Residential thermostatic virtual units flexibility under summer thermal stress and negative price conditions: An opportunity cost based analysis in joint spot and reserve markets

When high summer temperatures coincide with strong photovoltaic output, the spot market can experience short periods of negative prices, while the system must still secure enough upward and downward reserves. In these conditions, residential thermostatic virtual units (RTVU) face conflicting decisions: low spot prices incentivize higher consumption, downward reserve becomes infeasible as temperatures approach comfort limits, and providing upward reserve requires costly deviations from the spot-optimal schedule. Existing models cannot fully explain the resulting abnormal dispatch patterns, reserve shortfalls, or sharp increases in opportunity cost. This study begins from this operating scenario and first constructs a unified power-feasibility region based on RTVU thermal dynamics, linking spot consumption and reserve capability through the same physical constraints. Within this region, we jointly optimize spot power and reserve provision to obtain a time-resolved opportunity-cost profile that captures how the cost of deviating from the spot-optimal path grows with temperature stress. We then apply dual decomposition to the spot-balance and reserve-requirement constraints to design a practical distributed market mechanism. This mechanism allows each cluster to optimize its schedule locally without sharing thermal parameters or comfort bounds, while remaining consistent with the centralized solution. Using real New Zealand price data, the case study shows that opportunity cost rises sharply under tight thermal conditions, that limited downward flexibility is the primary driver of negative prices, and that the distributed mechanism closely reproduces the centralized dispatch, reserve allocation, and price signals.