Skip to main content

The Problems We Still Need to Solve in Gluten-Free Brewing

This is the frontier. Most of what we publish as guidance is a way of working around something the field has not solved — a malt that cannot convert on its own, a grain with no specification, a hold sized to cover variation nobody has measured. Here is that list, in the open, with what we know, what we don't, and the work that would move each one.

A knowledge base tells you what is known. This page tells you what isn't. Every problem below is drawn from a position this site already takes — the "why we take this position" record on the owning page names the problem that forces it — or from an open Research Ask, an intake record, a directory research-gap section, or a contradiction found in the evidence library; where we have proposed an explanation for a problem, it is on Open Theories, stated with the test that would falsify it; since 2026-09-06 a new intake record is checked against the Problems it touches at intake time (Runner 02, step 11). Where the honest answer is "nobody has measured it," that is what it says. An empty line is a fact about the record, not an invitation to invent.

Scope. Every gluten-free brewing grain. The register is sorghum-heavy, and the reason is plain: the positions this site takes, and the production record they rest on, are sorghum positions. The other grains' problems enter from the research directories' gap sections, and the Coverage Map shows in one line each where those fields run out. Where a problem is the same for every grain, it is stated once, for all of them.

Problems are ordered within each category by how much a solution would change published guidance, and we say why at the top of each group. The order is ours; the evidence is on the linked pages.

Raw materials​

Ordered by how much a solution would change what a brewer can buy. Supply is the constraint everything else sits behind.

Malting-sorghum seed and malt do not exist at qualifying scale

The problem. There is no identity-preserved malting-sorghum seed supply and no malted-sorghum supply in North America that meets a published truly-gluten-free standard at commercial scale. Why it matters. Every brewer who wants to run a 100% sorghum malt process must either malt their own or source from the single-digit number of producers we know of — none of which, to our knowledge, guarantees cultivar identity lot to lot. Where we are. We track identity lot by lot instead — see Cultivars for Malting — and keep a standing invitation to anyone who will make the malt. What we know / don't know. Known: which named cultivars malted well in the USDA Madison screening, at one location. Not known: whether any of them perform across locations and crop years; whether seed rights for the best lines can be secured. Open contradictions. None. Research asks. Ask 8; Ask 9. Status. Open — blocked on supply.

Sorghum malt data from other producers is not public

The problem. Sorghum malting has been done commercially and experimentally for decades, and almost none of the resulting malt analyses, screening data, or brewing trials are published. The record this site holds is one program's. Why it matters. Every new maltster and brewer starts from near zero, and every failure is repeated. Where we are. We publish everything we hold with DOIs and ask for the rest. What we know / don't know. Known: our own datasets, the published academic screenings, and the South African methods literature. Not known: what sits in the filing cabinets of programs that ended. Open contradictions. None. Research asks. Ask 8. Status. Open — standing.

No grain but sorghum has released malting-quality data, and buckwheat has no study at all

The problem. Millet, rice and corn each have a published cultivar-by-malt-quality study — 24 proso millet varieties under one malting regimen (2010), 19 rice cultivars malted and brewed (2024), and a maize varieties malt-quality study — and none of them releases the row-level tables a brewer could re-analyse, which is what we republish for sorghum in the Madison screening. Buckwheat has no cultivar malting study we have found at all. An earlier version of this problem said none of the four grains had any public malting-quality dataset; the rice study is one our own rice directory already links. Why it matters. A brewer choosing a millet or rice malt has the supplier's description and their own trials, and no shared numbers to compare cultivars, lots or maltsters against. Where we are. We publish the sorghum datasets we hold (Research Library). For the other grains we hold nothing to publish, and each directory now says what the literature holds (Millet, Rice, Buckwheat, Corn). What we know / don't know. Known: germplasm, agronomy and variety-trial data exist for each grain; rice cultivars vary in malting traits, so there is something to select on; maize malts comparably to sorghum at equal nitrogen. Not known: how any named buckwheat cultivar performs as malt; and for the other three, the underlying rows behind the published summaries, or how any of it maps onto the commercial malts US brewers actually buy. Open contradictions. None. Research asks. Ask 8, which covers every gluten-free grain, not sorghum alone. Status. Open — standing.

Malting​

Two problems: the one most of our process guidance is built around, and the grain much of the industry runs on.

