Incentivizing Energy Reduction for Emergency Demand Response in Multi-Tenant Mixed-Use Buildings

Emergency demand response (EDR), which is the last line of defense to avoid cascading failures during emergency events, has witnessed numerous crucial participants, including buildings and datacenters (DCs). However, even though the majority of DCs are physically located in mixed-use buildings (MUBs), the existing studies on EDR are non-coordinated approaches that separately focus on either buildings or DCs, hence ignoring that both DCs and non-DC (e.g., office) operations share the same MUB facilities (e.g., electricity supply). Furthermore, even when all MUB tenants (i.e., offices and DCs) are jointly considered, tenants will incur different costs to shed energy for EDR, thereby raising an issue of mis-aligned incentive for their participation. To overcome this non-coordinated energy shedding and mis-aligned incentives, we propose two incentive mechanisms in MUBs, such that the total incurred cost is minimized for energy shedding. The first mechanism, namely MECH-NA, is designed for non-strategic MUB tenants. In MECH-NA, the MUB operator provides a mechanism package, including reward rate and a commitment profile with deviation penalty, based on which the MUB tenants will shed energy to maximize the reward and minimize their energy-shedding and deviation costs. We also design a distributed algorithm to implement MECH-NA that can achieve the minimum MUB cost. The second mechanism, namely MECH-SA, is a VCG-Kelly-based mechanism tailored to handle strategic MUB tenants. In MECH-SA, the operator announces both reward and energy shedding rules, based on which the tenants strategically participate in an bidding game. For this game, we not only show that there exists an efficient Nash equilibrium at which the total MUB cost is achieved, but also design a distributed algorithm to implement MECH-SA. Simulation results show that both MECH-NA and MECH-SA can obtain the optimal MUB cost, which outperforms partially or non-coordinated approaches.