Building a New Digester? Here’s What You Can’t Fix Later
Alchemyca Signals | Issue #3 Published July 28, 2026 | By Alchemyca Biotech
A new digester gets designed around a series of numbers, and those numbers lock in a specific order: tank volume, hydraulic retention time, upgrader capacity, and most importantly, an output target. Every one of those numbers depends on a single input that, on most new build projects, has never actually been measured. What will the feedstock really produce?
The Assumption Everyone Builds Around
On a recent new build dairy manure RNG project in the United States, the developer’s model called for a specific daily gas output target. The engineering team’s own internal assessment, built on standard industry benchmarks, put the realistically achievable output 20 to 30 percent below that target. The design assumed optimistic values for volatile solids in the incoming manure, and did not account for how much that dilution shifts with farm to farm handling practices.
That 20 to 30 percent gap was not a modeling error. It was a data gap. Neither the developer’s target nor the engineering team’s benchmark was built on the actual feedstock, just on assumptions standing in for it.
The Upside Case, and the Objection to It
Characterizing feedstock precisely is usually framed as downside protection: get it wrong, and the shortfall costs millions in unmet output over the life of the loan. The upside case matters just as much. A system designed against the real energy content of the feedstock, rather than an optimistic assumption, gets sized to a target it can actually hit, and in some cases exceed, instead of chasing a number that was never achievable to begin with.
There is an obvious objection to this: what if the real number comes in lower than what the project needs to clear financing? A developer’s return case often depends on hitting a specific production threshold, and precise characterization can come back below it. That is not an argument for skipping the characterization. A project built around a number the feedstock cannot support is not more fundable for having avoided the data. It just fails later, after the money is spent, instead of at the term sheet.
Who Actually Owns the Gap
The engineering firm’s contract typically closes out at commissioning. The obligation to hit that production number, the one tied to revenue, loan covenants, and investor returns, sits with the owner for the next fifteen to twenty years. When output lands short, it’s the owner explaining it to the offtaker and the lender.
That is not a new problem. John Haeckel, Founder of Clean Fuel Partners, who spent years turning around a distressed anaerobic digestion facility before selling it, put it in blunt terms:
“The designs may be great, but they make the assumption that nothing’s ever going to go wrong. Something always goes wrong. It’s more expensive to design and build it to be maintainable, but it’s literally: you’re going to pay now, or you’re going to pay later.”
What Locks In, and When
Tank volume and hydraulic retention time get set at the design stage, before groundbreaking. Upgrader capacity gets purchased on a schedule that has to keep pace with construction. The output guarantee gets written into contracts before either of those decisions is finalized. In practice, all three commitments are typically made before anyone has run a single biomethane potential test on the actual feedstock, let alone anything closer to how the plant will really operate.
The first real feedstock data, when it comes at all, tends to arrive during commissioning or in the months after start up. By then, tank volume is concrete, upgrader capacity is fabricated, and the output guarantee is signed. The operator is the one left making the numbers work.
The Retrofit Math
Characterizing feedstock before design finalizes is a matter of several weeks and a modest testing budget. Correcting a feedstock gap after commissioning is a different order of problem. An undersized upgrader means a capital project to add capacity. A missed output guarantee means a contract dispute, a cash settlement, a hard conversation with a lender, or all three. A digester sized for the wrong hydraulic retention time cannot be resized without a construction project inside a live facility, and every month of that fix comes directly out of an operator’s cash flow.
Getting in front of the design numbers
The fix is not more conservative engineering assumptions, and it is not simply running a bench scale biomethane potential test and calling the feedstock characterized. A BMP test runs in triplicate, in optimal, controlled conditions, on a small batch of feedstock in a bottle. Not exactly at scale conditions. It’s useful, but not a number a continuous, fully loaded plant will actually reproduce.
Between the bench scale test and full scale operation sits a step most new build projects skip: a small continuous pilot, on the order of a few thousand dollars, that runs the real feedstock through a scaled down version of the intended process, continuously, under something closer to the loading pattern and retention time the full scale plant will see. That is the step that shows whether the biology holds up outside a bottle, not just whether it can produce gas once, under ideal conditions.
A bench scale biomethane potential test against the actual feedstock, a small continuous pilot that models how the scaled up plant will actually run, and continuous NIR based monitoring deployed ahead of commissioning, together give an owner a defensible output range to design and contract around, built from how the system will really operate rather than a single batch result run under optimal conditions.
The operators who insist on this are not doing it because the technology is new. They are doing it because they, or someone they know, has already lived through the alternative: a signed guarantee built on an assumption nobody tested, and the fallout, which unfortunately can include asset shutdown and lost jobs.
What to Take From This
The cheapest time to close a feedstock gap is before it becomes the design basis. Every number that gets locked in before groundbreaking is a bet on a feedstock assumption, and the owner is the one still holding that bet long after everyone else has moved on to the next project. The bet gets a lot safer, and the upside gets a lot more real, once it is backed by real data instead of a benchmark.
The next issue of Alchemyca Signals covers why bioaugmentation only works on biology that is already stable, and why biocatalysts are only as good as the biology they are working with.
Subscribe to Alchemyca Signals at https://www.alchemycabiotech.com/signals. One insight, every Tuesday.
Alchemyca Biotech develops the Claritix™ operational intelligence platform for anaerobic digestion facilities.
