Research Ongoing Funded Projects/ Recently Completed Projects

 Ongoing Funded Projects

Coordinated Ramping Product and Regulation Reserve Procurements in CAISO and MISO using Multi-Scale Probabilistic Solar Power Forecasts (Pro2R)
(B. Hobbs, PI; JHU Prime Contractor; IBM, NREL, and University of Texas-Dallas subcontractors), 2018-2021.

This project will develop an advanced data-driven platform for probabilsoistic short-term (0-6 hr ahead) and mid-term (day-ahead) solar power forecasting and will integrate them into California Independent Service Operator (CAISO) and Mid-Continent Independent Service Operator (MISO) market and control center operations.  It will advance the state-of-the-art using probabilistic solar power forecast information and net-load ramp forecasts to dynamically procure reserves.
 
Objective 1. To accommodate the exploding amount of data relevant to solar forecasting, the data management system of the Watt-sun forecasting system will be replaced by Physical Analytics Integrated Data Repository and Services (PAIRS), which is a completely scalable platform for geospatial-temporal data. It will be an improvement of more than 50-fold in terms of speed, maturity, and data processing throughput compared to the previous system. It will enable “automatic” fusing of satellite, weather and sensor data, and can inject data much faster than the current system. PAIRS can also be used to distribute the forecast data in a scalable matter. It will support a state-of-art REST API and SDK, enabling building of applications on top of the system. In addition, PAIRS will support WMS mapping services, which will help with visualization applications.
 
Objective 2. A new short-term solar forecasting module will be developed leveraging some of the initial work from Solar Forecasting I, where we enhanced “convection-based” forecasting-based GOES satellite observations with a 2D Navier-Stokes equation. This module needs to be more thoroughly tested and then put into operations for the continental US. We will leverage the new GOES-R data in this work; that database has significantly better spatial and temporal and spectral resolution than the previous GOES (-13/-14) satellite data.
 
Objective 3. A powerful innovation that resulted from the Solar Forecasting I project is situation categorization for the machine-learning. This allowed modelling of specific weather situations separately. In Solar Forecasting I, the situations were identified using FANOVA (functional analysis of variance). In Solar Forecasting II, we extend this work by using deep learning techniques, identifying situations based on full images or raster observations, such as from the GOES satellite.
 
Objective 4. All forecasting in Solar Forecasting I was deterministic. Here we will calculate probabilistic estimates for irradiance for points and regions, with > nine quantiles in each forecast.

 

Mid-Atlantic Regional Integrated Sciences and Assessments (MARISA)

If there’s anything that’s certain about the effects of climate change it’s how much is still uncertain. No one knows exactly how much sea levels will rise or local weather patterns (like storms) will change, making future planning extremely difficult for vulnerable coastal and floodplain areas. The Hobbs group and their colleagues are helping planners to better reckon with these uncertainties when making decisions about how we use and protect our coastlines and watersheds. MARISA is a project between JHU, the RAND Corporation, Penn State, and Cornell University. The center, currently funded for five years, was established by the National Oceanic and Atmospheric Administration to help stakeholders adapt to climate variability and change in the Chesapeake Bay Watershed. MARISA’s goal is to help water planners, transportation engineers, land-use developers, policy-makers, and other managers effectively deal with the questions that climate change is bringing about. A focus is on deciding how much adaptability to incorporate into a system while factoring in costs.

 

"PIRE: USA/Europe Partnership for Integrated Research and Education in Wind Energy Intermittency: From Wind Farm Turbulence to Economic Management"
(Project Director: C. Meneveau, Johns Hopkins University)

This project is a US-European partnership for an integrated research and educational program for graduate and undergraduate students, post-docs and faculty. The international collaboration will address pressing research questions that arise when adding inherently intermittent wind sources to power systems. With billions of dollars to be invested in renewable power, improved understanding and better tools for effective use of sustainable but intermittent power sources are crucial. Research will be tightly integrated with a training program that includes carefully designed international experiences.

