2016 to 2018

Daniel Burillo, Mikhail Chester, and Emily Bondank

Climate non-stationarity, including temperature and heat waves, raises significant challenges for electric power reliability. Power grids are designed to be robust to historical weather patterns, so unpredictability in climate conditions present significant planning challenges for ensuring adequate generation to meet peak demand. This is particularly true in the US Southwest, a region expected to experience significant temperature increases, and that heavily uses electric air conditioning and thermal power generation technologies. Global circulation models (GCMs) project temperature increases as large as 6°C for Maricopa county (Phoenix area) Arizona and Los Angeles county (California). Current US policies plan for weather variance by generally building power generation capacity to meet the 90th percentile, T90, hottest temperature projected peak demand, plus 15%. We assessed how future temperatures affect power demand and planning reserve margins. Under the highest GCM temperature projection scenarios we estimate that local reserve margins will reach 30% less than T90 levels – a 1.5 GW and 5 GW difference for the two counties. We calculated these values by creating engineering physics-based structural equation models that are congruent with historical observations. Paradigm shifts will be needed to plan for climate non-stationarity, that can ensure agility of the complex power system in the balancing of supply and demand.

2014 to 2018

Emily Bondank, Mikhail Chester, Ben Ruddell, and Daniel Burillo

With increasing heat and growing populations, desert cities are of interest for researchers who want to understand how water and electricity provision might be constrained with climate change and affect basic services. Past research has identified that there is a water-energy nexus at the macro level, highlighting that a city’s electricity consumption is associated with water provision and water consumption is associated with electricity provision. For example, in the US it has been found that roughly 4% of electricity generation is used in the treatment and conveyance of water, and 39% of water withdrawals are used for thermoelectric power generation. This interconnectedness implies that as extreme climate events become more frequent, any increased vulnerability in the wa ter system may propagate to the electricity system, and vice versa. To help water and electricity utilities proactively manage their systems in the face of climate change, we used reliability engineering methodology to develop stochastic models of failure in the coupled systems under the threat of increasing temperatures. We identified which components of the respective systems are sensitive to heat, and the mechanisms that may lead to large-scale failure. From failures in physical components (such as pumps, pipes, and transmission lines) to failures in quality (such as water quality due to changes in biological and chemical process, or electricity load balancing), the chance of failure propagating to large cross-scale outages was assessed. Water and electricity utilities in Phoenix, Arizona are used as case studies, because Phoenix is a city that is projected to experience significant increases in temperatures. The types of water infrastructure considered are bulk water ext r action and transmission, water treatment, water distribution, and wastewater conveyance. The types of electricity infrastructure considered are generation plants and transmission lines. Ultimately there are three main interdependencies identified between the water and electricity system that could increase in vulnerability with increasing temperatures. Namely water interruptions from blackouts or brownouts, decreased electricity generation capacity with a water interruption, and water pipe breaks causing electricity substation flooding. We find that though increasing temperatures has the potential to decrease the reliability of the coupled water and electricity systems in Arizona, proactive governance and strategic improvements to maintenance practices can mitigate the risk.

2014 to 2018

Emily Bondank, Mikhail Chester, and Ben Ruddell

Civil infrastructure systems are vital for delivering resources, providing protection, and facilitating many critical activities. Typically, these systems are designed to last for a long time, often on the order of decades, with some systems persisting for over a century. Infrastructure limits of operation are designed based on historical climate conditions, and global climate models project that these conditions no longer represent the exposure that infrastructure will see in the future. One particular hazard is temperature rise in already hot climates. Water treatment and distribution systems are particularly critical to the economic health of a city, especially in hot conditions. The delivery of safe and sufficient water to residents and commercial establishments is vital to almost all residential, commercial, industrial, and public operations. It is especially important that water systems remain reliable as temperatures rise because in addition to greater potable water consumption, the viability of many services may also require increased water consumption. The electricity generation and agriculture industries in particular may need increasing amounts of water into the hotter future, and the delivery of this water is dependent upon both the availability of the resource, and the reliability of the infrastructure. When both the availability and infrastructural reliability is stressed by rising temperatures, there could be a significant threat of provisional inadequacies and consequential economic losses in hot regions of the world.

Past research has shown that with the increase in ambient temperatures in hot regions from climate change, the frequency of outages to the consumer from infrastructural failures could increase by up to 131% b y 2050 under certain system designs, maintenance practices, and climate projections. It is unknown however, where the outages would occur within a region and if the spatial extent and durations of outages could also increase from temperature rise. Through assigning increasing stochastic failure rates to infrastructural components and running time simulations of the components in a geospatial water distribution network, the neighborhoods which could be affected and what the actual impact from water outages might be is estimated. A standard water network with the temperature projections of Phoenix, Arizona is used as a case study because Phoenix is a city that is projected to experience significant increases in temperatures. From a measure of how many nodes fall below the threshold of 20 psi throughout a simulation period of around 80 years, the duration and spatial extent of these water outages has now been projected out until 2099. We find that the annual number of nodes and t he length of time that nodes fall below the threshold steadily increases throughout this century. Thus, water utilities are faced with the challenge of adapting their system design and maintenance practices to reduce the impact of consumer outages into the future.

