In commercial and industrial facilities, rooftop units (RTUs) are usually turned off or greatly curtailed during unoccupied periods, for example, overnight and/or on weekends. Unfortunately, too little attention is given to the schedule RTUs are programmed for restart. If the desired setpoint is reached too early, energy is wasted, while reaching it too late adversely affects occupant comfort. Optimizing RTU start times means programming each RTU so as to reach its desired space temperature setpoint at the intended time. However, many facility managers fail to appreciate the energy and cost savings benefits of optimizing their HVAC startup schedule and simply program RTU start times to ensure they start early enough to guarantee the setpoints are reached before start of business.
But then there’s also the problem of peak demand charges, which are usually based on the 15-minute period during the billing cycle that had the highest average power consumption level. Coordinating the operation of packaged air conditioners and heat pumps (RTUs) to minimize duty cycle overlap has proven to be a highly effective strategy for reducing both peak demand and total energy consumption. The goal of duty cycle coordination is to time shift individual RTU duty cycles so as to minimize the number of loads operating during any 15-minute demand peak measurement period, while still maintaining an acceptable occupant comfort level.
However, during startup this peak minimization strategy can conflict with the goal of minimizing RTU runtime in advance of facility operating hours. In fact, because the RTUs are generally operating farther from their setpoints during startup than during normal operation, the peak demand problem is even further exacerbated. The consequence of this is that, while peak cooling demand is normally expected to occur during the warmest period of the day when the heat load is highest, we often observe that the daily demand peak occurs during the startup period. To further complicate this problem, lights are often turned on during the same timeframe.
So, how can a facilities manager deal with the conundrum of minimizing RTU startup times while also avoiding the creation of an artificial demand peak? This is admittedly a complicated problem, but let’s starts with minimizing peak demand. First, bear in mind that the utility probably uses 15 or maybe 30-minute time windows during which peak demand levels are determined. So, obviously, we want to stagger the RTUs so that at the very least they don’t all start running during the first 15-minutes. If the RTUs are of different tonnage, then start the higher tonnage ones first.
Consideration should also be given to multi-stage RTUs as to how long they run after startup in their 2nd or 3d stage modes so as to not bring additional RTUs online until the multi-stage units have settled into their 1st stage mode. If you know the time it takes each RTU to get within an acceptable target temperature, then this information can be used to prioritize the starting order, allowing the slowest RTUs to start first and working your way down to the quickest to setpoint starting last.
Those of you who would prefer to avoid the time and hassle of optimizing your RTU startup schedules yourselves might be interested in taking a look at Energy As a Service by Encycle or EASE™. Our patented Swarm Logic® enabled by Swarm Controllers and new SwarmStat smart energy solutions provide transparent and effective solutions to problems like scheduling RTU startup times to minimize the total energy costs associated with both energy consumption and peak demand.