How to Measure Temperature Rise on a Gas Furnace

Why this matters

Every gas furnace rating plate prints a temperature rise range, and it is there for the heat exchanger's sake. Too little air across a firing heat exchanger and the metal runs hot: the high limit starts tripping, the burner short-cycles, and over years the stress shows up as a crack. Too much air and the supply air comes out cool and drafty, and on a noncondensing furnace the flue products can cool enough to condense inside a heat exchanger and vent that were never built for water.

The rise is also the cheapest airflow check you will ever make on a heating call. Two thermometers and ten minutes tell you whether the blower and ductwork are letting this furnace run the way its maker rated it. Taken carelessly, though, it is one of the easiest readings to fool: a supply probe that can see the heat exchanger reads radiant heat, not air.

This article is the measurement. The formula that turns a rise into an airflow is owned by Temperature Rise and the Sensible Heat Airflow Formula; the gas side is Gas Pressure Set-Up Numbers Field Card and How to Clock a Gas Meter to Check Furnace Input.

Safety and start state

This test runs a gas furnace at full fire with the cabinet accessible. Before firing:

• Wear a personal CO monitor for the whole test, and confirm the vent connector is intact and connected, with no scorching at the draft hood or the burner vestibule. If you smell gas at any point, stop: everyone leaves, no switches or phones operated inside, call the gas utility from outside.
• Gas input confirmed first. The rise only means something if the burner is firing at its rated input. Manifold pressure set per the maker and the gas pressure card, and input clocked within a few percent of the plate, before you trust any rise. A furnace underfired by a fifth shows a rise about a fifth low and looks like it has plenty of air.
• Filter, registers and dampers as the house normally runs. All supply registers open, no furniture on the returns. If there is a bypass humidifier, close its damper for the test; it recirculates warm supply air into the return and pulls the measured rise down. Reopen it at the end.
• Your two probes checked against each other. Hold both probe tips together in the return airstream for a minute and note any difference. A fixed offset between two different instruments does not cancel in a subtraction; it adds straight into the rise. If the supply probe reads 0.8 F higher than the return probe in the same air, subtract 0.8 F from every rise you take.
• The plate read. Rise range, and for two-stage or modulating furnaces, whether the range applies to high fire, to each stage, or to both. See related: Furnace Rating Plate Read Before Start-Up.

Where the probes go

          SUPPLY TRUNK
   ======================+=========
                         |  SUPPLY PROBE: past the
                         |  first elbow or a few
                         |  feet along the trunk,
                         |  where air is mixed
   +---------------------+
   |  COOLING COIL       |  coil blocks line of
   |                     |  sight to the heat
   +---------------------+  exchanger
   |  FURNACE            |
   |  heat exchanger,    |  NOT here: a plenum
   |  burners            |  port between furnace
   |                     |  and coil sees radiant
   |  blower             |  heat and unmixed air
   +----------+----------+
              ^
   RETURN PROBE: in the return drop,
   ahead of the furnace, where the
   air entering the furnace is mixed

Return probe. In the return duct ahead of the furnace, where the air entering it is mixed. If there is a bypass humidifier or another duct joining the return, close the bypass for the test as above, or read downstream of the junction, so you are reading what the furnace actually takes in.

Supply probe. Out of line of sight of the heat exchanger, in a mixed airstream. On a furnace with a cooling coil, beyond the coil does both jobs: the coil blocks the view, and the air has started to mix. On a furnace without a coil, go past the first elbow or a few feet along the trunk. A port in the plenum straight above the heat exchanger reads the glow of hot metal and the hottest core of the airstream; it can read high by several degrees and convince you a healthy furnace is starved.

Traverse the supply. Take three or four points across the duct. If they spread by more than a couple of degrees, the air is not mixed; move farther downstream.

Steps

1. Fire on high. Set a call that holds high fire: a second-stage call from the thermostat, or the board's test or high-fire setting per the manual. A two-stage furnace measured on low fire is measuring a different input and a different blower speed.

2. Wait for the maker's settle time. The supply temperature climbs for several minutes after ignition as the heat exchanger and duct come up to temperature. The manual states how long to run before reading, commonly somewhere around 10 to 15 minutes. Watch the supply reading; when it stops climbing over a minute or two, the furnace is at steady state.

3. Read both points. Return at its port, supply as the average of the traverse. Write both down.

4. Subtract and correct. Supply minus return, minus any probe offset found at the start.

5. Compare with the plate range. Inside the range passes. Near either end is worth a note; outside the range is a finding.