2016 to 2017

Yeowon Kim, Daniel Eisenberg, Emily Bondank, Mikhail Chester, Giuseppe Mascaro, and Shane Underwood

As climate change continues to shift global climatological patterns with respect to temperatures and precipitation, built infrastructure is becoming more vulnerable. In the US Southwest, climate models predict hotter and drier climates, and in some scenarios, an increased frequency of extreme precipitation events and urban flooding. Roads in highly urbanized desert cities like Phoenix, Arizona are particularly vulnerable to extreme precipitation events given the intensity and frequency of events. Due to the inherent uncertainty of extreme weather predictions, assessing future flooding damages and adapting road infrastructure to manage them remains a difficult task. Currently, there is limited work focused on assessing climate change impacts on roadway infrastructure, and the few urban flooding studies largely overlook the US Southwest. More importantly, there is no systematic way to assess the effectiveness of roadway flooding solutions to manage unforeseen events into the future. In order to be useful over a wide range of potential future threats, flooding solutions themselves must be developed such that their failure to manage water does not compromise the rest of the urban system, e.g., designed according to a paradigm of “safe-to-fail”. Traditional “fail-safe” systems provide robust protection to infrastructures when the risks are accurately predicted and inflicted within the range of a designed safety factor. However, if the system receives a shock that is not foreseen with the historical data, it may lead to a shutdown of the entire system and thus cause an unmanageable and cascading failure. Moreover, the risks and uncertainties faced by urban infrastructures are becoming so great due to climate change that the “fail-safe” paradigm is quickly becoming economically and technically unsustainable. Coupling spatially-explicit flooding forecasts with site- and context-specific “safe-to-fail” strategies can help infrastructure managers make decisions with high uncertainties coming from non-stationary and the unpredictable characteristics of extreme weather events. In this study, we link climate and urban drainage models to predict future roadway vulnerability using the EPA Storm Water Management Model (SWMM) and assess potential flooding solutions based on multiple “safe-to-fail” criteria for Phoenix. The simulation results indicate increased flooding intensity at drainage junctions are the most vulnerable type of road. From the literature, we score 31 separate roadway flooding solutions based on the roadway types that they impact and 19 “fail-safe” and “safe-to-fail” criteria they may possess. The “safe-to-fail” scorecard and a multi-criteria decision analysis (MCDA) analytic hierarchy process (AHP) algorithm are used to rank flooding solutions for Phoenix. Different weight schemes for “safe-to-fail” and “fail-safe” infrastructure characteristics give various flooding solutions in MCDA, which implies that MCDA provides a framework for city governments to make decisions considering various factors that are not easily captured by climate models and/or engineering design criteria for more “safe-to-fail” infrastructure development.

2014 to 2018

Brandon Gorman, Derek Hamel, Emily Bondank, Shaun Atkinson, Daniel Carmody, Aaron Lajom, Mikhail Chester, Nathan Johnson

Utilities that manage critical infrastructures commonly use Supervisory Control and Data Acquisition (SCADA) tools to visualize and control their network. SCADA systems provide immediate knowledge of the system state to guide real-time decisions made by operators. The Resilient Infrastructure Simulation Environment (RISE) combines SCADA capabilities of electric power systems and water distribution systems in an easy to use geographical interface to identify, manage, track, manage, and predict interdependencies and vulnerabilities between infrastructures that are not apparent when looking at one infrastructure in isolation. An example of an interdependency is the 2009 accident at the Sayano-Shushenskaya hydroelectric dam where turbine explosion caused flooding, spilling of oil into the river, power grid blackout, economic loss, and loss of lives. RISE can model extreme events including natural or anthropogenic stressors. The open source infrastructure models OpenDSS (electrical networks) and EPANET (water distribution) are implemented.

RISE enables exploration of technical investigations for power systems, water systems, interdependent power and water systems, and social investigation in terms of the real-time operation of these technical systems. This is enabled by functionality including drag-and-drop infrastructure design, rapid connection of components, an interdependency wizard, real-time simulation and operation, post-scenario diagnostics, resiliency parameterization and quantification, and a client-server architecture that allows n-many users to work together on a single environment in real-time. Three modes of use are provided: an Editor mode for creating water and power system models, a Simulation mode to run models with users engaging with the system in real-time, and an Analysis mode for performance tracking and data analysis. With the n-many user functionality researchers can explore ways that expert operators and novice trainees manage systems and communicate among individuals, groups, and organizations within and across spatial and temporal scales. RISE is developed to be a revolutionary advancement for knowledge generation, knowledge dissemination, and engineering of interdependent power and water infrastructure models. In summary, the goal of RISE is to assist in answering three key research questions: [1] How do users identify and interpret interdependencies and vulnerabilities between infrastructures? [2] How do experts identify, track, and mitigate cascading failures before they happen? [3] What are the key decision variables for operating a resilient infrastructure system?

