Finishing & Filtering

Everything up to this point is water removal. You can do that badly and still recover — a slow boil just costs you an evening. But the last few degrees, the filter, and the jar are where the batch either becomes syrup you can sell or becomes a lesson. I boil on a turkey fryer in Michigan, buckets on spiles, sugar and red maple, and I sell what comes off it at a store and at a farmers market. So the end of the boil is the part I’ve actually lived, as opposed to the equipment page, where I’m frankly just shopping.

This page is about the finish: hitting the right density and knowing why the number in your head is probably wrong, reading a hydrometer that has two lines on it for a reason, understanding niter as a scorching problem rather than a clarity problem, filtering hot without ruining the batch in the filter, getting it into a jar that won’t mold, and recognizing the faults when they show up.

The single most useful thing I can tell you before any of it: the legal density number you’ve read online is almost certainly Vermont’s, and you probably don’t live in Vermont. I don’t. That mistake is in published sources, and it’s the reason this page exists in the shape it does.

Finishing — and the number the internet gets wrong about my state

Finish at 7.1°F above the boiling point of water. At sea level on a standard day that’s 219°F — but 219 is not a constant. It’s 212 + 7, and 212 is only true at sea level at standard pressure. Smoky Lake: “The boiling point of water/syrup is NOT a constant. It fluctuates with environmental conditions such as altitude and barometric pressure.”

So boil water and read your thermometer at the start of every session, then add 7.1°F to that reading. Barometric pressure moves day to day at a fixed altitude. Hard-code 219 and you’ll make under-density syrup on a low-pressure day (ferments, molds) and over-density on a high one (crystallizes). The delta is real; the absolute number is an artifact. (Sources say 7.0, 7.1, and 7.1–7.3 — rounding, plus the delta shifts slightly with target Brix. 7.1 is the figure tied explicitly to 66.0 Brix.)

🔴 66.9 °Brix is Vermont’s law. Not Michigan’s, and not the USDA’s.

You’ll read everywhere that “the legal minimum density is 66.9 °Brix.” That is Vermont-only.

Jurisdiction	Minimum	Maximum	Reference temp
Vermont (packaged)	66.9 °Brix / 36 °Bé Mod. 145	68.9	60 °F
Vermont (bulk)	65.9	68.9	60 °F
Maine	66% by weight	68.9%	68 °F
USDA federal	66% by weight	68.9% (Grade A)	—
Michigan	66.0 — adopts USDA by reference	per USDA	—

Vermont’s regulation (20-011-002 Code Vt. R., auth. 6 V.S.A. ch. 32): “All grades of packaged maple syrup shall have a minimum density … equivalent to 36 degrees Baume Modulus 145 or 66.9 degrees Brix at 60 degrees Fahrenheit.” Maine (7 M.R.S. §892-A): solids “may not be less than 66% by weight.” USDA’s 2015 standards: not less than 66 percent.

Michigan sets no number of its own. The Michigan Maple Syrup Act (MCL 289.531) adopts the USDA standard by reference — so Michigan’s floor is 66.0. That’s mine. When I say “66,” I mean in Michigan, and I name the state because the number changes with it.

Two published sources get this wrong, and they’re not fringe. A 2009 Maple Syrup Digest article says Vermont and Maine are both 66.9 — and the convenient excuse (“Maine changed in 2013”) fails, because Maine’s pre-2013 text also says 66% and the original 1991 grade text says 66.0. And a vendor hydrometer guide states “66.9 Brix at 60°F (per USDA)” — attributing Vermont’s stricter standard to the federal government. That’s exactly the error that propagates.

The resolution: finishing to 66.9 at 60°F satisfies every jurisdiction in the table. Finishing to 66.0 satisfies the feds, Michigan and Maine, and is illegal to package in Vermont. Know which you’re claiming. (I went looking for a published “Grade A minimum Brix” figure specifically and did not find one — so I’m not printing one.)

Both directions fail. OMAFRA: below 66 Brix brings “legal, quality, economic and potential food safety issues. Low density syrup may ferment or mould.” Above max: flavor problems and crystallization. And over-density isn’t the “safe” error — packing at 70 Brix instead of 67.5 means a 1-liter jug holds the sugar equivalent of 1.05 L. At $18/L that’s $0.90 given away per liter. At a market, that’s money handed across the table.

