What a reset control is doing
Outdoor reset replaces a fixed boiler setpoint with a line. You give the control two anchor points — a supply temperature at your outdoor design temperature and a lower one at a mild outdoor temperature — and it interpolates between them, sending water at whatever temperature the line says for the weather outside. The slope of that line is what controls call the reset ratio.
The reasoning is sound and the benefits are real. Emitters sized for design conditions are oversized every other day of the year, so at milder weather they can deliver the same heat with cooler water. Cooler water means lower standby losses, lower distribution losses, and on a condensing boiler a return temperature that spends far more of the season low enough to condense. On a heat pump it means a higher coefficient of performance, for the same reason.
With the defaults on this page, the line runs from 180 degrees supply at 5 degrees outdoors down to 90 at 60, which is 90 degrees of water over 55 degrees of weather: a ratio of 1.64 to one. At 30 degrees outdoors the line gives 139.1.
The mismatch nobody mentions
Building heat loss is very nearly proportional to the difference between indoors and outdoors. With a 70 degree room and a 5 degree design temperature, the design difference is 65 degrees, and at 30 degrees outdoors the difference is 40, so the load is 40 over 65, or 62 percent of design. That is a straight line, and it is the reason a straight reset curve seems like the natural shape.
Emitter output is not straight. It follows the temperature difference between the water and the room raised to an exponent above one, which is the relationship the radiator output calculator works with. At the design point in this example, the average water is 170 against a 70 degree room, so the design difference is 100. At 30 degrees outdoors, the supply is 139.1, the average water is 129.1, and the difference is 59.1. Raise 0.591 to the power 1.3 and you get 0.505: the emitters are delivering 50 percent of their design output while the building is asking for 62.
Eleven points of design output is not a rounding error, and it is not even the worst of it: the gap widens through the middle of the curve and peaks around 45 degrees outdoors, where the emitters give 25 percent against a load of 39. It is the difference between a house that holds its setpoint and one that drifts a degree or two cool through the part of the season with the most hours in it. And it is not a fault in the equipment or the installation. It is what happens when a straight line is used to track a curve that bends the other way.
| Outdoors | Supply on the line | Emitter output | Building load |
|---|---|---|---|
| 5°F | 180.0°F | 100% | 100% |
| 16°F | 162.0°F | 77% | 83% |
| 27°F | 144.0°F | 56% | 66% |
| 38°F | 126.0°F | 36% | 49% |
| 49°F | 108.0°F | 19% | 32% |
| 60°F | 90.0°F | 5% | 15% |
The two columns agree at both ends by construction and disagree everywhere in between. That is what the sag figure on the calculator is reporting.
Parallel shift, slope, and which one to reach for
Reset controls generally offer two adjustments. Parallel shift moves the whole line up or down without changing its angle, and it is the right tool when the house is uniformly a little cool or a little warm across the whole season. Changing the slope pivots the line, and it is the right tool when the house is fine in mild weather and cold on the coldest nights, or the reverse.
For the sag described above, a modest parallel shift upward is the usual practical fix, at the cost of running warmer than strictly necessary at both ends. Raising the warm end alone flattens the line and lifts the middle without touching the design point, which targets the problem more precisely if the control allows the two ends to be set independently. Some controls offer a curved characteristic rather than a straight line, which addresses it directly.
Adjust in small steps and give each one a full day of similar weather before judging it. Buildings have thermal mass measured in hours, and a change judged after twenty minutes is being judged on the wrong signal.
The limits of this arithmetic
The temperature drop across the emitters is treated as fixed here and in reality it shrinks at part load, which lifts average water temperature slightly and makes the true sag a little smaller than the table shows. Internal gains from people, cooking and appliances are a fixed number of BTU rather than a proportion, so they cover a growing share of the load as the weather mildens, which pushes the same way. Solar gain does too, unpredictably. And the exponent is a property of your emitters rather than a constant.
None of that changes the shape of the argument, only the size of the gap. For the emitter side, the baseboard length calculator and the radiator output calculator. For the load, the heat loss calculator. For what a low supply temperature does to a boiler that then cannot turn down far enough, the buffer tank calculator. And for how many hours a season actually sits in the middle of the curve, the heating degree day calculator.
Questions people ask
What is a reset ratio?
The slope of the line, expressed as degrees of supply water per degree of outdoor temperature. A curve running from 180 degree supply at 5 degrees outdoors to 90 at 60 covers 90 degrees of water across 55 degrees of weather, which is a ratio of 1.64 to one. A higher ratio is a steeper line, meaning the water temperature swings more for the same change in weather, which suits a building with a high heat loss relative to its emitter capacity. A lower ratio suits a building with generous emitters.
My reset curve is set correctly but the house is cool in mild weather. Why?
Very likely the mismatch this page exists to show. A straight line anchored at both ends produces emitter output that sags below the building requirement in the middle, because emitter output follows a power law while building load is nearly linear. With the defaults here, at 30 degrees outdoors the emitters give 50 percent of design output while the building wants 62, and the gap is wider still in the mid forties. A parallel shift upward, or a higher warm end, lifts the middle. Before adjusting, rule out the ordinary causes as well: a thermostat that is satisfying on a warm interior wall, a zone that is not calling, air at a high point, or a control that is limiting on return temperature.
Should I lower my reset curve to save fuel?
Lower water temperature does reduce distribution and standby losses and, on a condensing boiler, keeps return temperature in the condensing range for more of the season, all of which are real savings. What limits how far you can go is whether the emitters still deliver the load at the lower temperature, which is the comparison this page prints. Lower it in small steps, hold each setting through a stretch of similar weather, and watch the rooms that were marginal before. If the boiler is not condensing, the gains are smaller and come mainly from reduced standby and distribution loss.
What is warm weather shutdown and how does it relate to the curve?
It is the outdoor temperature above which the control stops calling for heat at all, and it is usually a separate setting from the two points that define the line. The warm end of the curve and the shutdown point are often set near each other, which makes intuitive sense, but they are doing different jobs: the curve says what temperature to supply while heat is wanted, and the shutdown says when to stop wanting it. Set the shutdown too low and the house runs on into weather where solar and internal gains have taken over; set it too high and the boiler fires for calls the building did not need.
Does outdoor reset work with a thermostat, or instead of one?
With it, and understanding the division of labour makes the settings easier. Reset decides what temperature the water is; the thermostat and the zone valves decide whether it flows to a given zone. A well-set curve means each call is met with water just warm enough, so zones run longer at lower temperature and room temperature is steadier. A curve set too high means the water is hotter than needed and the thermostat spends its time cutting calls short, which is the behaviour reset was meant to remove. A curve set too low means calls never satisfy, which is what the sag in the middle of a straight line produces.