2017 to 2018

Samuel Markolf, Mikhail Chester, Andrew Fraser, Christopher Hoehne, and Shane Underwood

Climate and technological change are two emerging forces that are likely to have profound effects on the structure, operation, and effectiveness of transportation systems. In particular, technological advances in the transportation system will result in increased complexity and interdependence with other infrastructure systems.

Given these concurrent changes, it is vital for transportation system practitioners to identify and plan for both the direct and indirect – via interdependence with other infrastructure systems – threats posed by climate change. For example, the continued growth of electric vehicles will increase the dependence of the transportation system on the electricity grid. Thus, any prolonged disruption to the electrical system could significantly inhibit mobility. Similarly, any disruption to the transportation system may hinder the ability of response crews to reach and repair failed components of other critical infrastructure systems.

We identify the impacts and vulnerabilities that extreme events from climate change may have on the transportation system – both directly and indirectly. Interdependencies between infrastructure systems are classified as either physical, cyber, geographic, and logical, while failures within and across these systems are classified as cascading, escalating, and common cause. Through this analysis, we seek to answer the following research questions: How does consideration of interdependence and complexity alter traditional vulnerability assessment completed by transportation agencies? Which types of interdependencies and failures appear to be most relevant to transportation systems? What are the most common causes of delay/loss of mobility in the transportation sector and are those causes dependent or independent of other infrastructure systems?

2017 to 2018

Mukunth Natarajan, Mikhail Chester, and Ben Ruddell

As concerns for food security grow, increasing attention is being focused on the ability of certain regains to maintain or increase agricultural production. However, the increasing of agricultural output is constrained not just by water availability but also by access to energy. While there have been many studies that have quantified the relationships between energy and water systems, there remains limited knowledge of the requirements and dynamics of the food-energy-water (FEW) system. Understanding the FEW nexus is critical as constraints on resources shift, and this is particularly true in the US state of Arizona, where access to water and climate change threaten agricultural enterprises. We estimate the water and energy inputs into agriculture in Arizona by identifying t he inter-connectivity between the systems.

The energy and water inputs for agriculture are estimated using the LCA framework accounting for production, transportation and storage. The direct and embedded energy and water invested in the various stages of an agricultural product’s life are estimated. In the production phase, water consumption is modeled using the Pennman-Montieth method, capturing requirements during plant growth stages specific to Arizona’s climate. Energy consumption is estimated using surveys from the Cooperative Extension of Arizona. Transportation energy is estimated using data from the Freight Analysis Framework (FAF) which provides data on food trade between US states. The energy invested in storage is estimated using the specific heats of the various crops. The energy-water nexus relationship in Arizona is used to determine the embedded energy in water as well as the embedded water in energy. The results from the various phases are aggregate d and amount to 26 PJ of energy and 2.6 billion cubic meters of water for the year 2014, or 1.7% and 28% of Arizona’s energy and water use. The efficiency of crops on a per kg and per acre basis was also estimated. Several crops consume large fractions of total agriculture energy but not water, or water but not energy, raising questions about effectively deploying conservation measures. Several conservation strategies were considered – No Tillage, Low Energy Precision Application (LEPA), Low Energy Spray Application (LESA), Subsurface Drip Irrigation (SDI), Replacing fertilizer with manure, off-grid renewable energy, truck fuel efficiency improvements, and storage technology improvements. When compared to the baseline case, LEPA systems are 20% more water efficient as they reduce evaporation while fuel efficiency standards can reduce the energy consumption of trucks by up to 40%. Thus LEPA (36% water savings) and EPA standard truck (8% energy savings) strategies are the best conservation strategies.

The results highlight the distributed nature of energy across the nexus while water is concentrated in the production phase. The results identify the crops and processes where conservation strategies should be focused. Given that energy is more distributed across the nexus than water, broad efficiency policies will be needed.

2016 to 2017

Janet Reyna and Mikhail Chester

Climate change could significantly alter consumer demand for energy in the residential sector, both in the timing and quantity of energy demand, as changing temperatures alter heating and cooling loads of buildings. Warming climates could also lead to the increased adoption and use of cooling technologies, resulting in higher electricity demand. We develop an assessment of changing residential energy demand in Los Angeles County, California, under several climate change temperature projection models and with a growing population. We subsequently investigate the potential for energy efficient technologies to offset some of the increased demand. We calibrate archetypal residential building energy simulation models with actual electricity demand data by neighborhood, accounting for difference in building materials and appliances. Under temperature increases, we find that without policy intervention, residential electricity demand might increases as much as 41-87% (depending on the climate model) between 2020 and 2060. Aggressive policies aimed at upgrading HVAC and appliances, however, could result in increases as low as 28%, potentially avoiding the installation of some new generation capacity. This aggressive intervention scenario also accounts for increased electrification of building technologies (e.g. water and space heating) as well as population growth of nearly 20%. Furthermore, most of these efficiency upgrades are cost-saving to the homeowner. We therefore recommend aggressive energy efficiency measures, in combination with investment in low-carbon generation sources and distributed generation, to mitigate impacts from projected increases in LAC’s residential energy demand.