Resilient Energy Allocation Model for Supply Shortage Outages

Miguel Alberto C. Mercado, Roy Dong, Allan C. NervesarXiv 2017 · arXiv

Introduction: The Paradox of Power Shortages

In many developing nations, power grids frequently experience Supply Shortage Outages (SSO) during peak demand periods. Paradoxically, while the grid may face a deficit of hundreds of megawatts, private commercial and industrial entities often possess significant unused backup generation capacity—sometimes exceeding the total grid shortage. The challenge for Distribution Utilities (DUs) is effectively "dispatching" this private energy to stabilize the grid.

The Failure of Traditional Economic Dispatch

Standard power system models focus on minimizing the total cost of generation. However, this research identifies a critical "Least-Cost Trap." Because fuel cost functions are monotonically increasing, a model that strictly minimizes cost will naturally drive energy output toward the absolute minimum required level ($P_{min}$). In an emergency shortage where supply cannot meet demand, this traditional economic approach yields the exact opposite of the desired outcome: it minimizes the very energy needed to keep the lights on.

A Multi-Objective Framework for Resilience

To address this, the authors propose a multi-objective convex optimization model that balances two competing goals:

  1. Cost Objective: Minimizing the total generation cost across all participating customers (PCs).
  2. Resilience Objective: Maximizing the total energy generated to mitigate the shortage.

The model introduces a regularization parameter, $\lambda \in [0, 1]$, to navigate this trade-off. When $\lambda = 0$, the model prioritizes maximum energy output (resilience); as $\lambda$ approaches 1, it shifts toward traditional cost minimization.

Visualizing the Pareto Frontier

The researchers validated their model through a numerical simulation involving five participating customers with varying capacities and cost structures. A key finding is the extreme sensitivity of the dispatch to the cost objective.

Figure 1: The Prioritization Pareto Frontier shows how individual PC dispatch levels drop sharply as the weight of the cost objective (Lambda) increases.

As shown in the Pareto Frontier above, the transition from a "resilience-first" to a "cost-first" strategy occurs much earlier than expected. In the simulation, the cost-minimization logic begins to dominate at $\lambda$ values as low as 0.05, and by $\lambda = 0.09$, the model prioritizes cost almost entirely.

Dispatch Sensitivity and Shortage Management

The model also provides insights into how dispatch levels should scale as the total grid shortage fluctuates. As the shortage increases, the utility must demand more energy from increasingly expensive generators once the more efficient units reach their maximum capacity.

Figure 2: Dispatch sensitivity analysis illustrating how energy requests to individual PCs scale relative to the total grid shortage.

Conclusion

The study concludes that during severe supply shortages, utilities cannot rely on standard economic dispatch. They must explicitly prioritize resilience—maximizing generated energy—over cost minimization to ensure the effectiveness of emergency load programs. Future research will explore scenarios involving market power, where large players might influence prices, and the implementation of different auction mechanisms to handle unknown cost valuations.

Summary updated

Abstract

Supply Shortage Outages are a major concern during peak demand for developing countries. In the Philippines, commercial loads have unused backup generation of up to 3000 MW, at the same time there are shortages of as much as 700 MW during peak demand. This gives utilities the incentive to implement Demand Response programs to minimize this shortage. But when considering Demand Response from a modeling perspective, social welfare through profit is always the major objective for program implementation. That isn't always the case during an emergency situation as there can be a trade-off between grid resilience and cost of electricity. The question is how the Distribution Utility (DU) shall optimally allocate the unused generation to meet the shortage when this trade-off exists. We formulate a combined multi-objective optimal dispatch model where we can make a direct comparison between the least-cost and resilience objectives. We find that this trade-off is due to the monotonically increasing nature of energy cost functions. If the supply is larger than the demand, the DU can perform a least-cost approach in the optimal dispatch since maximizing the energy generated in this case can lead to multiple solutions. We also find in our simulation that in cases where the supply of energy from the customers is less than shortage quantity, the DU must prioritize maximizing the generated energy rather than minimizing cost.
BibTeX
@misc{mercado2017resilient,
  author = {Miguel Alberto C. Mercado and Roy Dong and Allan C. Nerves},
  title = {{Resilient Energy Allocation Model for Supply Shortage Outages}},
  year = {2017},
  eprint = {1707.08477},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1707.08477},
}
Figure