Electricity sector models
Open electricity sector models are confined to just the electricity sector. These models invariably have a temporal resolution of one hour or less. Some models concentrate on the engineering characteristics of the system, including a good representation of high-voltage transmission networks and AC power flow. Others models depict electricity spot markets and are known as dispatch models. While other models embed autonomous agents to capture, for instance, bidding decisions using techniques from bounded rationality. The ability to handle variable renewable energy, transmission systems, and grid storage are becoming important considerations.
| Project | Host | License | Access | Coding | Documentation | Scope/type |
|---|---|---|---|---|---|---|
| AMIRIS | German Aerospace Center | Apache 2.0 | GitLab | Java | wiki | agent‑based electricity market modeling |
| Breakthrough Energy Model | Breakthrough Energy Foundation | MIT | GitHub | Python, Julia | website, GitHub | power sector modeling |
| DIETER | DIW Berlin | MIT | download | GAMS | publication | dispatch and investment |
| Dispa-SET | EC Joint Research Centre | EUPL 1.1 | GitHub | GAMS, Python | website | European transmission and dispatch |
| EMLab-Generation | Delft University of Technology | Apache 2.0 | GitHub | Java | manual, website | agent-based |
| EMMA | Neon Neue Energieökonomik | CC BY-SA 3.0 | download | GAMS | website | electricity market |
| GENESYS | RWTH Aachen University | LGPLv2.1 | on application | C++ | website | European electricity system |
| NEMO | University of New South Wales | GPLv3 | git repository | Python | website, list | Australian NEM market |
| OnSSET | KTH Royal Institute of Technology | MIT | GitHub | Python | website, GitHub | cost-effective electrification |
| pandapower | BSD-new | GitHub | Python | website | automated power system analysis | |
| PowerMatcher | Flexiblepower Alliance Network | Apache 2.0 | GitHub | Java | website | smart grid |
| Power TAC |
| Apache 2.0 | GitHub | Java | website, forum | automated retail electricity trading simulation |
| renpass | University of Flensburg | GPLv3 | by invitation | R, MySQL | manual | renewables pathways |
| SciGRID | DLR Institute of Networked Energy Systems | Apache 2.0 | git repository | Python | website, newsletter | European transmission grid |
| SIREN | Sustainable Energy Now | AGPLv3 | GitHub | Python | website | renewable generation |
| SWITCH | University of Hawaiʻi | Apache 2.0 | GitHub | Python | website | optimal planning |
| URBS | Technical University of Munich | GPLv3 | GitHub | Python | website | distributed energy systems |
| ||||||
AMIRIS
| Project | AMIRIS |
|---|---|
| Host | German Aerospace Center |
| Status | active |
| Scope/type | agent‑based electricity markets |
| Code license | Apache-2.0 |
| Data license | CC‑BY‑4.0 |
| Language | Java |
| Website | helmholtz |
| Repository | gitlab |
| Documentation | gitlab |
| Discussion | forum |
| Datasets | gitlab |
| Publications | zenodo |
AMIRIS is the open Agent-based Market model for the Investigation of Renewable and Integrated energy Systems. The AMIRIS simulation framework was first developed by the German Aerospace Center (DLR) in 2008 and later released as an open source project in 2021. [22]
AMIRIS enables researchers to address questions regarding future energy markets, their market design, and energy-related policy instruments. [23] In particular, AMIRIS is able to capture market effects that may arise from the integration of renewable energy sources and flexibility options by considering the strategies and behaviors of the various energy market actors present. For instance, those behaviors can be influenced by the prevailing political framework and by external uncertainties. [24] AMIRIS may also uncover complex effects that may emerge from the inter‑dependencies of the energy market participants. [25]
The embedded market clearing algorithm computes electricity prices based on the bids of prototyped market actors. These bids may not only reflect the marginal cost of electricity production but also the limited information available to the actors and related uncertainties. But also the bidding can be strategic as an attempt to game official support instruments or exploit market power opportunities.
Actors in AMIRIS are represented as agents that can be roughly divided into six classes: power plant operators, traders, market operators, policy providers, demand agents, and storage facility operators. In the model, power plant operators provide generation capacities to traders, but do not participate directly in markets. Instead, they supply traders who conduct the marketing and deploy bidding strategies on the operators behalf. Marketplaces serve as trading platforms and calculate market clearing. Policy providers define the regulatory framework which then may impact on the decisions of the other agents. Demand agents request energy directly at the electricity market. Finally, flexibility providers, such as storage operators, use forecasts to determine bidding patterns to match their particular objectives, for instance, projected profit maximization.
AMIRIS is based on the open Framework for distributed Agent-based Modelling of Energy systems or FAME. [26] [27]
AMIRIS can simulate large‑scale agent systems in acceptable timeframes. For instance, the simulation of one year at hourly resolution may take as little as one minute on a contemporary desktop computer. The researchers at DLR also have access to high-performance computing facilities.