6. If it is outside the range, fix airflow, not gas. Gas is set to give the rated input. Turning manifold pressure down to cool off a high rise underfires the furnace, drops the flue temperature, and on a noncondensing unit invites condensation in the vent; turning it up to warm a low rise overfires the heat exchanger. Both move the rise, and both are wrong. A high rise means not enough air: filter, blower speed tap, a closed or crushed duct, an undersized return, a dirty blower wheel, or a coil loaded on its entering face. A low rise means too much air for the input (or the input is low, which you already checked). Measure total external static to see where the restriction is (How to Measure Total External Static Pressure on a Furnace), and change the speed tap if the ductwork is reasonable (How to Set Blower Speed Taps for Cooling and Heating Airflow).

7. Re-measure after any change. Same probe locations, same high-fire call, same settle time.

8. Restore the house. Reopen any humidifier bypass damper you closed. Plug every test port you drilled with a proper plug, not tape, so the duct is not leaking at the hole. Return the thermostat to the customer's setting, and confirm the furnace finishes its cycle normally, with the blower running on through its off delay.

Worked example: a reading that lied, then a reading that did not

A two-stage, 96 percent condensing furnace, 80,000 BTU/h input on high fire, with an add-on cooling coil. The plate rise range (as read on this unit) is 30 to 60 F, applying to both stages. The customer reports the furnace "shuts off and restarts" on cold mornings. A previous visit's ticket says "rise 75 F, recommend duct work."

Gas first. Manifold pressure at the maker's setting for high fire, and the meter clocked at about 79,000 BTU/h against the 80,000 plate input: about 1 percent low, within the plate. Input is right, so the rise will mean something. Furnace short-cycling on the high limit is what the customer is describing, and that is an airflow symptom until proven otherwise.

Probe check. Both probes in the return together: 68.0 F and 68.0 F. No offset.

Where the old reading came from. The previous tech's port was in the plenum between the furnace and the coil, directly above the heat exchanger. Reading it again today, after 12 minutes on high fire: 143 F. Against a 68 F return, that is 75 F, the number on the old ticket. It is a radiant-heat reading, not an air reading.

The correct supply reading. New port beyond the coil, about 3 feet along the trunk past the first takeoff. Four points across the duct: 135, 136, 136 and 137 F, average 136 F, spread 2 F. Rise: 136 - 68 = 68 F.

So the old ticket overstated the rise by 75 - 68 = 7 F, but the corrected 68 F is still over the 60 F top of the plate range. The furnace is genuinely short of air, just by less than the old note said, and "recommend duct work" was a guess built on a bad reading.

Why it is short. The filter was a 1 inch pleated filter, loaded. Total external static, filter and coil included, with the loaded filter in place, on the high-fire heat call: 0.82 in w.c. against a plate maximum of 0.50. With a clean filter: 0.64 in w.c. Better, still over. The heat tap was on the board's medium-low setting; the fan table shows the next tap up moves substantially more air at this static.

The fix. Clean filter, heat tap moved up one step per How to Set Blower Speed Taps for Cooling and Heating Airflow. Gas untouched.

Re-measured. After another 12 minutes on high fire: return 68 F, supply average 120 F, rise 120 - 68 = 52 F. Inside 30 to 60, a little above the middle (45 F).

What the rise says in airflow. At 96 percent efficiency, output is about 76,800 BTU/h. Using the sensible heat relationship from the formula card, 52 F implies about 76,800 / (1.08 x 52) = 1,368 CFM, and 68 F implied about 76,800 / (1.08 x 68) = 1,046 CFM. A two-stage 80,000 BTU/h furnace's blower at around 1,350 to 1,400 CFM on high fire is where you would expect it; at about 1,050, it was well short. That conversion only restates the rise, so it is not an independent check. The independent one is the fan table: read the new tap's CFM at the measured static, and it should land near the same figure.

What is left on the ticket. Static was 0.64 in w.c. with a clean filter before the tap change, already over the plate's 0.50, and moving more air through the same ducts only raises it; re-read it on the new tap and record that figure. The tap change got the rise inside the range; it did not fix an undersized return. That is the duct recommendation, now backed by a measured static, not by a radiant reading.

When the rise does not settle the question

• A modulating furnace changes input and airflow together. Check the rise at the input the manual specifies for the test, usually a forced high fire.
• High altitude. Furnaces are derated for elevation; check the rise against the plate at the derated input the installer set, after confirming that input.
• A heat pump with a gas furnace backup (dual fuel). The coil is a heat pump coil and the board may run a different blower speed for the furnace than for the heat pump. Measure on the furnace's own call.

References

• Furnace installation manual and rating plate: temperature rise range, settle time, high-fire test procedure, maximum external static
• See related: Temperature Rise and the Sensible Heat Airflow Formula; Furnace Rating Plate Read Before Start-Up
• See related: Gas Pressure Set-Up Numbers Field Card; How to Clock a Gas Meter to Check Furnace Input
• See related: How to Measure Total External Static Pressure on a Furnace; How to Set Blower Speed Taps for Cooling and Heating Airflow