The project will develop improved knowledge about physical sources of variability and intermittency, such as atmospheric turbulence, and use it to develop next-generation tools for statistical characterizations of variability across multiple temporal and spatial scales. This information will be used for improving wind farm design, grid integration strategies and economic management of grid-integrated wind farms. Specifically, the study will focus on fluctuations at the seconds to hour time scale, for which turbulence within wind farms and micro-scale atmospheric flow variability are crucial. Next-generation computational fluid dynamics tools that (unlike traditional tools) describe fluctuation dynamics, will be developed and validated with laboratory and field observations. Results will be used to establish the statistical characterizations needed for next-generation modeling, forecasting, and control tools to manage fluctuations across various scales in the electric power system, from wind turbine dynamics and wind farm design to integration with power systems and energy markets. New optimal power flow tools that incorporate improved statistical characterizations of wind-farm output variability will be used to optimize resource siting and operations. Market mechanisms will be evaluated for incenting generation owners to respond rapidly to fluctuations and build flexible back-up plants, and for motivating consumers to adapt in real-time to changes in supply availability through demand response programs.

 

Yale-JHU Solutions for Energy, Air, Climate, and Health (SEARCH) Center
(Project Director: M. Bell (Yale); JHU Co-Director: B. Hobbs)

SEARCH's main objective is to investigate emerging energy transitions in the U.S. and resulting air pollution and health outcomes through state-of-the-science modeling and measurements to characterize factors contributing to emissions, air quality and health. Affiliated projects will estimate how these factors affect regional and local differences in air pollution and health today and under global change and will quantify the impacts of key modifiable factors on air quality and health and associated changes under current and future conditions.

Project 1 estimates how energy transitions in the U.S. affect emissions of air pollutants and how modifiable factors influence regional emissions, using state-of-the-art energy/emissions modeling, including critical feedbacks within the energy system. A broader set of emissions, such as from manufacture of energy technologies, will be estimated through Life Cycle Assessment (LCA). Project 2 assesses ambient levels and personal exposures of pollution corresponding to real-world energy transitions in a case study city, by developing novel highresolution monitors and considering modifiable factors, such as commuting patterns. Project 3 uses advanced ensemble on-line coupled climate-air quality models to estimate ambient concentrations under energy transitions and diverse global change scenarios, and quantifies how modifiable factors contribute to regional differences. Project 3 also improves characterization of multipollutants’ spatio-temporal variability through novel bias reduction, uncertainty quantification, and satellite data. Project 4 estimates health impacts of energy transitions, and incorporates epidemiological analysis of under-studied modifiable factors. Project 4 also estimates climate change impacts of emissions from energy scenarios. Global change modeling (Project 3) will be used to estimate how all Projects’ results are affected by future conditions.

SEARCH will produce detailed estimates of the health consequences and air quality conditions under diverse scenarios that integrate regional impacts of energy transitions, climate change, land use, and other modifiable factors. The integrated research projects address all four research questions in the RFA. Results will provide scientific evidence on how energy choices, climate policy, and other considerations can impact air pollution and health. The research framework will provide the basis for future synthesis of energy, air pollution, climate, and health and their critical linkages for other integrated policy strategies and updated scientific findings. SEARCH will have far-reaching impact by bridging multiple disciplines related to energy, climate, air, and health systems.



Recently Completed Porjects

"Performance and Effectiveness of Urban Green Infrastructure: Maximizing Benefits at the Subwatershed Scale through Measurement, Modeling, and Community-Based Implementation," EPA
(Project Director: Arthur E. Mcgarity, Swarthmore College)

Philadelphia's Green City Clean Waters program provides a context for research on how best to manage innovative urban stormwater practices that reduce runoff volume at the source. Implementing this green infrastructure (GI) approach presents municipal officials in charge of urban sewer systems with new and complex challenges compared to the "gray" alternatives that are specified and designed by a handful of technical experts and approved by a committee of decision makers. This project intends to resolve the complexities of GI implementation by engaging in transdisciplinary research on methodologies, the results of which will guide municipal managers and regulators in the development of strategies to establish conditions in the various urban districts that lead to decisions being made, by the thousands of individual decision makers, that cumulatively contribute positively to the desired program goals.

Leading by Dr. Arthur McGarity with a team of experts from multiple disciplines to bring advanced techniques of measurement, modeling, and community based implementation, the project team bears on this problem with the primary goal of maximizing benefit within a context of realistic constraints on overall cost, equitable distribution of benefits, and political feasibility. A fundamental aspect of our research approach is the idea that these methodologies should be developed and applied from the “bottom-up” by engaging our municipal and community partners in all stages of the research and project design.