Sorghum malt does not supply reliable endogenous conversion enzymes

The problem. Gluten-free brewing malt carries measurable enzyme activity, but not reliably enough, lot to lot, to plan a commercial mash around. Sorghum malt's beta-amylase and total diastatic power run markedly below barley malt's. Why it matters. It forces every brewer onto external enzymes, which cost money, add process steps, and make the malt's own conversion potential irrelevant to how it is designed. Where we are. Malt for character; external enzymes own conversion. The three routes to conversion, and why the koji route is documented and not adopted, are on Where Conversion Power Comes From. That position says plainly that it is a compensation for the malt, not the ideal. What we know / don't know. Known: the spread of diastatic power and alpha-amylase across 25 named cultivars at one location (Madison screening); that some cultivars are markedly better than others. Not known: how much of the variation is genetic versus malting method; whether any cultivar under a standardized malting protocol reaches a level a process could rely on. Open contradictions. Three, none of which moves the doctrine for this protocol: a Wrocław study brewed gluten-free beer from sorghum malts without external enzymes and found the starch saccharified fully under decoction mashing; koji-inoculated sorghum malt out-yielded barley malt by 5%; and the α-glucosidase finding points at a native route the DP number does not see (Sorghum; Koji and Enzymes). Recorded 2026-09-06 (B2). Open theories. The fourth germination day buys alpha-amylase, not diastatic power. Research asks. Ask 9; Ask 6. Status. Open — workaround holding.

Millet malt carries much of US gluten-free craft brewing on the thinnest science in the family

The problem. Commercial millet malt is produced at scale and brewed on today, and its published brewing science is thinner than sorghum's. Production runs ahead of the literature (Millet directory). Why it matters. A brewer on millet works from supplier practice and their own record. When a lot behaves differently there is no published baseline for the commercial malt to say whether it was the malt, the cultivar or the process; the varietal studies that exist were run on research malt under research regimens. Where we are. Millet Malt in Gluten-Free Brewing gives qualitative guidance and a lot-evaluation checklist, and says plainly not to infer extract, diastatic power, FAN or usage rates from it; Sorghum vs Millet Malt compares what can be compared. What we know / don't know. Known: that millet malt works, commercially, at scale. Not known: the numbers behind the commercial malt (the previous problem), and whether millet malt's own enzyme activity is any more reliable lot to lot than sorghum's; we have found no dataset that says. Open contradictions. None. Research asks. None yet. Ask 8 for the data. Status. Open.

Mashing and conversion​

Ordered by how directly a result would change a number in The Reliable Mash.

The 190°F / 60-minute hold may be hotter and longer than conversion needs

The problem. The baseline hold is sized to cover malt-lot, equipment, and scale variation that nobody has measured. A 30-minute hold worked in most production runs. The floor is unknown. Why it matters. Energy, time, and — at production scale — throughput. A brewer who could validate a lighter hold on their own system saves all three on every batch. Where we are. 190°F for 60 minutes, validate anything lighter across several batches and lots. What we know / don't know. Known: the baseline transferred to two outside breweries and was highly reliable in commercial experience. Not known: where temperature or time actually stops working, on any malt. Open contradictions. None. Research asks. Ask 7; Ask 3. Status. Open — workaround holding.

The external-enzyme dose floor is unknown

The problem. Published doses are conservative production insurance. Nobody has published the curve for this malt and this system showing where conversion, runoff, or fermentation breaks as dose comes down; two curves exist for other substrates (see contradictions). Why it matters. Enzymes are a real line item. If half the dose works, that is money back for every gluten-free brewer. Where we are. The documented protocol doses, stated as a starting point. What we know / don't know. Known: the doses that worked in production. Not known: the minimum. Open contradictions. Two published dose curves exist, so "nobody has published the curve" is true only of this malt and this system: a response-surface optimisation of commercial enzyme doses on buckwheat malt, and Desobgo's response-surface modelling on unmalted sorghum (External Enzymes; Buckwheat). The floor for this protocol is still unmeasured. Recorded 2026-09-06 (B2). Research asks. Ask 6. Status. Open.

the Maltose Mash, the maltose mash, was never tested against the Reliable Mash — and the reason the Reliable Mash won was never recorded

The problem. A protease rest plus a fungal-alpha coolback was written up and played with at Bard's; the glucoamylase finish won in production, and nobody wrote down why. Everything downstream of wort sugar profile — fermentation behavior, body, drinkability — rides on this. Why it matters. If the Maltose Mash makes a materially better beer from sorghum malt, the production recommendation is wrong. If it doesn't, the field should know why. Where we are. The Maltose Mash is published as proposed, untested, and a revival. What we know / don't know. Known: the proposed method and the reasoning. Not known: the side-by-side result. Open contradictions. None — this is an untested alternative, not conflicting evidence. Research asks. Ask 4 — the first problem this site is working by hand, with outside testers. Status. Under investigation.