Hydrometer vs. refractometer

A maple hydrometer has two lines because it’s calibrated at two temperatures. Hot test: 211°F, reads 59.1 °Brix (32.1 Baumé). Cold test: 60°F, reads 66.9 °Brix (36 Baumé). Syrup is less dense hot, so the hydrometer floats lower and reads lower. Same syrup, read hot. (⚠ An AI-written summary I hit claimed “syrup becomes less dense as it cools.” Backwards. It gets denser — which is exactly why the cold line sits higher.)

The split: the hydrometer tests hot, during the boil — that’s its whole advantage, and it’s what lets you finish a batch in real time. The refractometer can’t (cool and filter first) but reads at most temperatures, needs a drop, calibrates against distilled water every session, and works on sap. Get both. For hot-test corrections one vendor gives 0.2 Brix per °F from 211 — single source, uncorroborated, treat as a hint; safer is a Murphy compensation cup or reading the thermometer at the same instant and using the maker’s chart.

⚠ Re-check density right before bottling, not just at draw-off — more evaporation happens between evaporator and bottle. Temperature gets you to draw-off; it doesn’t certify the packed product. (Nobody made a defensible numeric accuracy claim comparing the two instruments. Genuine gap.)

Niter — and why it’s a scorch problem, not a clarity problem

Sugar sand / niter is what precipitates as water leaves. Chemically, per the 1963 J. Food Sci. analysis: 34–86% small sugar crystals, 1–50% calcium malate, plus calcium, malic acid, and undetermined material probably including silica. Those ranges are wide because the variability is real — composition varies by site and by point in the season.

Two mechanisms. (a) In the pan, on concentration — minerals exceed solubility as water leaves; calcium malate is “one of the least soluble salts in the concentrated syrup” and comes out first. (b) In the jar, on cooling — “solubility of various salts remains high, but when the finished product is cooled to room temperature, the solubility of certain salts drops below the saturation limit, and crystals precipitate.” (b) is why a jar that left the filter clear goes cloudy on the shelf. Filtering removes only what has already precipitated.

The part that scared me straight, from Vermont’s state grader: niter scale insulates the syrup from the fire. The metal under the scale overheats and syrup burns against it — at a bulk temperature that would otherwise be fine. Marckres calls the result burnt niter: “a burned taste from the niter rising off the front pan and the syrup burning,” plus “a niter flavor, which has a slightly fizzy affect like baking soda on the tongue.” Niter management is scorch prevention, not clarity.

His advice is commercial — “the producer switches draw-off sides as needed, or changes front pans.” ⚠ I can’t do that on a flat pan and neither can you. The hobby equivalent (drain and scrub between boils, finish in a separate pot) is what everyone does, but I did not find it stated in a citable extension source. That’s a real gap in the literature, and I’d rather say so than dress folklore as guidance.

Filtering

Temperature is the whole ballgame, and the sources span 180–219°F — which isn’t sloppiness, it tracks the method. Gravity/cone wants syrup as hot as you can manage, because gravity is the only driving force. A filter press supplies pressure and works at 180°F. Vacuum runs above 200. UW–Madison Extension gives 180–190 generally. Never quote a filtering temperature without naming the method. Hot syrup passes; cool syrup won’t pass at all, or passes so slowly the filter clogs and you lose the batch into the medium.

The hobby method, cone and Orlon: inspect every filter for holes, every time. Nest 3–4 cone prefilters inside the main wool or Orlon cone, openings up. Hang over a pot. Pour hot syrup slowly into the innermost. When flow stalls, lift out the innermost prefilter and keep going with the remaining layers — that’s the technique, you shed clogged layers rather than restart. Wear rubber gloves. Syrup at 180–219°F causes serious burns; that’s not a garnish on the instructions.

Wool filters better but runs slower; synthetics work best damp. ⚠ Wet or dry is unresolved — the same vendor says “dampen the filter bags with clean, hot water before use” in one place and “pour it into a dry cone-style felt filter” in another. The reconciliation is probably material-dependent, but no source says it cleanly, so I won’t present either as the rule.

🔴 The finding that would have cost me a batch, from Marckres:

“During the manufacturing process, these filters pick up and retain a slight chemical odor and flavor. Before use, they should be boiled in clear water and dried thoroughly. If not, they impart a chemical flavor to the syrup.”

