The Sliding Window and Fixed Window algorithms are based on keeping the demand over an interval below a user-specified target. Figure 1 illustrates an example of demand relative to time. In the example, instantaneous demand is the actual average power during the meter sampling period, which is fixed at one minute. The amount shed is the power that is currently turned off by DLLR. The uncontrolled demand (U) is the sum of the instantaneous demand (P) and the amount shed (S).
The shaded areas in the figure show how much energy is actually consumed. The white areas show how much energy is not consumed due to equipment which DLLR has turned off. Note that even though energy costs are saved, actual energy usage may not be reduced for two reasons. First, DLLR would not know whether equipment failed to turn off, or some other event caused it to turn on again. Second, the reduction of energy usage during the peak time may require an increase in energy usage after the peak time expires to return a zone to acceptable comfort levels. Often, cost savings is based on when the energy is used, not how much energy is used.
Item |
Description |
---|---|
t |
Present time |
t-1 |
One minute ago |
t+1 |
One minute into the future |
S |
Amount shed |
P |
Instantaneous demand |
U |
Uncontrolled demand |
|
Demand at time t |
|
Amount Shed at time t |
|
Ut = Pt +
St
|