The calcium baseline's original reason — recovered September 2026; what the process needs is still unmeasured

The problem, as it stood. Production ran a complete-process calcium condition of about 100 ppm, with period enzyme-distributor guidance behind it. The enzyme itself is documented to work below 20 ppm. Why production used the higher figure was not in the record. The problem, answered. It was in the record, in the November 2019 correspondence: the 100–150 ppm advice was given for Termamyl Classic, which needs 50–150 ppm; the protocol switched to SC DS two days later — an enzyme with no calcium dependency — and the note carried over unrevised. Craig confirmed on 2026-09-04 that it carried over and was kept as production practice. That answers why. It does not answer whether the complete process needs it. Why it still matters. A brewer targeting 100 ppm may be adding salts — and the chloride or sulfate that comes with them — for a reason that belonged to a different enzyme. Whether the malt enzymes, the mash pH, runoff or the beer benefit from the calcium anyway is unmeasured. Where we are. Target about 100 ppm as the complete-process baseline; do not read it as an enzyme minimum — the enzyme needs none. What we know / don't know. Known: what production ran, what the enzyme maker states, and now why the two differ. Not known: what the complete process actually requires. Open contradictions. None — the apparent contradiction was a carried-over setting, now stated as history on the owning page. Research asks. Ask 10 — the calcium ladder. Status. Open — evidence request standing.

Protein and fermentation​

One recorded problem. This category is thinner than the field deserves.

Sorghum malt FAN runs about half of barley malt's

The problem. Free amino nitrogen across the sorghum malts we screened ran roughly half the level a barley-malt wort gives the yeast. Fermentation support that barley brewers get for free has to be designed in. Why it matters. Sluggish or incomplete fermentation, off-flavors, and yeast health all trace back here. Where we are. Treat yeast nutrition as a process goal, not an afterthought. What we know / don't know. Known: the FAN spread across 25 cultivars and the published literature agreeing on the pattern. Not known: how much protease, malting modification, or cultivar choice can close the gap, and at what cost to foam and body. Open contradictions. Two, pointing opposite ways: koji-inoculated sorghum malt gave 24% more FAN after double mashing, so the gap is not fixed by the grain (Koji and Enzymes); and kafirin cross-links further under exactly the wet heat a mash applies, so the mash itself widens it (Kafirin and Protein Digestibility). Recorded 2026-09-06 (B2). Research asks. Ask 11 — closing the FAN gap, one variable at a time. Status. Open.

Process​

The rice-hull figures actually run in production were never reconciled

The problem. Production ran about 4.8% rice hulls and lautered clean; the mature sorghum operations we know run under 2%, some none. The figures actually run at one outside brewery — 2% versus 5% — and whether a true 0% trial ever occurred, are not firmly reconciled in the record. Why it matters. Hulls are cost and bed volume. A brewer who can go lower with a coarse grind should; one who can't needs to know where the floor is. Where we are. Start around 5% when bed behavior is uncertain; experience moves the number down. What we know / don't know. Known: the direction. Not known: the floor on a given malt and system. Open contradictions. None. Research asks. Ask 12 — the split batch that finds the floor. Status. Open — evidence request standing.

Sensory​

Hop performance against sorghum malt is unmapped

The problem. Hop variety guidance is barley guidance. Which varieties and rates work with sorghum malt's character, and which fight it, has never been systematically compared. Why it matters. Flavor is what sells the beer, and the usual rules of thumb were written for a different grain. Where we are. Hops and yeast for sorghum malt gives the working guidance; the 5-IBU hop-extract rule covers foam. What we know / don't know. Known: what worked in production. Not known: the comparison. Open contradictions. None. Research asks. Ask 1; Ask 2. Status. Open — seeking testers.

Quality, safety, and testing​

Ordered safety first. Nothing in this group may be worked around below the standard: no wheat, no barley, no rye, no oats, ever.