Breakthrough Energy Model
| Project | Breakthrough Energy Model |
|---|---|
| Host | Breakthrough Energy Foundation |
| Status | active |
| Scope/type | power sector modeling |
| Code license | MIT |
| Data license | CC‑BY‑4.0 |
| Language | Python, Julia |
| Website | science |
| Repository | github |
| Documentation | breakthrough-energy |
The Breakthrough Energy Model is a production cost model with capacity expansion algorithms and heuristics, originally designed to explore the generation and transmission expansion needs to meet U.S. states' clean energy goals. The data management occurs within Python and the DCOPF optimization problem is created via Julia. The Breakthrough Energy Model is being developed by the Breakthrough Energy Sciences team.
The open data underlying the model builds upon the synthetic test cases developed by researchers at Texas A&M University. [28] [29] [30]
The Breakthrough Energy Model initially explored the generation and transmission expansion necessary to meet clean energy goals in 2030 via the building of a Macro Grid. [31] Ongoing work adds and expands modules to the model (e.g. electrification of buildings and transportation) to provide a framework for testing numerous scenario combinations. Development of and integration with other open-source data sets is in progress for modeling countries and regions beyond the United States.
The model was applied subsequently the 2021 Texas power crisis, in which winter power outages resulted in hundreds of deaths and billions of dollars in economic losses. [32] : 1
DIETER
| Project | DIETER |
|---|---|
| Host | DIW Berlin |
| Status | active |
| Scope/type | dispatch and investment |
| Code license | MIT |
| Data license | MIT |
| Language | GAMS |
| Website | www |
DIETER stands for Dispatch and Investment Evaluation Tool with Endogenous Renewables. DIETER is a dispatch and investment model. It was first used to study the role of power storage and other flexibility options in a future greenfield setting with high shares of renewable generation. DIETER is being developed at the German Institute for Economic Research (DIW), Berlin, Germany. The codebase and datasets for Germany can be downloaded from the project website. The basic model is fully described in a DIW working paper and a journal article. [33] [34] DIETER is written in GAMS and was developed using the CPLEX commercial solver.
DIETER is framed as a pure linear (no integer variables) cost minimization problem. In the initial formulation, the decision variables include the investment in and dispatch of generation, storage, and DSM capacities in the German wholesale and balancing electricity markets. Later model extensions include vehicle-to-grid interactions and prosumage of solar electricity. [35] [36]
The first study using DIETER examines the power storage requirements for renewables uptake ranging from 60% to 100%. Under the baseline scenario of 80% (the lower bound German government target for 2050), grid storage requirements remain moderate and other options on both the supply side and demand side offer flexibility at low cost. Nonetheless, storage plays an important role in the provision of reserves. Storage becomes more pronounced under higher shares of renewables, but strongly depends on the costs and availability of other flexibility options, particularly biomass availability. [37]
Dispa-SET
| Project | Dispa-SET |
|---|---|
| Host | EC Joint Research Centre |
| Status | active |
| Scope/type | European transmission and dispatch |
| Code license | EUPL 1.2 |
| Data license | CC‑BY‑4.0 |
| Website | www |
| Repository | github |
| Documentation | www |
Under development at the European Commission's Joint Research Centre (JRC), Petten, the Netherlands, Dispa-SET is a unit commitment and dispatch model intended primarily for Europe. It is written in Python (with Pyomo) and GAMS and uses Python for data processing. A valid GAMS license is required. The model is formulated as a mixed integer problem and JRC uses the proprietary CPLEX sover although open source libraries may also be deployed. Technical descriptions are available for versions 2.0 [38] and 2.1. [39] Dispa-SET is hosted on GitHub, together with a trial dataset, and third-party contributions are encouraged. The codebase has been tested on Windows, macOS, and Linux. Online documentation is available. [40]
The SET in the project name refers to the European Strategic Energy Technology Plan (SET-Plan), which seeks to make Europe a leader in energy technologies that can fulfill future (2020 and 2050) energy and climate targets. Energy system modeling, in various forms, is central to this European Commission initiative. [41]
