Comparisons of two-stage models for flood mitigation of electrical substations

Brent Austgen,Erhan Kutanoglu,John J. Hasenbein,Surya Santoso

DOI: https://doi.org/10.1287/ijoc.2023.0125

2024-04-29

Abstract:We compare stochastic programming and robust optimization decision models for informing the deployment of ad hoc flood mitigation measures to protect electrical substations prior to an imminent and uncertain hurricane. In our models, the first stage captures the deployment of a fixed quantity of flood mitigation resources, and the second stage captures the operation of a potentially degraded power grid with the primary goal of minimizing load shed. To model grid operation, we introduce adaptations of the DC and LPAC power flow approximation models that feature relatively complete recourse by way of an indicator variable. We apply our models to a pair of geographically realistic flooding case studies, one based on Hurricane Harvey and the other on Tropical Storm Imelda. We investigate the effect of the mitigation budget, the choice of power flow model, and the uncertainty perspective on the optimal mitigation strategy. Our results indicate the mitigation budget and uncertainty perspective are impactful whereas choosing between the DC and LPAC power flow models is of little to no consequence. To validate our models, we assess the performance of the mitigation solutions they prescribe in an AC power flow model.

Optimization and Control,Systems and Control

What problem does this paper attempt to address?

### Problems the paper attempts to solve

This paper aims to compare two decision - making models - stochastic programming (SP) and robust optimization (RO) to guide the deployment of temporary flood mitigation measures before an upcoming and uncertain hurricane to protect electric power substations. Specifically, the research objective is to optimize in the following two aspects:

1. **First stage**: Deploy a fixed number of flood mitigation resources.
2. **Second stage**: In the potentially degraded power grid operation, the main objective is to minimize load shed.

To model power grid operation, the author introduced adaptations to the DC power flow and LPAC power flow models. These adaptations achieved relatively complete recourse through indicator variables. The study applied two geographically realistic flood case studies, one based on Hurricane Harvey and the other based on Tropical Storm Imelda.

### Main contributions

1. **Model extension**:
- Proposed an alternative to the two - stage stochastic programming (SP) model based on DC power flow and considered a two - stage robust optimization (RO) model of the LPAC power flow model based on three dynamic topology variants.

2. **Feasibility indicator variables**:
- Extended the previous recourse problem by introducing a binary decision variable. If this variable is enabled, it allows for a simple recourse problem solution, that is, no power is generated and no load is satisfied. This "infeasibility indicator variable" serves as a rough substitute for complex discrete controls (such as transmission line switching and generator commitment) to ensure the feasibility of power flow.

3. **Case studies**:
- In the computational study, the proposed two - stage models were applied to two geographically realistic flood case studies caused by hurricanes to evaluate the impact of the selected surrogate PF model, uncertainty perspective, and mitigation budget on the optimal flood mitigation strategy. The results show that the mitigation budget and uncertainty perspective have a significant impact, while the choice of DC model or LPAC model has little impact.

4. **Model validation**:
- To verify the validity of the model, the performance of the mitigation solutions recommended by the model in the ACPF model was evaluated. Although the model has some inaccuracies, the results show that they can effectively guide flood mitigation.

### Conclusions

This paper provides decision - support for flood mitigation in electric power substations by comparing different power flow models and uncertainty perspectives. The research results show that in time - sensitive resilience applications, the DC approximation model is more popular due to its lower computational cost.