The flexibility a resource can offer in its power consumption is valuable to the power grid. Indeed, it can for example be used as reserve power to help stabilize the power network. Through methods developed in our research, we have been able to evaluate the amount of reserve power the LADR platform can provide, and emulating its participation to secondary frequency control, one of the mechanisms by which the control of the grid is performed.
Fig 1. In secondary frequency control, the power operator computes a request signal called Area Generation Control signal (AGC) which is sent to reserve providers. In response to this signal, the provider adapts its power consumption up and down in response to positive and negative AGC requests respectively. The provider needs to announce its baseline consumption and its flexibility which determines the magnitude of the AGC requests it will receive.
Experiments have been conducted where the flexibility and baseline consumption for the system are computed for a period of one day. Then, from the start of operation, an a priori unknown AGC signal coming from historic samples is dispatched to the building which adapts its power consumption accordingly, resulting in operation as depicted in Figure 2.
Fig 2. In the top plot, the power consumption in each room is reported with a different color. The middle plot shows the temperature in each room and the comfort constraints, while the last plot shows the AGC request received by the building. The total power consumption [blue] in the top plot is adapted so that it matches the baseline [black] plus the AGC.
An important finding of our research was how flexibility and comfort trade off. A summary of all experiments and simulations is reported in Figure 3, comparing real experiments and simulations for different levels of comfort constraints and experimental configurations. One of the most important factors is how much time in advance the baseline consumption needs to be announced: on average we have observed that for the same level of comfort, the flexibility offered can be increased by 20% of the installed power when the baseline is fixed one hour in advance rather than one day in advance.
Fig 4. The flexibility offered against the level of comfort (as measured by the ALD, a standard metrics for comfort). Solid markers are real experiments and line markers are simulations. Each color correpsonds to a different level of comfort constraints. "With Intraday" refers to the case where the baseline is fixed one hour in advance and "Without intraday" to the case where the baseline is fixed one day in advance.