Worst case: depressurization and backdraft testing. Episode thirty-three of the Gas Service Animated series.

On the bench this water heater drafts perfectly. In the home, it spills. The difference is the rest of the house. Run the dryer, range hood, bath fans, and furnace blower together, and they pull more air out than can leak back in. The house goes negative, the weakest vent reverses, and exhaust rolls back down into the room. That is a backdraft.

Why the water heater, and not the furnace? An atmospheric draft-hood appliance leans on the weakest force in the building: warm flue gas rising by buoyancy alone. There is no fan behind it. Just a few pascals of house depressurization can stall that rise and pull exhaust back down through the draft hood. Fan-assisted and sealed appliances push much harder and resist far more. The natural-draft appliance fails first.

To find the failure, you build the worst case on purpose. Turn on every exhaust. The clothes dryer moves about one hundred fifty cubic feet per minute. The kitchen range hood, another one hundred to one hundred fifty. Add every bath fan, and the air handler if its ducts leak into unconditioned space. Then work the interior doors, open and closed, to find what drives the combustion appliance zone most negative. That worst case becomes your test condition.

Now measure it. Put a manometer on the combustion appliance zone, referenced to outdoors. First read the base pressure with everything off. Then subtract that base, so you see only what the fans create. You want the zone pressure under worst case, in pascals, compared against the appliance limit. Negative means the room is pulling against the vent trying to push exhaust out.

Every appliance has a depressurization limit. Cross it, and you can no longer trust the draft. Verify these against the current B P I standards before you quote them. An orphaned water heater fails at about minus two pascals. A boiler/furnace common-vented with the water heater, minus three. An individual natural-draft boiler/furnace, minus five. A fan-assisted appliance holds to about minus fifteen. A sealed or direct-vent unit, minus fifty. Weakest draft, tightest limit.

Order matters. Fire the smallest-input appliance first. The smallest burner makes the least heat, so its draft is the weakest and coolest, and the easiest to backdraft. If the weakest one holds under worst case, that tells you the most. Test it first, then bring on the larger appliances, one at a time.

With the appliance firing, test for spillage at the draft hood. Hold a mirror or smoke source at the diverter. If spillage clears within sixty seconds, it passes; longer than one minute, it fails. Some programs allow two minutes, so cite the standard you follow. Then confirm draft inside the vent. A common minimum is the outdoor temperature divided by forty, minus two point seven five, in pascals. Fall below that, and the draft is too weak to trust.

One more danger. When the furnace blower runs, leaky return ducts in the mechanical room pull on that space and can spread spilled exhaust through the whole house. The very blower that heats the home becomes a carbon monoxide distribution system, carrying backdraft products into every room. An appliance that passes alone is not proof the house is safe.

Depressurization is how real homes fail. The weakest vent reverses first. Worst case is built on purpose. And the blower can spread whatever spills. An appliance passing in isolation proves nothing. Next episode: carbon monoxide response, and the red-tag.