Flexibility Estimation and Control of Thermostatically Controlled Loads with Lock Time for Regulation Service
- BATTELLE (PACIFIC NW LAB)
A virtual battery model is a simple and general method to quantify aggregate flexibility from thermostatically controlled loads (TCLs), enabling grid operators to effectively coordinate a large number of flexible building loads with supplyside resources in power systems. Lockout controls are designed to avoid wear and tear resulting from short-cycling of hardware. The lock on/off time could significantly affect aggregate flexibility from TCLs to provide ancillary services and may even fail control algorithms designed without considering the lock time constraints. This paper focuses on flexibility estimation and control design for TCLs with lock time constraints to provide frequency regulation service. We first investigate the potential impacts of lock time on TCLs’ aggregate flexibility and control performance. Both control-dependent and control-independent power bounds are derived, based on either previous TCL switching operations or regulation signals. While the control-dependent method provides aggregate flexibility for a given control method, the control-independent method calculates the theoretical maximum of power bounds. Two control algorithms are proposed to better distribute flexibility over time and thereby improve signal tracking performance. The proposed methods are illustrated and validated through simulations.
- Research Organization:
- Pacific Northwest National Laboratory (PNNL), Richland, WA (United States)
- Sponsoring Organization:
- USDOE
- DOE Contract Number:
- AC05-76RL01830
- OSTI ID:
- 1646828
- Report Number(s):
- PNNL-SA-146718
- Journal Information:
- IEEE Transactions on Smart Grid, Journal Name: IEEE Transactions on Smart Grid Journal Issue: 4 Vol. 11
- Country of Publication:
- United States
- Language:
- English
Similar Records
Impacts of Lock-off Time on Virtual Battery Model from Thermostatically Controlled Loads
Scheduling and Control of Flexible Building Loads for Grid Services Based on a Virtual Battery Model