Climate Impacts on New Connected Infrastructure Vulnerabilities

Monday, May 12, 2014 - 07:00
Add to Calendar

Managing the risks posed by climate change to energy production and delivery is a challenge to communities worldwide. Quantity and proximity of available energy supply to meet this changing demand is needed for power grid planning. As climate conditions change, populations will shift, and demand will re-locate; and networked infrastructures will evolve to accommodate new load centers, and minimize vulnerability to natural disaster. Displaced population creates new demand for built infrastructure that in turn generates new economic activity that attracts new workers and associated households to the new locations. Infrastructures and their interdependencies will change in reaction to climate drivers as the networks expand into new population areas and as portions of the networks are abandoned as people leave degraded locations. This study attempted to assess the vulnerability of these new configurations by loosely coupling Connected Infrastructure Dynamic Models (CIDM), technology insertion models, and Energy-Water hydrology models to estimate the emerging vulnerabilities. To achieve this objective, we developed nationwide demand maps for electric power at 1 km resolution using the projected 2057 configuration under temperature and precipitation changes within scenario RCP 8.5 at 12 km resolution through WRF dynamical downscaling of CMIP5 CESM simulations. This map resulted when simulations let power customers migrate and their power usage change from temperature and precipitation drivers. We selected an area of study surrounding Champaign IL as a sensitive area to show the possible insights drawn from coupling these estimates to technology and water forecasts. This new topology was coupled to a CIDM of the interconnected water infrastructure to characterize impacts of reduced water deliveries/thermal discharge constraints on thermoelectric power generation during drought. We employed a quantitative approach to predicting technology evolution, which combines historical data sets with formalized insights of domain experts. Efforts focused on modeling potential technology changes bounded by demand polygons assigned to infrastructure assets. We used a sliding time window over the population of specification vectors indexed by time of invention. Then, a mixture model density function is fitted to the vectors invented. A time-series fit model of the mixture model changes as the window slides forward in time. The results demonstrated the feasibility of generating high resolution demand maps under different climate scenarios, the ability to couple power systems to changes in the water availability and hydrology changes, and the ability to perform what if analyses based on technology policies.