A brand-new filter used straight from the bag will taint your syrup. And on reuse: “never washed with any detergent.” After the season, wash in water, dry thoroughly, store dry and away from odors — “filters not dried thoroughly will mold, creating musty off-flavor when hot syrup is filtered through them the next season. Never store filters with mothballs.”

Presses and DE. Diatomaceous earth goes into the syrup before the press; it traps impurities and gets caught by the papers, building the actual filter medium in place. The papers alone don’t do the fine work — the DE cake does. Must be food grade. CDL’s precoat: 4 cups of filter powder in 3½ gallons of hot water, which wets the filters, warms the press and lays the first coat; feed syrup gradually so the powder distributes before the papers clog. On dosage, Smoky Lake rejects the question I was about to ask: “Don’t ask yourself ‘How much DE do I need per gallon syrup?’ Instead, ask yourself ‘How much DE do I need per window plate?’” Their 7” press takes 3¼ cups per window plate. A per-gallon figure doesn’t exist and, per Smoky Lake, shouldn’t. A press isn’t hobby equipment for most backyard producers — cone plus prefilters is the hobby answer, and the press is where you go when cone filtering becomes the bottleneck. Selling at a market, I can see that day coming.

Bottling — the little gift jars are the risky ones

Reheat to 190°F after filtering (UW–Madison). Syrup must be 180–190°F at bottling to kill bacteria, yeast and mold (UVM). Don’t exceed 200 — it darkens.

But temperature alone is not the spec, and hobby sources omit this entirely. UVM’s Mark Isselhardt: “how long it needs to be at that temperature range depends greatly on different types of packaging size and material.” And the result is backwards from intuition:

Container	Thermal behavior	Spoilage risk
Gallon / large	Cools slowly	Lower
Quart & pint plastic	Held temperature longest of those examined	Lower
Pint and smaller	”did not have enough mass of hot syrup to maintain 82 °C for at least three minutes after bottling with 85 °C syrup”	Higher
Half-pint glass	”acts as a heat sink and therefore cools syrup quickly”	Highest

Bottling at 185°F into a half-pint glass jar does not deliver the same lethality as bottling at 185°F into a quart jug. The cute gift-size bottles — the exact ones that sell best at a market — are the ones most likely to mold. UVM’s remedies: raise the bottling temperature, extend time in the critical range, or water-bath the filled containers to hold heat after capping. If you sell small glass, this is your problem, and it’s mine.

Headspace: ¼ inch, as little as possible — but ⚠ every headspace figure I found came from general canning blogs, not maple extension; UW–Madison’s bottling page is silent on headspace, inversion and cooling entirely. A headspace figure from a maple extension source: not found. The reasoning is sound (less oxygen; the hot syrup sterilizes the small air volume; trapped air contracts on cooling and pulls a vacuum) but treat the number as soft. Same for inversion — cap and turn upside down ~5 minutes so hot syrup contacts the whole interior and the cap. It’s attributed to Ohio State by a third-party blog and I couldn’t verify it against an OSU publication.

Containers: food-grade, sealable, clean — and rated for hot fill. Marckres: “packaging syrup in containers not designed for hot filling creates a bitter flavor or a flavor that tastes the way some plastics smell.” That includes the pail you carry syrup from the evaporator to the filter in. Metal cans get a three-month rule — “only pack what will be sold in a three-month period” — longer brings a metallic off-flavor and possibly a greenish tinge. On shelf life the sources answer three different questions and don’t actually conflict: UW–Madison’s “shelf-stable indefinitely” is about microbial stability, MSU’s 2 years unopened / ~1 year refrigerated is about quality, and Marckres’s 3 months is container-specific. The one point of unanimity: once opened, refrigerate.

Failure modes — the ones nobody warns you about

The best document in this entire subject is Henry Marckres’s Maple Syrup Off-Flavors — he’s Vermont’s state grader, and UMaine’s grading school hosts it. Read that if you read one thing after this.

Scorch is “a burned flavor with a very strong bite on the tongue and in the throat,” from “operating the evaporator with too low a level of product in the front pan.” Low pan level and niter scale are two independent causes.

