P90, capex, tariff escalation, battery degradation, and regulatory drift in banking and settlement. The outputs arrive as ranges with sensitivity bands, not single-point estimates — every headline metric carries its P50/P90 spread, and the tornado view shows which assumption moves the answer most.

The financial characterisation was built the way a lender reads it: 15-minute cash flows rolled up to annual statements across the asset life, surfacing LCoE, NPV of savings against capex and opex, equity IRR, payback, and — most importantly for debt — DSCR year by year, under the base case and under stress. The committee could ask the questions committees ask: what does coverage look like in a P90 generation year with tightened banking? Which single assumption, moved adversely, breaks the covenant first? The model had answers because the uncertainty was in the model, not appended to it.

• A defensible base case: every cash-flow line traceable to block-level simulation under the market's actual rules, rather than to an averaged assumption — a base case that survives being taken apart
• Quantified downside: DSCR and returns expressed as P50/P90 ranges under regulation-aware stress scenarios, so the committee prices the risk it can see instead of discounting for the risk it can't
• Curtailment made legible: surplus energy, banking losses, and curtailment exposure derived from the asset's shape and the state's rules — a specific number per scenario, not a generic haircut
• A faster diligence conversation: when every sensitivity a committee raises has already been run, diligence shifts from contesting assumptions to examining results — and the questions converge instead of multiplying

Underwriting a captive renewable asset is not a generation estimate with a discount rate attached. It is a claim about cash flows that settle 15 minutes at a time, under rules that will be amended, in years that will not all be average. Modelling the asset at that resolution — with uncertainty propagated rather than assumed away — is what turns a developer's pitch into a credit decision. The pro-forma tells a lender what the project hopes to earn; the simulation tells it what the project can survive.