Press Coverage:
EPA Announces $5 Million in Grants for Green Infrastructure Research (1/21/2014)

 

"Coastal SEES Collaborative Research: Morphologic, Socioeconomic, and Engineering Sustainability of Massively Anthropic Coastal Deltas: the Compelling Case of the Huanghe Delta"

Owing to their extraordinary natural resources and ecosystem services, river-delta coastlines host hundreds of millions of people worldwide. However, the sustainability of society on delta landscapes is uncertain, due to significant human influences including: 1) reduction of sediment - the life sustaining resource for any delta system - as a result of damming and leveeing of river channels, 2) disrupting natural sediment dispersal and deposition patterns within and along river-delta coastlines, 3) accelerated sinking of low-lying deltaic landscapes due to sub-surface water and fossil fuel extraction, and 4) sea-level rise, which threatens to drown deltaic landscapes. The overarching goal of this project is to evaluate river-delta sustainability by merging science that examines physical aspects of delta growth with socio-economic decisions. This research will provide guidance for the sustainable use of vulnerable delta resources, while promoting best engineering practices that protect society and infrastructure from disasters including river flooding, ocean storms, and sea-level rise. Additional broader impacts include training future scholars for the interdisciplinary field of coastal sustainability, creating an internet-based interface to promote global-citizen awareness in coastal sustainability, and developing teaching modules with complementary workshops intended for high-school courses on coastal science and sustainability for underrepresented groups in Houston and Los Angeles. This project is supported as part of the National Science Foundation's Coastal Science, Engineering, and Education for Sustainability program - Coastal SEES.

The crucial resource in building sustainable deltaic coastlines is sediment, and the key control on sediment delivery is river channel avulsions, relatively rapid displacements of river channels and the formation of new river channels. A multi-investigator, cross disciplinary team of researchers will address the following questions of fundamental importance to river-delta coastal sustainability: What are the socioeconomic consequences of altering river channel pathways on a highly utilized delta? Are current delta land loss mitigation strategies sustainable over long-range (decades to centuries) timescales? Can the location of significant future flooding events be predicted? These questions will be addressed using the Huanghe (Yellow River) delta, China as a case study. The Huanghe delta is a compelling region because it is one of the most dynamic and heavily urbanized coastal landscapes in the world. Lessons learned from the Huanghe delta will be exportable to evaluate the sustainability of delta coastlines worldwide. This project will build predictive models for coastal sustainability by bringing together the mechanics of avulsion on deltas, associated channel-shoreline interaction, socio-economic response to natural and engineered avulsions, and the resulting coupled human-natural system dynamics. U.S. researchers in cooperation with Chinese colleagues will create a template for multi-disciplinary coastal sustainability research to help guide future governance and decision making that integrates human-delta dynamics, societal objectives and uncertainty, hazard and land use engineering, coastal morphodynamics, and educational outreach. This project will evaluate whether massively anthropic coastal landscapes can be managed using engineered avulsions to minimize coastal erosion in the face of reduced sediment supply and rising sea level.

 

"Transmission Investment Assessment Under Uncertainty Using a Multi-Stage Stochastic Model Approach with Recourse," US Department of Energy, Consortium for Electricity Reliability Technology Solutions

The introduction of electricity markets, together with increasing interregional trade and the integration of renewables, has made transmission expansion planning more complicated. Uncertainty about fuel prices, the location, amount, and type of new generation and about electricity demand means that transmission investments today may later be regretted as being the of the wrong type, amount or in the wrong location. Traditional deterministic planning methods cannot value the optionality and flexibility associated with particular investments as compared to alternatives whose consequences may be irreversible. Policy models, such NEMS or IPM, that can simulate interregional additions of transfer capability and generation investment and operation response to them implicitly assume that market players have perfect foresight and consistent beliefs over the entire time horizon and that they must commit irreversibly to a particular expansion path today. The resulting projections of transmission investment may be quite different from what investors will do in the face of pervasive economic, technological, and policy uncertainty, particularly if some of the alternative expansion paths allow investors to revise their choices in the future when the value of present uncertainties become better known. Unfortunately, most of the existing literature on transmission planning under uncertainty focuses on either simple one-period decision problems, one-period game-theoretic models with interactions between transmission and generation investment, or multi-period decision problems without these interactions that consider a highly simplified set of scenarios.

