How Tradvolt builds a cost model

Every study on this site follows one discipline: separate what a supplier quoted, what a public source states, what is a scenario assumption, and what is a calculated result. Where an input is not available, it is written down as an open item for the buyer to confirm — it is not filled with an invented number.

This methodology hub gives engineering, finance, and procurement teams the exact accounting rules behind every Tradvolt landed-cost and total-cost-of-ownership (TCO) study so you can audit our numbers or stress-test a live supplier quote. Across all studies, Tradvolt computes present value of costs, net present value (NPV), internal rate of return (IRR), simple and discounted payback, Levelised Cost of Energy (LCOE), Levelised Cost of Storage (LCOS), and EV-vs-ICE TCO per kilometre from a single year-by-year cash-flow architecture. To prevent hidden optimism, every model enforces a strict four-way boundary between supplier-quoted figures, cited public facts (OEM datasheets, IEC/UL standards, WCO HS headings, IRENA/BNEF benchmarks), explicit bounded scenario defaults (specific yield, duty cycle, port utilisation, WACC, tariff escalation), and unresolved customs duty or VAT rates flagged as PENDING — lookup. Use the four asset-class templates below—Landed Cost, Battery & Storage TCO, Solar PV TCO, and EV Charging TCO—together with our verification protocol to decide which cost lines belong in your tender evaluation and which open inputs must be confirmed before contract award.

The shared principle

The overall approach, including how evidence dates and exclusions are handled, is described on how we model landed cost and total cost of ownership. The pages below apply that approach to specific equipment classes.

Across all equipment classes, the computation engine in tco_model.py builds a year-by-year cash-flow table from Year 0 through the stated analysis horizon (validated between 1 and 60 years) on a single consistent basis—either nominal/undiscounted or real/discounted, never mixed. Capital outlays enter at Year 0 or their scheduled installation year, fixed annual operating costs start in Year 1 and compound at the stated annual escalation percentage, scheduled component replacements enter in their designated year, and any end-of-horizon residual value is recorded as a negative cost in the terminal year. Each year's net flow (benefit - cost) and delivered energy are discounted by 1 / (1 + discount_rate)^year so that net present value (NPV), bisection internal rate of return (IRR), cumulative-sum simple and discounted payback, and levelised energy cost (LCOE or LCOS) are mathematically consistent.

Method pages by equipment class

What is deliberately excluded

Models do not monetize benefits that the buyer has not quantified — resilience, avoided outage cost, carbon value or brand effects — unless a study says so explicitly and states the basis. They do not assume a certification, a warranty term or a lead time that no document in the study establishes.

Baseline scenario calculations also omit unquoted financing fees, foreign-exchange hedging costs, peak demand-charge reductions on energy-only storage runs, ancillary-service or capacity-market payments, and national customs duties that depend on binding tariff rulings. Where a project's payback depends on those excluded streams, the study publishes the negative or conditional baseline result transparently and identifies which input levers must change before procurement.

How to choose which model applies

Start from the question you are being asked to answer. If it is what the goods cost to have delivered, use landed cost and stop there. If it is what the equipment will cost over its life, add the class that matches the equipment — storage, solar or charging — and note that a hybrid site needs more than one.

Before relying on any single point estimate, test the three primary sensitivity levers defined for each asset class in the model engine—discount rate plus tariff and specific yield for solar PV, discharged-energy value and charging cost for battery storage, or diesel price per litre and electricity tariff for EV fleets—under plus-or-minus 20% shocks to see how NPV and payback respond.

All four share the same evidence discipline described on how we model landed cost and total cost of ownership: what is quoted, what is public, what is assumed and what is calculated are kept separate, and an unavailable input is recorded as open rather than filled in.

Start here: flagship studies & method notes

Browse by buyer decision

1. Building the Year-0 landed-cost stack (FOB, freight, HS duty & VAT)

Decompose the charges between a Chinese factory gate and site delivery, and see how unresolved tariff headings are flagged for broker confirmation.

2. Modelling stationary battery LCOS, fade & augmentation reserves

Audit how daily cycling duty, round-trip efficiency, annual capacity fade, and mid-life augmentation reserves govern 10- to 15-year storage economics.

3. Computing 25-year solar PV LCOE, yield & inverter replacement

Examine how location-specific yield, linear module degradation, tariff escalation, and Year-12 inverter replacement reserves drive PV payback.

4. Auditing EV charging utilisation, grid make-ready & claim verification

Evaluate port utilisation, transformer and civil works CapEx, EV-vs-ICE fleet TCO per km, and how sourced facts are separated from assumptions.

Frequently asked questions about Tradvolt cost modelling

1. How does Tradvolt discount costs, benefits, and energy throughput in its LCOE and LCOS formulas?

In every computed Tradvolt study, both annual cash flows (costs and benefits) and annual energy throughput (kWh generated or discharged after annual degradation) are discounted at the stated project discount rate (WACC). LCOE and LCOS are calculated as the present value of discounted lifecycle costs divided by the present value of discounted energy throughput over the 10- to 25-year horizon, keeping the time basis strictly aligned with Net Present Value (NPV).

2. What items are deliberately excluded from Tradvolt's baseline TCO models?

Unless a study explicitly states a monetised basis, baseline models exclude unquantified co-benefits such as avoided outage or SLA downtime penalties, resilience value, demand-charge savings on energy-only runs, ancillary services revenue, carbon credits, FX hedging costs, and project financing fees. Similarly, models never assume customs duty exemptions, product certifications, or warranty terms that are not established by a cited document.

3. How does the battery TCO model avoid double-counting charging electricity cost?

In stationary storage studies (such as the 10 MWh Southeast Asia data center BESS and 3 MWh MEA hospital ESS models), the cost of grid or solar electricity drawn to charge the battery—adjusted for DC/AC round-trip efficiency (typically 88%–92% for LFP)—is already included inside the annual Cost column of the year-by-year cash-flow table. Therefore, charging electricity is not added a second time when computing LCOS from the present value of costs.

4. Why does the solar TCO model include a mid-life cash-flow spike between Year 10 and Year 15?

While crystalline silicon PV modules are typically modelled over a 25-year operating horizon with 0.4%–0.55% annual linear degradation, three-phase string inverters carry shorter standard warranty envelopes. Tradvolt's 25-year solar PV models (such as the 2 MWp Indian convention center rooftop study) book an explicit inverter replacement or overhaul reserve—for example, 10% of initial CapEx in Year 12—so lifecycle LCOE and NPV reflect realistic power-electronics replacement.