The transition towards more sustainable energy systems poses new requirements on energy system models. New challenges include representing more uncertainties, including short-term detail in long-term planning models, allowing for more integration across energy sectors, and dealing with increased model complexities. SpineOpt is a flexible, open-source, energy system modelling framework for performing operational and planning studies, consisting of a wide spectrum of novel tools and functionalities. The most salient features of SpineOpt include a generic data structure, flexible temporal and spatial structures, a comprehensive representation of uncertainties, and model decomposition capabilities to reduce the computational complexity. These enable the implementation of highly diverse case studies. SpineOpt’s features are presented through several publicly-available applications. An illustrative case study presents the impact of different temporal resolutions and stochastic structures in a co-optimised electricity and gas network. Using a lower temporal resolution in different parts of the model leads to a lower computational time (44%–98% reductions), while the total system cost varies only slightly (-1.22–1.39%). This implies that modellers experiencing computational issues should choose a high level of temporal accuracy only when needed.

The Spine Toolbox is open-source software for defining, managing, simulating and optimising energy system models. It gives the user the ability to collect, create, organise, and validate model input data, execute a model with selected data and finally archive and visualise results/output data. Spine Toolbox has been designed and developed to support the creation and execution of multivector energy integration models. It conveniently facilitates the linking of models with different scopes, or spatio-temporal resolutions, through the user interface. The models can be organised as a direct acyclic graph and efficiently executed through the embedded workflow management engine. The software helps users to import and manage data, define models and scenarios and orchestrate projects. It supports a self-contained and shareable entity-relationship data structure for storing model parameter values and the associated data. The software is developed using the latest Python environment and supports the execution of plugins. It is shipped in an installation package as a desktop application for different operating systems.

Ireland and other countries in the EU have binding targets for production of energy from renewable sources by 2020. Ireland’s Renewable Energy Action Plan aims to meet this target by producing 40% of electrical energy from renewable sources and most of this will come from wind power. In order to forecast the amount of wind power capacity required, it is necessary to forecast the amount of wind power curtailment that will arise from the need to maintain a certain amount of conventional generation online to provide system services such as reserve, inertia and system balance. Estimation of future levels of wind power curtailment is also necessary for investors. In this paper, a stochastic scheduling model is used to study the impact of forecast error related uncertainty on wind power curtailment estimation. Results are shown illustrating the impact of uncertainty on final energy production from wind power and the impact improvements in forecasting could have on these estimates.

With pressure to reduce greenhouse gas emissions from the energy and other sectors, policymakers are increasingly seeking to mobilise investment in renewable electricity with support mechanisms. The objective of this paper is to assess the value of utility-scale solar PV using a multidisciplinary approach. In many cases renewable energy strategies are developed by investors and policymakers on the basis of a relatively narrow analysis of costs. However, as policies should be designed to maximise societal welfare, a wider paradigm is needed in decision-making. This paper argues for an impact assessment of renewable electricity generation that integrates its value from investor, utility, policy maker and end-user perspectives. We propose a multidisciplinary approach to the assessment, combining engineering production cost estimation with financial and market analysis of utility-scale solar PV in Ireland. The results show that while from an investor's perspective solar PV generation is, and will likely remain so for the foreseeable future, relatively expensive in a country with low solar irradiation such as Ireland, there are benefits for the electricity system in terms of reduced wind curtailment and, to a lesser extent, CO2 emissions. This demonstrates why the wider costs and benefits of integrating a renewable electricity technology to the system need to be considered in assessment of renewable policy support mechanism. Policy makers can then seek to maximise the system benefits while minimising the cost of any policy supports.

As the penetration of variable renewable production increases on systems, the potential need to curtail the output of variable renewables to ensure system integrity increases. The need for these curtailments is primarily due to three factors: firstly, the need for system balance, secondly the need for sufficient system services such as inertia and lastly, as a result of congestion arising from limitations of the transmission network. This paper presents a methodology for modeling these types of curtailments in production cost studies. The methodology is applied to the power system of Ireland and examines the sensitivity of curtailment estimates to spatial and temporal wind data.

