Building hurricane-resilient urban distribution networks by climate-informed system hardening optimization Part B - Climate-Informed Life-Cycle Grid Hardening Demo

Abstract

# Climate-Informed Life-Cycle Grid Hardening Demo ## Scope This capsule presents a compact example of the manuscript's life-cycle economic planning module. It uses a Williamsburg distribution network with five illustrative fault scenarios and example frequency and probability values. The example demonstrates the input format, stochastic-model interface, result structure, and planning visualization. The five bundled scenarios are demonstration inputs. They are independent from the formal scenario set and event weights used for the manuscript results. Formal reproduction of the manuscript results uses inputs constructed through the complete workflow described below. ## Formal Manuscript Inputs The formal manuscript inputs should be constructed through the full climate-to-planning workflow: ```text Climate-conditioned CHAZ events and annual occurrence weights -> asset-resolved wind exposure and HRSRA failure sampling -> representative outage-scenario selection (Supplementary Note 10) -> risk-based 40-year life-cycle planning (Supplementary Note 11) -> climate-conditioned economic aggregation (Supplementary Note 13) ``` Supplementary Note 10 defines the stratified HRSRA procedure used to select representative outage states for formal manuscript inputs. Each selected state approximates the converged mean customer interruption cost for its hurricane event. Supplementary Note 11 defines the two-stage stochastic planning model. First-stage decisions select line undergrounding and distributed-generation investments. Second-stage decisions represent scenario-dependent operation, reconfiguration, load service, interruption cost, and conditional value-at-risk. For formal event frequency `w_e`, the model uses total annual event rate `Lambda = sum(w_e)` and conditional probability `p_e = w_e / Lambda`. Their product preserves each event-specific annual occurrence weight. Supplementary Note 13 defines the propagation of these climate-conditioned event weights through the complete economic analysis. ## Bundled Example Inputs The bundled workbooks contain five illustrative line-failure scenarios. A value of `1` in a scenario column denotes a failed line before hardening is applied. | Example scenario | Example frequency | Example probability | | --- | ---: | ---: | | Scenario1 | 0.0988 | 0.380 | | Scenario2 | 0.0637 | 0.245 | | Scenario3 | 0.0494 | 0.190 | | Scenario4 | 0.0351 | 0.135 | | Scenario5 | 0.0130 | 0.050 | The example frequencies sum to `0.26 yr^-1`, and the example probabilities sum to `1.0`. These totals describe only the demonstration dataset. The model reads the values from `data/Data/Scen_data/Williamsburg/Williamsburg_Scen_Pro_Info_multi.xlsx` exactly as supplied. ## Included Files - `code/Plan_Configuration.py` loads the example settings and calculates the life-cycle yield factor. - `code/Plan_Model.py` builds the two-stage Pyomo/PySP model and reads the five illustrative scenarios. - `code/Plan_Run.py` provides Extensive Form and Progressive Hedging execution routes. - `code/Plan_Result_Write.py` exports PySP scenario-tree solution variables. - `code/Visualization.py` maps distributed-generation sites and hardened lines from the bundled planning result. - `data/Data/Scen_data/Williamsburg/` contains the illustrative fault states, frequencies, and probabilities. - `data/plan_results/Williamsburg/Plan_Results.xlsx` contains the bundled example planning result. - `data/plan_results/Fig/Williamsburg.jpeg` contains the reference figure. ## Example Parameters The example parameters are defined in `code/Plan_Configuration.py`. They include a 40-year planning horizon, a 3.75% annual discount rate, and a 0.5% baseline operation and maintenance proportion. The configuration also uses a 95% CVaR confidence level and risk-neutral weighting (`rho = 0`). The complete manuscript parameterization and discount-rate sensitivity analysis are reported in Supplementary Notes 11 and 14. ## Execution The Code Ocean entry point executes `Visualization.py`. It reads the bundled `Plan_Results.xlsx` and writes `Williamsburg.jpeg` to `/results`. `REPRODUCING.md` provides the Docker commands and the optimization-example workflow. The optimization environment uses Python 3.8, Pyomo 6.0, PySP 6.0.0, and Gurobi 11.0.3. The Extensive Form route is selected in `Plan_Run.py`. The visualization uses OpenStreetMap tiles through `contextily`. OpenStreetMap data are credited to OpenStreetMap contributors under the applicable licensing terms. ## Formal Data and Cost Sources Formal climate-conditioned tropical cyclone tracks for the manuscript are available from the ISIMIP3b CHAZ repository: https://data.isimip.org/10.48364/ISIMIP.808980. The formal event-frequency adjustment used by the manuscript follows Meiler et al. (2026). The full formal input workflow is documented in the manuscript and its Supplementary Information. Undergrounding cost information was obtained from Dominion Energy Virginia Strategic Underground Program reports: https://www.scc.virginia.gov/DocketSearch//Home/Document/12/379283. The network and derived example data are subject to `data/LICENSE`. Source code is subject to `code/LICENSE`. ## References 1. Lee, C.-Y. et al. ISIMIP3b tropical cyclone tracks (CHAZ). ISIMIP Repository (2025). https://doi.org/10.48364/ISIMIP.808980 2. Meiler, S., Lee, C.-Y., Camargo, S. J. et al. Global coastal wind hazard maps from the CHAZ tropical cyclone model. Scientific Data 13, 136 (2026). https://doi.org/10.1038/s41597-025-06452-0 3. Poudyal, A., Poudel, S. & Dubey, A. Risk-based active distribution system planning for resilience against extreme weather events. IEEE Transactions on Sustainable Energy 14, 1178-1192 (2023). https://doi.org/10.1109/TSTE.2022.3220561 4. Virginia Electric and Power Company. Virginia Electric and Power Company's Strategic Underground Program annual report (31 March 2025). https://www.scc.virginia.gov/DocketSearch//Home/Document/12/379283 5. Larsen, P. H. A method to estimate the costs and benefits of undergrounding electricity transmission and distribution lines. Energy Economics 60, 47-61 (2016). https://doi.org/10.1016/j.eneco.2016.09.011 6. Industrial Economics, Incorporated. The benefits, costs, and economic impacts of undergrounding New York's electric grid (2023). https://dps.ny.gov/system/files/documents/2023/09/final-report-ny-undergrounding-2023-06-27.pdf 7. Watson, J. P., Woodruff, D. L. & Hart, W. E. PySP: modeling and solving stochastic programs in Python. Mathematical Programming Computation 4, 109-149 (2012). 8. Watson, J. P. & Woodruff, D. L. Progressive hedging innovations for a class of stochastic mixed-integer resource allocation problems. Computational Management Science 8, 355-370 (2011).