Introduction to Optimization

The document describes the optimization capabilities of LEAP, which are powered by the open-source NEMO framework. It details the difference between partial cost optimization for single modules and full energy system optimization (ESO) across all transformation modules and selected demand end-uses. The guide explains that optimization aims to find the least-cost capacity expansion and dispatch by minimizing the total net present value of social costs, subject to constraints like emission caps. It also covers the implicit Reference Energy System (RES) in LEAP and the options for perfect versus limited foresight optimization.

• LEAP performs optimization through integration with NEMO, an open-source framework written in the Julia programming language. NEMO is a separate, free installation that connects to LEAP to calculate the least-cost capacity expansion and process dispatch.
• Optimization in LEAP is available in two primary modes: partial cost optimization, which targets a single Transformation module (such as electricity generation), and full energy system optimization (ESO). ESO allows for simultaneous least-cost calculations across all Transformation modules, optional optimization of demand end-uses with useful energy demand, and the modeling of transmission systems like pipelines and electricity networks.
• The optimization framework defines the 'optimal' system as the configuration with the lowest total net present value of social costs over the calculation period. This calculation considers capital costs, salvage values or decommissioning costs, fixed and variable O&M costs, fuel costs, and environmental externality values, while adhering to constraints such as emission caps, resource reserves, and technology penetration limits.
• LEAP utilizes an implicit Reference Energy System (RES), meaning it automatically creates connections between resources, processes, and end-use demands based on module properties and ordering, rather than requiring the user to explicitly specify the network.
• The software supports two types of foresight: perfect foresight, where all years are optimized simultaneously, and limited foresight, where years are divided into groups and optimized separately, with each group providing the starting conditions for the next.

APA

Stockholm Environment Institute (n.d.). Introduction to Optimization. https://leap.sei.org/help24/Optimization/OptimizationIntroduction.htm

Chicago

Stockholm Environment Institute. Introduction to Optimization. n.d. https://leap.sei.org/help24/Optimization/OptimizationIntroduction.htm.

Wikipedia

{{cite report |author=Stockholm Environment Institute |title=Introduction to Optimization |url=https://leap.sei.org/help24/Optimization/OptimizationIntroduction.htm |access-date=17 August 2026 |via=Climate Insights Directory}}

BibTeX

@techreport{stockholmenvironmentinstitutendintroduction, author = {{Stockholm Environment Institute}}, title = {{Introduction to Optimization}}, institution = {Stockholm Environment Institute}, url = {https://leap.sei.org/help24/Optimization/OptimizationIntroduction.htm}, urldate = {2026-08-17}, note = {Indexed by Climate Insights Directory} }