GridNavigator collects data from one or many Building Automation System(s) and provides analytics and a cloud base trending display, but the predictive demand engine is what is unique and useful. The idea here is that the Building Automation System (BAS) can work in concert with the GridNavigator prediction engine. It is reality based, not model-based. Using historical data and NOAA expected weather data, the next day's demand and KW are computed at 1 AM, and recomputed based on TRUE usage and actual weather data every hour from 5:30 AM until 4:30 PM. This creates a very precise prediction that tracks the actual values to 97.2% accuracy. The BAS can use the current and predicted KW and KWH values to do any demand-limiting strategy that the customer desires.
The "Traditional" demand management is usually done by just looking at quasi-real-time values (either campus/building demand, or utility demand) and then doing demand management at that time. The demand LIMIT is usually set lower than the true target, so that the controls have time to operate and shed loads or limit power usage. The BMS is always "working behind" -- it is never in front of energy usage; it is always reactive, with gross controls. The idea of forecasting demand is that you not only know what you will peak, but you will know when you will peak, and the rate-of-change of your usage and demand. This allows the Building Automation System to do more intelligent controls: widen setpoints, but only for as long as necessary to get through a demand peak; intelligently shed loads or reset CHW or HW supplies; limit flows, clamp air exchanges and limit maximum flows, etc. With predictive control, the operations do not have to be as "bang-bang" and control as gross as with the traditional demand management approach that only relies on latent values. Much more occupant comfort is possible, and finer controls can be achieved. Some critical peak situations, where the peaks are predicted to be short, would have little effect on occupant comfort, whereas traditional demand management would have initiated more Draconian equipment shutdowns, and the concomitant discomfort for humans.
With an existing BAS, the demand and usage predictions are written to the Building Automation System (twelve 15-minute values for each). The Building Automation System (BAS) is programmed to use these values, along with the usual schedules, resets, and logic. The logic for demand management would be at a higher priority (typically taking advantage of the 16-level BACnet priorities) than the normal controls, and released when demand curtailment is no longer needed. The energy and cost savings would be maximized within equipment and human-comfort constraints, which the BAS is in complete control of. If there is no BMS system, the demand and usage predictions would still occur, since the metering and energy monitoring is still taking place. But if there is no BAS, we will then use low cost GridNavigator's smart thermostats and Looche Lighting System to control the load based on the forecasted energy.