Verifying a gluten-free claim in a shared facility

The problem. A brewery that also handles gluten grain can produce a beer with no gluten grain in it — and still cannot easily prove, batch by batch, that cross-contact did not occur. The evidence chain that would prove it is expensive and not standardized. Why it matters. The claim is a safety claim to a celiac reader. Getting it wrong harms people. Where we are. Work the question in order and prevent cross-contact upstream. Dedicated equipment is the control; testing is the check. What we know / don't know. Known: what dedicated production proved, batch by batch, over four years (Production QA Log). Not known: a validated, affordable evidence chain for a shared facility. We do not publish a workaround here. Open contradictions. The FDA's rule of 13 August 2020 already replaced finished-product assay with a records route for fermented and hydrolysed foods, and names sorghum beer as its example; "no validated method" is the regulator's stated basis for that route, not an open gap (Testing and Certification). Recorded 2026-09-06 (B2). Research asks. None yet. Status. Open — no workaround published.

No shared method for evaluating sorghum malt

The problem. Commercial labs run the barley panel or the feed panel on sorghum malt, because that is what they have. Neither measures quite the thing. Suppliers and brewers cannot compare lots. Why it matters. Without a common method, a malt specification cannot exist, and without a specification, every lot is a test. Where we are. A proposed Sorghum Malt Evaluation Standard, draft, open for criticism, and a lot-reporting standard for what a lot must declare. What we know / don't know. Known: what each existing method measures and misses (Evaluation Methods). Not known: whether the proposed standard holds up when someone other than us runs it. Open contradictions. The method problem is already producing contradictory numbers in our own record: Muoria's 4.37 °ASBC and Madison's median 31.2 °ASBC are the same unit on the same grain, both cited on this site, and Hartwick's total-protein method is undeclared (Evaluation Methods; Madison screening). Recorded 2026-09-06 (B2). Research asks. Ask 5. Status. Open — seeking review.

Per-lot reporting is missing for every gluten-free malt, not only sorghum

The problem. The sorghum, millet, rice, buckwheat and corn directories each record the same finding: no widely used per-lot reporting practice for malt buyers. The numbers that predict mash behavior, protein, germination, moisture and mycotoxins, are measured at grading, at the seed lab and at the allergen lab, and the grain ships without them. Why it matters. A reported unusual figure is a process adjustment. The same figure unreported is a failed batch. Where we are. The Lot Reporting Standard proposes what a lot must declare, across gluten-free grains. It asks no seller to generate new information, only to forward what their own paperwork already says. What we know / don't know. Known: which figures exist and where they are measured. Not known: whether any maltster or grain seller will adopt the standard, and whether buyers will ask for it. Open contradictions. None. Research asks. None. This is an adoption problem; the sorghum method half of it is Ask 5. Status. Open — proposal standing, no adopter recorded.

Industry and supply chain​

The market for malting sorghum has to be created before it can be supplied

The problem. Growers will not plant malting sorghum without a buyer; maltsters will not malt it without a grower and a brewer; brewers will not commit without malt. Nobody moves first. Why it matters. Every technical problem on this page is easier with a supply chain and impossible without one. Where we are. We are trying to break the deadlock ourselves, and we say on that page and on the cultivars page what commercial interest we would hold if it works. What we know / don't know. Known: what the agronomy and the market data say (Sorghum Agronomy and Supply). Not known: whether demand at the scale a maltster needs exists yet. Open contradictions. None. Research asks. None — this is a coordination problem, not a research problem. Status. Open — blocked on coordination.

How this page changes​

A problem is added when a published position names one that isn't here. A problem moves when a research ask opens or closes, when a contradiction is raised or resolved, or when someone outside this site sends evidence that changes a position. Each change is dated below. When a problem is solved, it stays on the page, marked resolved, with the date — the record of what got fixed is part of the map.

Changes.

  • 2026-09-07 — Scope decided (DR-0082): the register covers every gluten-free brewing grain and says why it is sorghum-heavy. Three cross-grain problems added from the directories' gap sections: the malting-quality dataset gap, millet's science deficit, and per-lot reporting. Sixteen problems.
  • 2026-09-06 — The reliable-enzymes problem gains the routes page and its koji position (DR-0080).
  • 2026-09-06 — Open Theories opened between this page and the asks; the reliable-enzymes problem links to its first theory.
  • 2026-08-23 — Research Asks 10 (calcium ladder), 11 (FAN gap), and 12 (rice-hull floor) opened and linked from their problems; three "None yet" entries retired.
  • 2026-08-22 — Page created with 13 problems across eight categories, drawn from the first three published position records (external enzymes, the 190°F baseline, cultivar identity) and Research Asks 1–9.
Research needed