The model power system is managed by a single operator with full knowledge of the economic and technical characteristics of the generation units, the loads at each node, and the heavily simplified transmission network. Demand is deemed fully inelastic. The system is subject to intra-period and inter-period unit commitment constraints (the latter covering nuclear and thermal generation for the most part) and operated under economic dispatch. [39] : 4 Hourly data is used and the simulation horizon is normally one year. But to ensure the model remains tractable, two day rolling horizon optimization is employed. The model advances in steps of one day, optimizing the next 48 hours ahead but retaining results for just the first 24 hours. [39] : 14–15
Two related publications describe the role and representation of flexibility measures within power systems facing ever greater shares of variable renewable energy (VRE). [42] [43] These flexibility measures comprise: dispatchable generation (with constraints on efficiency, ramp rate, part load, and up and down times), conventional storage (predominantly pumped-storage hydro), cross-border interconnectors, demand side management, renewables curtailment, last resort load shedding, and nascent power-to-X solutions (with X being gas, heat, or mobility). The modeler can set a target for renewables and place caps on CO2 and other pollutants. [39] Planned extensions to the software include support for simplified AC power flow [c] (transmission is currently treated as a transportation problem), new constraints (like cooling water supply), stochastic scenarios, and the inclusion of markets for ancillary services. [40]
Dispa-SET has been or is being applied to case studies in Belgium, Bolivia, Greece, Ireland, and the Netherlands. A 2014 Belgium study investigates what if scenarios for different mixes of nuclear generation, combined cycle gas turbine (CCGT) plant, and VRE and finds that the CCGT plants are subject to more aggressive cycling as renewable generation penetrates. [45]
A 2020 study investigated the collective impact of future climatic conditions on 34 European power systems, including potential variations in solar, wind, and hydro‑power output and electricity demand under various projected meteorological scenarios for the European continent. [46]
Dispa-SET has been applied in Africa with soft linking to the LISFLOOD model to examine water‑energy nexus problems in the context of a changing climate. [47]
EMLab-Generation
| Project | EMLab-Generation |
|---|---|
| Host | Delft University of Technology |
| Status | active |
| Scope/type | agent-based |
| Code license | Apache 2.0 |
| Website | emlab |
| Repository | github |
EMLab-Generation is an agent-based model covering two interconnected electricity markets – be they two adjoining countries or two groups of countries. The software is being developed at the Energy Modelling Lab, Delft University of Technology, Delft, the Netherlands. A factsheet is available. [48] And software documentation is available. [49] EMLab-Generation is written in Java.
EMLab-Generation simulates the actions of power companies investing in generation capacity and uses this to explore the long-term effects of various energy and climate protection policies. These policies may target renewable generation, CO2 emissions, security of supply, and/or energy affordability. The power companies are the main agents: they bid into power markets and they invest based on the net present value (NPV) of prospective power plant projects. They can adopt a variety of technologies, using scenarios from the 2011 IEA World Energy Outlook. [50] The agent-based methodology enables different sets of assumptions to be tested, such as the heterogeneity of actors, the consequences of imperfect expectations, and the behavior of investors outside of ideal conditions.
EMLab-Generation offers a new way of modeling the effects of public policy on electricity markets. It can provide insights into actor and system behaviors over time – including such things as investment cycles, abatement cycles, delayed responses, and the effects of uncertainty and risk on investment decisions.
A 2014 study using EMLab-Generation investigates the effects of introducing floor and ceiling prices for CO2 under the EU ETS. And in particular, their influence on the dynamic investment pathway of two interlinked electricity markets (loosely Great Britain and Central Western Europe). The study finds a common, moderate CO2 auction reserve price results in a more continuous decarbonisation pathway and reduces CO2 price volatility. Adding a ceiling price can shield consumers from extreme price shocks. Such price restrictions should not lead to an overshoot of emissions targets in the long-run. [51]
EMMA
| Project | EMMA |
|---|---|
| Host | Neon Neue Energieökonomik |
| Status | active |
| Scope/type | electricity market |
| Code license | CC BY-SA 3.0 |
| Data license | CC BY-SA 3.0 |
| Website | neon-energie |
EMMA is the European Electricity Market Model. It is a techno-economic model covering the integrated Northwestern European power system. EMMA is being developed by the energy economics consultancy Neon Neue Energieökonomik, Berlin, Germany. The source code and datasets can be downloaded from the project website. A manual is available. [52] EMMA is written in GAMS and uses the CPLEX commercial solver.
