One element of building automation that I would like to explore further is the concept of just-in-time temperature control. In my last post I gave an overview of how KSX fits into the overall building automation sphere. Now I would like to press further into some details.

Typical temperature control systems are very simple. They normally include seasonal modes: summer and winter as we as occupied and unoccupied modes. This is certainly better than maintaining the same temperature all the time, but leaves much to be desired. Consider for a moment the winter case, where the temperature set point is set to 72 degrees at 7:30 AM and then down to 64 degrees at 6:00 PM. There are several areas where improvement can be made, and when scaled up to the size of a large office tower, significant savings are available.

Firstly, occupancy can be considered. While many controllers allow for the day of week to be included, so that the office is not heated on weekends, we can add in holiday schedules, so days off can be considered. Real optimization can be performed by creating a heating model of each individual office. Every time the temperature set point is changed, the response of the heating system is measured. With KSX we can gather the data from the temperature sensor to see how fast the office can come to set point. This will be different for each office, as their physical configurations are different. Size, heating infrastructure, number of windows, compass orientation will all affect the temperature response. Also, external factors will factor in, such as outside temperature, sunlight exposure, atmospheric pressure etc. KSX can model the temperature response of the office using all of these variables. Over time, the unmeasured variables will be incorporated to as the model adjust to minimize their errors.

Once a model has been created for the office temperature response, we can be smart with our set points. If, on a given day, it will take us 22 minutes to warm up from 64 to 72, we can change the thermostat set point at 7:38 AM, rather than 7:30 AM and save ourselves 8 minutes of increased heating system operation. As the season changes, and sunlight hits our office earlier in the morning, the temperature may increase due to passive solar, and rather than heating from 64 degrees, we may heat from 65. The model will then help us to know we can turn up the set point later. The figure below shows the change in the office temperature response on a typical day (top) versus a cold day (bottom).

When occupancy schedules are known, and we have an accurate temperature response model for an office we can also optimize our unoccupied temperature set point. Is it more efficient to set the unoccupied set point at 64 degrees or should it be lower or higher? And is this always correct, or does it vary with conditions also? KSX can answer these questions, and implement those answers on your building, in real time.

Lastly, KSX can integrate security system information into its decision making. Rather than always turning down the thermostat at 6:00 PM, we can turn it down when the last one in the office leaves, which could be at 6:00 PM, or could be at 5:47 PM, and we can save ourselves 13 minutes of maintaining the temperature. Now imagine this scenario, played out across hundreds or even thousands of offices across your company, or the companies whose buildings you manage. The savings are significant. The investment is small, simply some servers, software and some knowhow.