I have a simulation based research vs. practical experience type question:

There are a number of academic research papers available which claim 20% to 40% cooling energy savings by increasing the cooling setpoint by 5°F. The savings change by location. Some research papers also provide thumb rules like (approx.) 10% energy savings for 1°C rise in cooling setpoint. The studies I have come across are based on energy simulations.

On the other hand, some 'on the field' engineers (retro-commissioning/energy engineer) I have spoken to suggest, these savings are inflated or at least not achievable with just change in the cooling setpoint. To achieve these savings you have to make changes in VAV controls (basically reduce the minimum airflow) along with changing the cooling setpoint. But these changes (reduction in VAV min airflow during reheat) can not be applied in actual buildings for various practical reasons. These reasons vary from burning up the electric reheat coil, high diffused air temperature leading to stagnation, not enough diffusion due to low velocity, IAQ issues in conference rooms, supply fan stalling, low air throw length etc.. There are savings but of the range 2-5% for changing just the setpoint by 3F (no VAV controls change as it is impractical for reasons mentioned earlier).

However my sample size for 'on-field experience' is small (a couple of engineers), and there is at least half a dozen of research papers based on simulations I could find which give results as I mentioned earlier, which suggest a good potential in energy savings by increasing the cooling setpoint.

Has anyone actually implemented these retrofits (changing the cooling setpoints) in regular existing office buildings (20,000 sqft to 70,000 sqft)? If yes, which camp do you agree with?


Any time you are given rules of thumb you have to understand the variables that were eliminated to make that simple relationship.

Claim: 30% energy savings at 5 degree F reduction in office temperature.


  • Office temperature
  • Outdoor temperature
  • Building geometry
  • Insulation R-value of walls
  • Insulation R-value of roof
  • Insulation R-value of floor
  • Thermal mass of contents
  • Solar radiation
  • Shading
  • Outside air exchange rate
  • $/kWhr
  • Cooling system efficiency

The calculations would be quite involved and vary greatly based on the data you input and the assumptions you make. The biggest variable here is that the rule is stating a percentage rather than an amount.

For example the claim would work for the following scenario:

  • 65F office temperature
  • 70F outdoor temperature

If you changed the office temp to 70F you would have zero heat transfer and you would save 100% of cooling costs.

To defeat the claim consider another scenario:

  • 65F office temperature
  • 110F outdoor temperature

In this scenario if you changed the office temp to 70F you would save the same amount of energy as in the previous scenario, but the savings would not necessarily be 30% of the total cooling energy used.

It is best practice to not trust rules of thumb that you have not created or tested. In this case a better rule of thumb metric might be W/sqftF. This isn't an industry standard; I made it up just now. It still takes out the massive amount of heat transfer calculations, but gives us a much more meaningful answer than a percentage.

Lets say my house is 1500sqft and takes a continuous 3kW to keep it 5F below ambient. So the savings metric would be 3000/(1500*5) = 0.4 W/sqftF

So now I know that for a similar house, dropping the thermostat 1 degree F will roughly save 0.4 watts per square foot. Multiplying by the house square footage gives me the total savings. I could then take that number and divide it by that buildings actually cooling energy use to give a simple (and much more accurate) percentage.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.