This project involves the development and application of a stochastic two- and three-stage modelling approach to capture the multistage nature of the planning problem together with the interaction of demand, generators, and transmission investors in the market in response to future uncertainties and possibilities for obtaining information. Both transmission investment and generation investment in response to transmission availability would be modelled. Multiple scenarios representing broadly different futures concerning fuel prices, load growth, carbon policy evolution, and the location, amount, and types of renewable investment will be defined. Within-year variability in load and wind output, accounting for interregional diversity, would be represented using an OPF framework. The OPF model would initially be a linear programming DC load flow representation, while a decomposition scheme will be developed that would enable use of the SuperOPF or similar models. This model would allow the following questions to be addressed. Are investments made considering uncertainties significantly and systematically different from investments resulting from a deterministic (single scenario) market or planning model? Do uncertainties in the 20202040 timeframe have implications for transmission investments being made now? What are the costs and benefits of increasing flexibility in transmission plans? Are there no-regrets transmission investments in the near-term that are beneficial under most or all scenarios? What are the economic costs that would result from disregarding uncertainties in transmission planning, and what is the value of better information concerning future uncertainties? Preliminary results for the UK available here.

 

"System Dynamics Analysis of Obesity", NIH (PI: Youfa Wang, JHSPH)

Lead by researchers from the Johns Hopkins Global Center for Child Obesity, we are developing and applying systems dynamics and agent-based models that represent the economic, behavioral, and physiological subsystems whose interactions determines the prevalence of obesity.

 

"National Center for Earthsurface Dynamics," NSF Science & Technology Center, University of Minnesota

 

"EFRI-RESIN: Development of Complex Systems Theories and Methods for Resilient and Sustainable Electric Power and Communications Infrastructures", National Science Foundation

Microgrids are small-scale grids (~0.1-10 MW in size) that can be operated independently of the main grid. They can be supplied by a mix of renewable and fossil-generating units for improved sustainability and resiliency during both equipment failures and natural disasters. Microgrids, operated by businesses interacting in retail markets, will provide customers with appropriate incentives to participate in energy savings and grid survivability during emergency conditions. We consider the economic, environmental, and reliability benefits of such grids, considering how their interaction with regional power markets results in changes in operations and investments elsewhere. That work uses regional market models (based on the ECN COMPETES model, which was developed in cooperation with JHU) plus economic, emissions, and thermodynamic models of microgrid operation. More recently, we have focussed on the potential reliability and black-start benefits of microgrids to regional power systems. We are also using cooperative game theory to identify regulatory strategies to ensure that microgrids that have positive net social benefits are profitable in retail markets, while those with negative net benefits would not be.

The stated purpose of NCED is to transform management of ecosystems, resources, and land use, in part by enhancing the usefulness of channel and landscape modeling. By explicitly linking modeling to decision making, scientific information can be used effectively to inform managers and stakeholders about the consequences and performance of management alternatives, and simultaneously, decision analyses can provide feedback on what information would be most valuable for improving management. My NCED work focuses on the use of decision analysis to improve ecosystem restoration decisions. The major questions that are now being addressed by my groups NCED-funded research are: (1) How can we improve elicitation of expert judgment concerning probabilities? (2) What are the optimal mixes and designs of land building projects in the Mississippi Delta? (3) How can stakeholder value judgments be integrated with sediment budgets and engineering analysis to identify preferred strategies for reducing sediment loss in the Minnesota River basin, considering scientific and economic uncertainties? (4) What are the most effective ways to integrate expert judgments, stakeholder preferences, and scientific information in stream restoration?

 

"Model-based Methods for Debiasing Individual Probability Assessments: Theory and Experiments", National Science Foundation (With Robert Clemen, Duke U.)

Subjective probability elicitation is often conducted by risk and decision analysts to obtain probability distributions of uncertain variables that cannot be quantified by other means. Elicitations can be conducted either through interviews or surveys. There is a lot of guidance available on best practices for interview-style elicitations, however, this guidance is lacking for survey-style elicitations. Survey style elicitations are prone to cognitive biases, some of which can be minimized through survey design and some which are present even in the most carefully designed survey. Even expert judgments are subject to variety of biases such as partition dependence, carryover, and overconfidence. The objectives of this research study are: (1) to identify cognitive biases present in elicitation surveys (including partition-dependence, over-confidence, and carry-over bias), (2) to develop model-based approaches to quantify the magnitude and correct for the biases, and (3) to offer guidance on structuring elicitation surveys to minimize biases. Our testing of the models is based upon a survey of Duke University students and a more extensive web-based survey designed to test hypotheses concerning the effect of survey design on the biases.