The test procedure is quite simple. Test the temperatures in each of the rooms. If it’s a hot or cold day and all the rooms are all within two degrees of each other, the building is balanced. Of course there’s more to it than that.

Let’s take a closer look at temperature balance and see how it really should be done in the field. To prepare for the testing, start the system by setting the thermostat for normal conditions required by the current outside temperature. Let the equipment run until it has either cycled once or has run for at least 15 minutes.

Start by recording the time of day and then measure the outdoor ambient air temperature (be sure to test in the shade on hot sunny days). Also record the outside humidity.

Typically, because of speed and ease it’s best to use an infrared thermometer to measure the room temperatures. There are some things to remember when measuring room temperature with an infrared thermometer. It doesn’t measure air temperatures, so you have to choose a standard location in each room so your readings will accurately represent the average room temperature.

When using an infrared thermometer it’s important to understand how your particular instrument reads. The more expensive the instrument, the smaller the area the beam will read. All infrared thermometers are rated by the distance to spot ratio. The ratio is the size of the area being evaluated by the infrared thermometer as it relates to distance. The area being measured becomes larger as the distance increases.

Cheap infrareds costing $100 or less may have a 1 to 4 ratio. This means that it averages the temperature of nearly a 20-sq.ft. area at a wall 10 feet away. Under the same conditions, a good quality infrared thermometer costing $300 to $500, will have a ratio of 1 to 50 and will measure the temperature of an area less than one square foot. The key is to understand how the instrument you are using is reading temperature.

The best test location in a room is chest high, on an inside wall, away from the airflow of a supply register. The test location should also be away from any direct sunlight or other heat source. Measure the temperature in each room.

Next, find the average temperature in the building by averaging the temperatures that you have just taken. It’s best to throw out the highest temperature reading and the lowest temperature reading. Then average the rest of the readings to get a fair average of the temperatures.

Finally subtract the difference between the average building temperature and each room temperature to find the degree difference from average for each room. Adjust airflow for each room that has a temperature difference of greater than two degrees.

In the winter, if the room is too warm, close the damper and vice versa. This test can also be used after installation when the system has been balanced to measure the accuracy of manual J load calculations. Some assumptions of wall or ceiling R-Values and other building components are made when collecting the data for a manual J calculation. Verify the accuracy of your calculations during extreme weather by measuring room temperatures.

One last note, temperature testing is only valid during extreme weather with a temperature difference of 15 degrees or more between inside and outside. If the house is 70F inside and it’s 70F outside, it’s not the right time to effectively temperature balance.

To most of our customers, temperature is our product. Try temperature balancing to evaluate the performance of one of your systems. It’s a fairly fast, simple process that provides evidence of a job well done.

Rob “Doc” Falke serves the industry as president of National Comfort Institute a training company specializing in measuring, rating, improving and verifying HVAC system performance. If you're an HVAC contractor or technician interested in a free temperature balancing test procedure, contact Doc at robf@nationalcomfortinstitute.com or call him at 800-633-7058. Go to NCI’s website at nationalcomfortinstitute.com for free information, technical articles and downloads.