The historical time series data or Monte Carlo simulation approaches that are often used to represent wind power in transmission planning models will lead to large-scale optimisation problems. The complexity of such problems will be further compounded if advanced techniques for wind variability and wind forecast uncertainty management are also endogenously included, corresponding to a merging of the traditionally separate 'real-time operations' and 'long-term planning' analysis timeframes in power system analysis. A stochastic mixed-integer scheduling model is applied here to investigate the likely transmission planning model formulation impacts of advanced wind forecast techniques, and to determine whether any additional optimal transmission planning model precisions offered justify the associated very-large-scale computational burden. Results indicate that power-flow modelling is only significantly influenced in a small subset of the network branches associated with major interconnections and flexible/inflexible conventional generation locations. Model sensitivity analysis also suggests that even at high wind penetrations, such power-flow modelling differences may be overshadowed by the impact of general uncertainty in fuel price volatility and demand profile that is systemic to long-term planning problems. Such trade-offs have significant practical relevance to the many researchers currently investigating formulations of this class of optimisation problem.

The rapid deployment of renewable sources of electricity (RES-E) is transforming power systems globally. This trend is likely to continue with large increases in investment and deployment of RES-E capacity over the coming decades. Several countries now have penetration levels of variable RES-E generation (i.e., wind and solar) in excess of 15% of their annual electricity generation; and many jurisdictions (e.g., Spain, Portugal, Ireland, Germany, and Denmark; and, in the United States, Colorado) have experienced instantaneous penetration levels of more than 50% variable generation.1 These penetration levels of variable RES-E have prompted many jurisdictions to begin modifying practices that evolved in an era of readily dispatchable, centralised power systems. Providing insights for the transition to high levels of variable RES-E generation is the focus of this document, which is the final report of the RES-E-NEXT project commissioned by the International Energy Agency’s implementing agreement on Renewable Energy Technology Deployment (IEA-RETD). It presents a comprehensive assessment of issues that will shape power system evolution during the transition to high levels of variable RES-E generation. While policy will be a central tool to sustain the growth of RES-E capacity and to enable power system transitions, the scope of the report extends beyond policy considerations to include the related domains of regulation, power market design, and system operation protocols. This broad scope is in recognition that a changing resource mix with greater penetration levels of variable RES-E has broad implications for grid operations, wholesale and retail power markets, and infrastructure needs. The next decade will be a critical transition period for power system stakeholders, as global deployment of RES-E capacity (and especially variable RES-E capacity) continues to scale-up in many regions of the world. To address increased penetration levels of RES-E in power systems and the new challenges that could emerge, coordinated portfolios of policies, market designs, regulations, and operational protocols are essential. The goal for policymakers is to facilitate investment in RES-E technologies and to enable efficient and reliable system operation, costeffective service delivery, and continued public acceptance. Although the factors that impact the speed and scale of RES-E deployment manifest uniquely in each power system, in the transition to high shares of variable RES-E this report identifies four critical domains and the changing drivers that will shape next-generation policy for each. These domains are introduced in Table I, and comprise the major sections of this report.

An artificial neural network algorithm for generator scheduling is proposed. The algorithm employs an unfeasible Lagrangian dual maximum solution to initialise the neurons of an augmented Hopfield network. The proposed algorithm produces cheaper solutions when compared with Lagrangian relaxation or a randomly initialised augmented Hopfield network. The algorithm also has shorter convergence times than the augmented Hopfield network, but is not as fast to converge as Lagrangian relaxation.

The augmented Lagrangian function has better convergence and solution properties for mixed integer non-linear programming problems than either standard Lagrangian function, or penalty function approaches. The Hopfield neural network based approaches that have been proposed to solve this function have used continuous neurons to represent discrete variables. The augmented Hopfield network has both discrete and continuous neurons. Here a Lagrangian augmented Hopfield network (LAHN) is constructed by including augmented Lagrangian multiplier neurons in the augmented Hopfield network. This new network is applied to the generator scheduling problem (a mixed integer non-linear programming problem) and results illustrate that improved solutions are obtained.

Power systems are typically scheduled at least cost subject to operational and security constraints. Generally, no account is taken of generator reliability when scheduling units. Also, the security criteria, which include reserve, are usually deterministic in nature. This paper proposes a method to consider generator reliability explicitly in the scheduling problem. A competitive structure is proposed that includes a market for reserve. This is formulated as an augmented Lagrangian dual function and is solved using a new recurrent neural network. The price for reserve is used, along with the unit reliability, to find a balance between the cost of reserve and the risk of not providing it.