**6. Discussion**

Table 4 shows the annualized costs from the two simulations performed with the current control method and the simulation performed with the proposed steady-state optimal control method. The costs from the simulations were annualized using cooling degree days for the College Station, TX area. The first simulation in the table is the UBO as it is currently operated. The building technician defined temperature of 23 °C was used as the zone temperature setpoint. While this simulation used the least amount in utilities, it also had the greatest cost in terms of loss productivity due to discomfort. The technician defined setpoints is above the PMV optimal range for loss productivity, thus this is to be expected.


**Table 4.** Annualized economic costs of the different simulation scenarios.

The second simulation in the table is the UBO and its current control method but with PMV optimal temperature setpoints (setpoints that give 0 PMV). Worth noting is the significant decrease of 93.1% in the cost of loss productivity just by changing the user-defined setpoints. This of course comes at an increase (approximately 15.4%) in utility cost; however, the total cost was reduced by \$5307.85, or 38.1%. The third simulation is the UBO with the proposed steady-state optimal control method. This method resulted in the greatest decrease of the cost of lost productivity of 95.6% with a slightly higher cost in utilities of 5.4%. The steady-state optimal control method also gave the greatest decrease in overall cost, saving \$6189.48, or 44.5%, of the original cost. This translates to utility savings of \$704.47, or 8.7%, and total cost savings of \$881.63, or 10.2%, over the current control method with PMV optimal setpoints.

A significant observation is that, just by changing the current zone temperature setpoints, the UBO building operators could have immediate savings in terms of increased productivity for a slight increase in utility cost with no change in control methods. Furthermore, additional savings can be had through the use of the advanced steady-state optimal control method. Other benefits of using the advanced controller are that, over time, the modeling identification algorithm used will update and improve the steady-state prediction models automatically, providing for the potential for further savings over time. In addition, the models will adapt as seasonal climate shifts occur and equipment efficiency changes while the current control method would require manual tuning as the system parameters change to maintain the same level of performance.

In the UBO, user overrides of the current control system and setpoints are common. While these overrides may reduce in frequency with a change in zone temperature setpoints to PMV optimal values, the impact of overrides would still be greater with the current control system compared to the proposed steady-state optimal controller. The steady-state optimal controller balances the optimal economic zone setpoint with the user-defined setpoint, theoretically reducing the number of overrides. The advanced controller also has the added benefit of allowing the building operator to easily prioritize zones to maximize comfort by adjusting the weight of the annual salary of respective zones.

The authors acknowledge that not all building operational situations call for maximum comfort and productivity, but propose that the importance of occupant comfort and its significant economic impact on businesses and organizations merits further investigation. As building design and control move forward, optimizing occupant comfort should be considered a priority as opposed to maintaining comfort within a predefined range.
