Half-Barrel Pints, Foam Loss, and What a Draft Beer Actually Costs

The first variance count I ever ran on draft beer came back looking like somebody was robbing me.

The number on the spec sheet says a half-barrel keg gives you 124 sixteen-ounce pints. The number on my count book said something closer to 95. That is a thirty-pint gap. Run that math at indie-bar pricing, seven dollars a pint, and that is two hundred bucks per keg unaccounted for. On a four-keg-a-week well IPA, that is real money.

My first thought was theft. My first thought is always theft. That is how new managers think, and I was a new manager.

My first thought was wrong. The kegs were losing twenty-five pints to foam, line cleaning, and a regulator pressure that nobody had touched in months. Nobody was stealing anything. The bar was just bleeding.

This post is the math I wish I had on a printed sheet pinned above the keg cooler when I started.

How many pints are actually in a keg #

Specs first. Then the real-world adjustment.

The standard sizes:

• Half barrel (the standard American keg): 15.5 gallons = 1,984 ounces = 124 sixteen-ounce pints (Boulevard Brewing, Micromatic).
• Quarter barrel (sometimes called a “pony”): 7.75 gallons = 992 ounces = 62 pints.
• Sixtel (the small one, common for craft taps): 5.16 gallons = 660 ounces = 41 pints.

Those are the numbers on the keg spec sheet. They are the numbers Toast and BinWise use in their wholesale calculators. They are the numbers your distributor will quote you when you complain about a keg that ran light.

They are also the numbers that do not match what you will actually pour.

A real-world half-barrel after foam loss, line cleaning, and end-of-keg sediment will give you somewhere in the 110-118 pint range (Toast, BinWise). That is a 5-10% gap between theoretical and actual. The gap is not a flaw in the keg. The gap is the cost of running a draft program. You build it into your math or it eats your margin without your permission.

Where the missing pints go #

Three places. In rough order of size:

Foam. The first pour from a fresh keg is mostly foam. So is the first pour after a regulator pressure shift. So is every pour from a line that has been sitting all day on a slow Tuesday. Foam is air mixed with beer; you cannot sell it. A poorly tuned bar bleeds 5-8% to foam alone (Brewers Association).

Line cleaning. Beer lines need cleaning every two weeks at minimum (Brewers Association draft beer guidance). Each cleaning pulls a few ounces of beer out of every line. Across eight or twelve taps biweekly across a year, it adds up to 30-50 pints in cleaning loss alone. Necessary cost. Not a leak.

End-of-keg sediment and the change-out. The last 8-16 ounces of every keg is unsellable foam-and-sediment slurry. The change-out itself spills another 4-8 ounces while the bartender swaps lines. Per keg that is a half-pint to a pint of waste just for the change. Across 200 kegs a year, that is 100-200 pints.

Add it up: 5-10% loss is not a sign you are being robbed. It is the cost of doing draft. The 5% number is for a tuned program. The 10% is for a program where the regulator has not been touched since the lease was signed, the lines are cleaned “when we remember,” and the bartender on the well shrugs and says “it foams a little, it’s fine.” That last one is not normal drift. That is a bar that needs a Saturday afternoon and an actual sit-down with the keg.

What a pint actually costs you #

The math is simple. Wholesale price of the keg, divided by real pints, not theoretical:

Cost per pint = wholesale keg price ÷ servable pints (110-118 for a half barrel)

A typical regional craft half-barrel runs $180-$220 wholesale (BinWise). Call it $200.

• $200 ÷ 124 theoretical pints = $1.61 per pint (the spreadsheet number)
• $200 ÷ 115 actual pints = $1.74 per pint (the honest number)

That thirteen-cent difference does not sound like much. Multiply by 115 pints, multiply by 4 kegs a week, multiply by 50 weeks. That is roughly $3,000 a year the spreadsheet was hiding. Run that across six taps and you are looking at a real number that nobody on the leadership team is talking about because the spreadsheet says everything is fine.

Profit per pint and the break-even point #

Draft beer pour cost should run in the 18-24% band, with draft on the lower end of liquor’s range (Toast, USBG, BinWise).

At $1.74 actual cost per pint and a $7 retail price, your pour cost is 24.9%. Slightly hot. Healthy range, but at the top.

Profit per pint at that math: $5.26.