The ones you can’t fix. Marckres groups them deliberately: “Even though they are naturally occurring, they damage the normal maple flavor and are considered unacceptable.” You cannot filter or boil them out.

Metabolism — “changes in the metabolism of the tree due to a warming of temperatures,” possible “at any time during the sugaring season, from the first run on,” and “usually a change to colder temperatures reverses its effect.” Tastes “woody, peanut butter, or popcorn,” maybe “cardboard.” ⚠ Marckres’s own honesty deserves reproducing: “This is mostly speculation from working with syrup for many years and research is being done to determine the exact cause.” The state grader doesn’t claim the mechanism. Don’t write it as settled.

Buddy — late season, the tree makes buds, and “usually the production of this type of syrup signals the end of sugaring for the season.”

⚠ And the sources genuinely fight about what it tastes like. Marckres: “chocolaty, almost a tootsie roll type flavor,” going “bitter chocolate” if strong, and “breathing in and out normally intensifies the flavor.” The chemistry literature: “unpleasant cabbage-like flavor,” driven by dimethyl disulfide (DMDS) and dimethyl trisulfide, formed when yeasts convert sulfur-containing amino acids. Chocolate and cabbage are not the same descriptor. Maybe it’s intensity-dependent; maybe grading tradition and chemistry just use different vocabularies for the same molecule. Unresolved — I’m giving you both. (Marckres blames bud formation, the chemists blame yeast on amino acids. Compatible — budding changes the sap’s amino acid profile — but no source actually states that link.)

Four faults that look alike and aren’t: Cloudy = sugar sand or filter failure — filtration. Gritty sediment = sugar sand — chemistry. Crystallized = sucrose dropping out of over-dense syrup — density. Ropy/slimy = sour sap, microbial: sap held warm makes syrup with “a ropy appearance when poured. The flavor is very sour and leaves a slimy coating on the roof of the mouth.” Foamy = ferment, from under-density or dirty barrels.

⚠ Don’t conflate crystallization with sugar sand. They look similar in the jar and have nothing to do with each other — and they get confused precisely because sugar sand is itself 34–86% sugar crystals.

And the list nobody gives hobbyists, all Marckres:

Off-flavor	Cause	What it tastes like
Chlorine/Sodium	Chlorine used on tubing, not rinsed	Destroys maple flavor; watering of the tongue; salty. Clean tubing with water only
Detergents	Non-approved detergents — including used glass jars that kept soap	Soapy; perfume
Rust-preventative paint	Painted bucket/tank interiors, often fish-oil base	Oily, like cod liver oil
Metallic	Long storage in metal cans or poor barrels	Sharp on the sides of the tongue, “like biting tin foil”; greenish tinge
Plastic	Non-food-grade, or not hot-fill rated — including the transfer pail	Bitter; tastes how plastic smells
Filters / musty	New filters not boiled; used ones detergent-washed, stored damp, or moldy; poorly sealed containers	Chemical; musty; back of tongue and throat
Defoamers	Too much	Oily/waxy on the roof of the mouth
Earthy	Tapping into punky, dark, stained, or cracked wood	Tastes like garden soil

Notice the pattern. Most of that enters after the boil — in the transfer pail, the filter, the container. Marckres frames it as “These off-flavors can occur anywhere from the tree to the containers” — and the container end is where I have the most control and the least awareness. I can spend a whole page shopping for an arch. Half the ways to ruin the product live in a five-dollar bucket.

⚠ On mold in a finished jar — the sources contradict each other outright, and it’s a food-safety question, so I default to discard. I’m not telling anyone to reheat and skim. That question belongs over on what to do with it, where the disagreement gets laid out properly rather than tucked into a footnote here.

What this actually costs you to get right

Not much, which is the joke of it. The arch is thousands of dollars and the RO is twelve hundred and the pan costs more than the firebox — and then the whole batch turns on a hydrometer, a thermometer you re-baseline against boiling water every single session instead of trusting 219, a filter you boiled before its first use, and a hot-fill-rated pail. Call it a hundred dollars of gear and about four habits.

That’s the asymmetry I keep coming back to. Nothing on the equipment page can save a batch that gets filtered through a fresh unboiled cone or packed at 64 Brix into a half-pint of glass. The evaporator decides how fast you make syrup. This page decides whether it’s syrup.