2. In dry and marine climates, and for sensors errors representa- tive of well maintained sensors and poorly maintained sensors, differential dry-bulb temperature control yields the lowest annual coil load and almost always yields the lowest peakhourly coil load of any of the economizer strategies considered. In addition, the annual and peak hourly coil loads resulting for differential dry-bulb temperature control are relatively insensi- tive to the magnitudes of the sensor errors.
3. In humid climates, no one strategy consistently produced lower annual and peak hourly coil loads over the range of sensor errors simulated. With well maintained sensors, differential enthalpy control, model-based control, and optimization-based control performed better than differential dry-bulb tempera- ture control when both the annual coil loads and peak hourly coil loads are taken into consideration. With poorly maintained sensors, differential dry-bulb temperature control yielded the lowest annual coil loads and peak hourly loads that were only slightly higher than the corresponding values for the other three economizer control strategies.
All of the economizer control strategies investigated are suscep- tible to sensor errors. Incorrect control decisions that result from sensor errors reduce the energy savings that can be achieved and lead to higher peak hourly loads on the coil. For a scenario repre- sentative of poorly maintained sensors, the lowest annual coil loads of the four economizer strategies investigated are 3.0– 40.3% higher than those resulting from the ideal scenario of opti- mization-based control and ideal sensors, and the lowest peak hourly loads are 6.8–84.1% higher than those corresponding to optimization-based control and ideal sensors.
The model-based and optimization-based control strategies are also susceptible to modeling errors; however, sensitivity to model- ing errors was not investigated here. Such a sensitivity study should ideally include data from a number of coils of different sizes with variations in the air inlet temperature, humidity ratio and flow rate as well as variations in the inlet water temperature and flow rate. The study should evaluate the error associated with using a single optimal bypass factor for all conditions, as well as the sensitivity of the results to suboptimal values of the bypassfactor. The sensitivity of the results to the return conditions should also be assessed for all the economizer strategies using a more de- tailed model of the conditioned space.
Although not investigated, self-optimizing economizer control merits further attention [8,9]. Because sensors and models are not used, this strategy has the potential to achieve the ideal results corresponding to optimization-based control and ideal sensors.