Optimal HVAC Dispatch for Demand Response: A Dynamic Programming Approach

Demand Response is the generic title given to a range of options
customers have to reduce costs - or even generate profits - by
changing their pattern of energy consumption. Typically this is
achieved by deliberately moving consumption away from peak
times or onto off-peak times. It is generally accepted that Heating,
Ventilation and Air Conditioning (HVAC) provides one of the
largest possible source for energy shifting potential in European
countries. In this paper, steps towards an efficient and partially
decentralised approach to aggregated DR for HVAC units in
individual or blocks of buildings are considered. Specifically, the
paper focuses upon the dispatch problem for individual HVAC
units in the presence of a tertiary DR programme. The dispatch
problem for an individual HVAC zone is formulated as a non-linear
constrained predictive control problem, which manipulates On/Off
controls to optimize thermal comfort and electricity consumption
across a future time horizon. A Dynamic Programming (DP)
algorithm is developed to solve the dispatch problem with fixed
memory and computation time overheads. A simulation study is
presented to illustrate the operation of the algorithm, and the paper
concludes with a discussion of future work.