Using Control Charts To Monitor Room Temperature
- The x-chart control limits:
- The s-chart control limits:
- Process in-control and capable (~2% of the cases). No need for change.
- Process in-control but not capable (~13%). Look at choral room (Figures 1, 2, and 3). Two options exist: (i) reduce the process standard deviation and (ii) revise the specification limits. In the choral room case, revision of the limits is a preferable solution, as this will not affect the customer satisfaction. The choral room is spacious, used by many people, and if the tolerance is increased from 2° to 3°, the process will be almost capable (Cp = 0.91, Cpk = 0.84) and with higher customer satisfaction.
- Process out of control due to identifiable special cause (~49%). The process is out-of-control due to three main causes: (i) turning the HVAC system off during the weekends, (ii) turning it off during the holidays, and (iii) high variability in external temperature. Figures 4 and 5 illustrate the situation when the temperatures go above the upper control limit every seventh day. In July, the facility turns the air-conditioning units off for the weekend, as the classroom is not in use. The weekend readings during unoccupied times should be removed from the analysis and new control limits calculated.
- Figures 6 and 7 illustrate similar situations with average temperatures on the first, second , and third of January, way below the lower control limit. During these three days, the university was closed for the winter break and the heating was turned off. These three-day readings should be removed from the analysis and new control limits calculated.
- Figures 8 and 9 present a situation where the temperatures are out of control due to rapid changes in the outside temperature, usually during the spring and fall months. One may argue that the outside temperature is not a special cause, as the goal of the heating/air-conditioning is to keep the inside temperature constant despite the outside temperature amplitudes. Internal and external temperatures, however, are highly correlated. That is why the external temperature during the months with high amplitudes is a special cause. Adjusting the process during these months is difficult, as the heating/air-conditioning units are set up to respond to the changes in the external temperature. The situation is even more complicated when the system transitions back and forth between heating and cooling modes.
- Process out of control due to unidentifiable special cause (~36%). Figures 8 and 9 illustrate that the auditorium’s November out of control temperatures are not related to the outside temperatures. Management needs to investigate and correct the problem on that specific date. Common problems include power outages (requiring restarting the variable frequency drives controlling the fan motors of the air-handling unit) or failure of the building air compressor (provides control air).
- Process standardization. The setpoint for the meeting room 1 was 74° and for meeting room 2 was 70°. The more consistent the setpoints, the easier the process control.
- Tolerances. Many of the control charts were out-of-control, much higher or lower than the specification limits because of the narrow specifications limits. When the specification limits are set with one-degree tolerance (auditorium and classroom), increase the tolerance from one to two degrees. Small increase in the tolerance, irrelevant to the customer, will make the process capable.
- Process mean higher than the specification target. When the process mean is higher than the specification target, reset the setpoint. For example, in the library reading room, the temperature was consistently much higher than the setpoint of 72°. It is a closed space, with no windows, with computers and a printer constantly in use, generating a lot of heat, which makes the space hot and uncomfortable to use. Reducing the setpoint to 70° or 68° will partially solve the problem, as the heat around the space during the heating season is enough.
- Process mean lower than the specification target. Similarly, for those spaces where the process mean is lower than the specification target, the setpoint could be reset based upon occupancy. Meeting room 1 had temperatures much lower than the specification limit, especially during the summer months. The setpoint was 74°, but the room was kept much colder with the airconditioning running at full power, even when the space was not in use. Similarly, the classroom was not in use during May, June, July, and August, but had a cooling setpoint of 70°. In this particular case, the university is wasting money cooling a space that is not in use. Either reset the setpoint to a higher temperature, based upon room occupancy, or turn the unit off when the space is not in use. Shade the windows to block the sun during the summer months.
- Redesigning spaces. The cafeteria is a large space with an automatic door at the main entrance. The door is constantly open as people go in and out. In this case, the outside temperature influences the inside temperature, making the place cooler in winter months (December) and warmer in summer/fall months (June, July, August, September, and November). Currently, the employees are using individual electric heaters to stay warm during winter months, causing more fluctuations in the inside temperature and incurring additional costs for the university. One possible solution is installing an air curtain serving as a shield against the cold blast coming in through the doors. Another solution, although more expensive, is redesigning the exterior entrance so the west winds are not entering the lobby when people enter the building. Placing window shades will also reduce the sun impact and improve the temperature control during the summer months when the space tends to overheat. Similarly, meeting room 2 was directly exposed to outside airflow through the entrance doors, resulting in the process being out-of-control all the time. Adding a second set of doors could improve the temperature control in this space.
- Allowing users to adjust the temperature. In several spaces, the thermostats were either broken (classroom) or deliberately removed (meeting room 1) to limit the users control over the temperature. The idea was to prevent people from adjusting the setting and to interfere with the room temperature control. Allowing the classroom users to reduce the temperature in November, however, would have saved the university money on heating.
- Consider the thermostat location. The thermostat location tremendously affects the readings. For example, the library reading room thermostat is located directly under the air vent. The alumni house office’s thermostat is located high in the center of a wall picking up the warmer air in the room, resulting in readings that are higher than the temperature that people actually feel in the room. When not practical to change the thermostat location, consider it when determining the room setpoints.