AMERICAN TK-310C-NI 1**

The registry lists this model with a UEF, the ratio of hot water delivered to energy used, of 0.81. This site combines that value with the DOE test's daily hot-water draw for the high draw pattern. The result is 209 therms of natural gas a year. The registry does not publish this annual figure itself, so it is a site calculation. The underlying data come from DOE CCD, retrieved 2026-09-19. Among the 397 models registered as a gas tankless unit at the high draw pattern, this one places 323rd lowest energy use of 397. The class median is 182 therms, and this model is 27 therms above it. The calculation steps are described in how the annual cost is calculated.

The 209 therms figure applies only to the high draw pattern, which describes a larger household, or one where showers and laundry overlap. A household that uses less hot water than the test assumes would use a different amount, and that value is not available on this page. The registry lists a maximum flow rate in gallons per minute of 4.6 for this unit. This is a flow rating and is separate from the energy figure. The model sits in the highest quarter of its class for energy use, at 27 therms above the 182 therms median. It fits a reader who is checking a model already chosen. It fits less well a reader whose priority is the lowest gas use in the class, who can try the selector by fuel, space and size or find a same-class replacement by model number.

The running cost is not written here because it depends on the state average residential gas price where the reader lives. The calculation multiplies the annual energy of 209 therms by that price. The cost calculator on the reader's own ZIP does this for this model. Any cost figure carries the conditions of the energy figure: the high draw pattern, natural gas, and the federal test's daily hot-water draw. Real use in a home can differ from the test draw, and this page has no data on actual household consumption. At the same gas price, a model that uses 27 therms more than the 182 therms class median costs more to run than the median model, in proportion to that difference.