Methodology

The model walks a plant from feedstock to finished fuel. At each stage it computes a small set of quantities, and each of those is argued against a published authority — or, where there is not one yet, is declared as our own banded estimate.

The bodies below are the reference set the model is argued against, with what each one actually backs. None of them has reviewed or endorsed cbg-sim; naming them is a statement about our reading, not about their approval.

The model has no way to express a bare number. Every quantity carries a status, and arithmetic takes the weakest input — so a placeholder cannot launder itself into a confident headline, and a computed figure is never sourced however good its ingredients were.

1. sourcedSomebody published this exact figure. The citation travels with it.
2. derivedComputed from other values by a stated formula.
3. pending verificationMeasurement is scheduled or in flight.
4. projectionA forward assumption. Defensible, not measured.
5. assumedA placeholder. The widest band, and the first thing to replace.
6. mandate ≠ capacityA policy target. Never quotable as installed capacity.

The discipline is borrowed. Petroleum reserves have been reported as proved, probable or possible for decades precisely because an uncalibrated model quoted to a lender is a liability. This is that habit applied to biogas.

The model reads 164 parameters. Five are sourced — four physical constants and the CEA grid emission factor. 26 are derived, 9 are placeholders, 4 have measurement in flight, and 120 are projections: defensible estimates inside declared bands, not published measurements. The seven feedstock biomethane-potential values are all in that last group, and each carries a note naming the study still needed for it.

That is stated before anything flattering because a consultant checks biomethane potential first and should find the answer rather than catch us out. Sourcing those projections is active work, and this paragraph is written to shrink.

The knowledge base behind the tool is further along. Of 258 reviewed answers in Learn, 192 are sourced and 218 carry a citation a reader can look up, with 49 derived and the remainder flagged as projections — or as policy mandates, which are never quotable as installed capacity.

The capex build-up is fitted against three Indian project cost points — ₹14 crore at 2 TPD CBG, ₹65 crore at 10 TPD and ₹140 crore at 25 TPD — declared in the model and re-checked by the test suite on every run. Residuals at all three sit inside their stated bands.

The smallest anchor is a demonstration plant whose reactor loading is past the acidification threshold at its stated volume, so it anchors cost and not process; the suite says so rather than quietly fitting through it. Three calibration items remain open and each needs named-operator review: empirical conversion efficiency for Indian Napier, the soil-carbon credit, and the biochar credit. All three are bracketed conservatively.

• No dynamic time-step simulation — steady state only
• No microbial population dynamics or detailed kinetics
• No explicit ammonia, VFA or H₂S inhibition
• No feedstock seasonality; year-round availability at design rate
• No stochastic Monte Carlo — the run is deterministic
• No financing detail: no debt, tax or working-capital cycle
• No CFD or hydraulics inside the digester

A fuller specification — notation, equations, parameter tables, calibration and sensitivity, with a reviewer cover note — is available on request.