Break-even on a $200 keg? 38 pints sold. Anything past 38 is gross profit. A half-barrel that sells out gives you $805 in revenue, $200 in COGS, and roughly $605 in gross profit per keg. That is if you actually got 115 pints out of it.

If you only got 95 pints because the regulator is wrong and the lines have not been cleaned in two months, you sold fewer pints, you turned the same revenue into less profit, and your pour cost just hit 30%. That is the same math that would flag a calculator like ours red.

Failure modes (the order to investigate) #

When draft cost runs hot, here is the order to check. Not theft first. Theft is almost never first. Not because bartenders never steal. They occasionally do. But variance flags trained-incompetence and equipment-drift weekly. It flags theft maybe once every five years.

1. Regulator pressure. Should be 12-14 PSI for most ales, slightly higher for some lagers (Brewers Association draft beer guidance). Anything above 16 PSI is foam city.

2. Line cleaning schedule. When was the last clean? If the answer is “uh, around the holidays,” that is your problem.

3. Glass temperature. A warm glass produces a beer that is 80% foam. Frosted glasses cost less than your variance gap.

4. Pour technique. New bartenders tilt the glass wrong. Angled pour against the side, then straighten at the end. Show them once on a slow shift. Watch them on the next busy one.

5. End-of-keg discipline. Some bars push every keg to the last possible drop. Some change them at the foam point and accept the half-pint of loss. The first costs you customer trust on the last six pours. The second costs you a half-pint. Pick the second.

6. Then you look at theft. Not before. Five steps before. The bartender on the well is not your problem. The PSI on the regulator is.

What about shelf life #

A pristine pint is worth nothing if the keg goes flat in the cooler.

Pasteurized domestic kegs hold quality for 90-120 days at proper cold storage. Unpasteurized craft kegs hold 45-60 days (Brewers Association). After that, the beer is technically still safe, but the carbonation is gone and the customer can taste it. You can pour from a stale keg. You will not pour from it twice to the same regular.

The implication for cost-per-pint: kegs that move slowly get more expensive even when nothing is wrong. A sixtel of a slow-moving sour that gets tapped on day 1 and runs out on day 50 is paying its way. The same sixtel that runs out on day 70 lost a percentage of its quality somewhere around day 45, and the last fifteen pints poured tasted worse than the first.

Match keg size to weekly turn rate. A draft line that sells eight pints a week does not need a half-barrel. It needs a sixtel. The math on a sixtel is tighter (you lose more proportionally to foam and the change-out), but a sixtel that sells out in three weeks beats a half-barrel that sells out in fourteen and tastes flat for the last four.

Sixtel vs half-barrel: when to pick which #

Quick rule of thumb. Take your weekly draft sales for that line in pints.

• Under 10 pints/week: sixtel. Fresh beer beats theoretical cost-per-pint efficiency.
• 10-30 pints/week: quarter barrel (“pony”), if the distributor stocks it for that brand. Otherwise sixtel and don’t lose sleep.
• 30+ pints/week: half-barrel. The cost-per-pint advantage is real and you will move it before it goes flat.

A bar that puts a half-barrel of a slow craft sour on tap “to give it a chance” is paying $200 for a keg that will pour eighty pints across two months and end with the last twenty going down the drain. The variance report flags it as a kitchen problem because the math comes out hot. It is not a kitchen problem. It is a tap-assignment problem.

Run the numbers weekly #

A half-barrel that should give you 115 pints and gave you 95 is either telling you the regulator is off, the lines need cleaning, or the bartender training needs a Saturday. The number does not tell you which one. The number just tells you something needs to be addressed.

The math is the easy part. The discipline of running it weekly, not at tax time, not when the distributor short-ships, not when the owner finally asks why margin is flat, that is the part that keeps you in business through a slow quarter.

If you want a tool that does this math for you without an email gate, we built one. Plug in your keg size, wholesale cost, retail price, and an honest shrinkage estimate. It will tell you cost per pint, profit per pint, and break-even pints sold. No upsell. No sales rep. The math runs in your browser. If you also want to look at the wider picture across spirits, beer, and wine, the liquor pour cost calculator handles that too. And if you have read this far and you are now nervous about the other leaks on your bar, the five real reasons pour cost actually leaks post covers what variance is actually catching when it flags hot.

The number doesn’t lie. The number doesn’t have feelings about your distributor relationships. The number just tells you whether your draft program is actually paying for itself, or whether it has been quietly pouring into the floor mat for the last six months.