Project purpose

Typical existing expansion planning methods are primarily scenario-based forward simulation, i.e. the designer proposes an expansion scenario and evaluates the impact and cost via simulation. These methods do not incorporate modern methods of optimization-based system expansion. Optimization-based methods are even more critical because the emergence of natural gas as an economically viable and environmentally preferred fuel for electric power generation has altered the mix of the generation fleet and created strong dependencies between the electric power and natural gas infrastructures. Advanced system expansion planning theory, methods and tools are needed to provide system planners with cross-infrastructure visibility of system stress and constraints. Without these advancements to optimization-based expansion planning, future infrastructure designs are vulnerable to coupled collapse of power systems and natural gas pipeline driven by extreme scenarios, e.g. extreme cold or extreme heat that create correlated stress on both infrastructures.

 

Technical approach

In this project, we integrate the individual elements of power system planning and natural gas system modeling into a formulation of a complete joint gas-grid expansion planning problem. The approach includes an elasticity model to couple the gas and grid physical networks with financial markets and includes extreme event resilience that represents the gas-grid physical infrastructures when they are placed under high stress. The resulting formulation is nonlinear and nonconvex making it computationally difficult to solve. To overcome these computationally challenges, this project develops relaxations, approximations, decompositions and monotonicity properties to restore tractability. The prototype algorithms are tested on gas-grid models of increasing complexity and network size and build to an ISO-relevant optimal planning problem.

  1. Russell Bent. Gas-Grid Resilience Planning. Sandia Grid of the Future Workshop, Albuquerque, NM. August 2018.
  2. Russell Bent. Expansion Planning of Joint Electricity and Gas Networks. INFORMS Annual Meeting, Phoenix, AZ. November 2018.
  3. Russell Bent. Expansion Planning of Joint Electricity and Gas Networks. Institute of Industrial and Systems Engineers Conference, Orlando, FL. May 2019.
  4. Russell Bent. jl: Convex Relaxations for Gas Systems Modeling. The International Council for Industrial and Applied Mathematics Congress, Valencia, Spain. July 2019.
  5. Russell Bent. Resilient Network Design of Electricity and Gas Networks, INFORMS Annual Meeting, Seattle, WA. October 2019.
  1. DOE Xlab Innovation Summit. Seattle, WA. Jan. 2019 (invited panelist)
  2. Tri-Laboratory Applied Energy Workshop: Modeling and Analysis of Current and Future Energy Systems. NETL. April 2019.
  1. Bent, K. Sundar, and C. Coffrin. GasModels.jl: An Open-Source Modeling Framework for Natural Gas Flow Formulations. Power Systems Computation Conference, under review.
  2. Ahumada, K. Sundar, R. Bent, and A. Zlotnik. N-k Contingency Analysis for Natural Gas. Power Systems Computation Conference, under review.
  3. Sundar, S. Misra, A. Zlotnik, and R. Bent. Robust Gas Network Expansion Planning. Power Systems Computation Conference, under review.
  4. Sundar, H. Nagarajan, S. Misra, M. Lu, C. Coffrin, and R. Bent. "Optimization-Based Bound Tightening using a Strengthened QC-Relaxation of the Optimal Power Problem," under review (Computational Optimization and Applications)
  5. Carleton Coffrin, Russell Bent, Byron Tasseff, Kaarthik Sundar and Scott Backhaus. "Relaxations of AC Maximal Load Delivery for Severe Contingency Analysis." IEEE Transactions on Power Systems, 34 (2): 1450-1458, 2019.
  6. Bent, S. Blumsack, P. van Hentenryck, C. Borraz-Sanchez, M. Shahriari. "Joint Electricity and Natural Gas Transmission Planning with Endogenous Market Feedbacks." IEEE Transactions on Power Systems, 33 (6): 6397 – 6409, 2018.
  7. Lu, H. Nagarajan, R. Bent, S. Eksioglu, and S. Mason. "Tight Piecewise Convex Relaxations for Global Optimization of Optimal Power Flow." Power Systems Computation Conference, 2018.

2019

In 2019, development efforts focused on releasing the initial model which couples natural gas and electric power systems into a single joint expansion planning formulation. There were three fundamental research directions that were pursued. First, we developed a novel global optimization method that will allow us to scale joint gas-grid expansion formulations to problems with > 1000 nodes. Our initial results, which focus on the electric power system alone, demonstrate the promise of the approach. In this paper, we closed the optimality gaps beyond what was achievable by any prior method. Second, we extended the planning model to incorporate robustness to increased, but uncertain, demand. Here, we relied on recently development monotonicity arguments that allow us to focus attention on extreme points in the uncertainty set of the demand for gas. Finally, we in 2020 are developing models of extreme event induced contingencies and using them to generalize the set of events for which joint gasgrid networks are designed for resilience.

 

2018

In 2018, we developed an initial formulation of optimization-based expansion planning of
natural gas pipelines and power system operations during periods of increased simultaneous demand (i.e. extreme cold weather). These preliminary models demonstrated the technical feasibility of coordinating the expansions of these systems and demonstrated the efficiency and resiliency gains that can only met through coordinated expansion planning. The core elements of electric power and natural gas expansion planning and modeling were also released.