EMMA models electricity dispatch and investment, minimizing the total cost with respect to investment, generation, and trades between market areas. In economic terms, EMMA classifies as a partial equilibrium model of the wholesale electricity market with a focus on the supply-side. EMMA identifies short-term or long-term optima (or equilibria) and estimates the corresponding capacity mix, hourly prices, dispatch, and cross-border trading. Technically, EMMA is a pure linear program (no integer variables) with about two million non-zero variables. As of 2016 [update] , the model covers Belgium, France, Germany, the Netherlands, and Poland and supports conventional generation, renewable generation, and cogeneration. [52] [53]
EMMA has been used to study the economic effects of the increasing penetration of variable renewable energy (VRE), specifically solar power and wind power, in the Northwestern European power system. A 2013 study finds that increasing VRE shares will depress prices and, as a consequence, the competitive large-scale deployment of renewable generation will be more difficult to accomplish than many anticipate. [54] A 2015 study estimates the welfare-optimal market share for wind and solar power. For wind, this is 20%, three-fold more than at present. [55]
An independent 2015 study reviews the EMMA model and comments on the high assumed specific costs for renewable investments. [33] : 6
GENESYS
| Project | GENESYS |
|---|---|
| Host | RWTH Aachen University |
| Status | active |
| Scope/type | European electricity system |
| Code license | LGPLv2.1 |
| Data license | LGPLv2.1 |
| Language | C++ |
| Website | www |
GENESYS stands for Genetic Optimisation of a European Energy Supply System. The software is being developed jointly by the Institute of Power Systems and Power Economics (IAEW) and the Institute for Power Electronics and Electrical Drives (ISEA), both of RWTH Aachen University, Aachen, Germany. The project maintains a website where potential users can request access to the codebase and the dataset for the 2050 base scenario only. [56] Detailed descriptions of the software are available. [57] [58] GENESYS is written in C++ and uses Boost libraries, the MySQL relational database, the Qt 4 application framework, and optionally the CPLEX solver.
The GENESYS simulation tool is designed to optimize a future EUMENA (Europe, Middle East, and North Africa) power system and assumes a high share of renewable generation. It is able to find an economically optimal distribution of generator, storage, and transmission capacities within a 21 region EUMENA. It allows for the optimization of this energy system in combination with an evolutionary method. The optimization is based on a covariance matrix adaptation evolution strategy (CMA-ES), while the operation is simulated as a hierarchical set-up of system elements which balance the load between the various regions at minimum cost using the network simplex algorithm. GENESYS ships with a set of input time series and a set of parameters for the year 2050, which the user can modify.
A future EUMENA energy supply system with a high share of renewable energy sources (RES) will need a strongly interconnected energy transport grid and significant energy storage capacities. GENESYS was used to dimension the storage and transmission between the 21 different regions. Under the assumption of 100% self-supply, about 2500 GW of RES in total and a storage capacity of about 240000 GWh are needed, corresponding to 6% of the annual energy demand, and a HVDC transmission grid of 375000 GW·km. The combined cost estimate for generation, storage, and transmission, excluding distribution, is 6.87 ¢/kWh. [57]
A 2016 study looked at the relationship between storage and transmission capacity under high shares of renewable energy sources (RES) in an EUMENA power system. It found that, up to a certain extent, transmission capacity and storage capacity can substitute for each other. For a transition to a fully renewable energy system by 2050, major structural changes are required. The results indicate the optimal allocation of photovoltaics and wind power, the resulting demand for storage capacities of different technologies (battery, pumped hydro, and hydrogen storage) and the capacity of the transmission grid. [58]
NEMO
| Project | NEMO |
|---|---|
| Host | University of New South Wales |
| Status | active |
| Scope/type | Australian NEM market |
| Code license | GPLv3 |
| Language | Python |
| Website | nemo |
| Repository | github |
| Documentation | nbviewer |
NEMO, the National Electricity Market Optimiser, is a chronological dispatch model for testing and optimizing different portfolios of conventional and renewable electricity generation technologies. It applies solely to the Australian National Electricity Market (NEM), which, despite its name, is limited to east and south Australia. NEMO has been in development at the Centre for Energy and Environmental Markets (CEEM), University of New South Wales (UNSW), Sydney, Australia since 2011. [59] The project maintains a small website and runs an email list. NEMO is written in Python. NEMO itself is described in two publications. [60] : sec 2 [61] : sec 2 The data sources are also noted. [60] : sec 3 Optimizations are carried out using a single-objective evaluation function, with penalties. The solution space of generator capacities is searched using the CMA-ES (covariance matrix adaptation evolution strategy) algorithm. The timestep is arbitrary but one hour is normally employed.
NEMO has been used to explore generation options for the year 2030 under a variety of renewable energy (RE) and abated fossil fuel technology scenarios. A 2012 study investigates the feasibility of a fully renewable system using concentrated solar power (CSP) with thermal storage, windfarms, photovoltaics, existing hydroelectricity, and biofuelled gas turbines. A number of potential systems, which also meet NEM reliability criteria, are identified. The principal challenge is servicing peak demand on winter evenings following overcast days and periods of low wind. [60] A 2014 study investigates three scenarios using coal-fired thermal generation with carbon capture and storage (CCS) and gas-fired gas turbines with and without capture. These scenarios are compared to the 2012 analysis using fully renewable generation. The study finds that "only under a few, and seemingly unlikely, combinations of costs can any of the fossil fuel scenarios compete economically with 100% renewable electricity in a carbon constrained world". [62] : 196 A 2016 study evaluates the incremental costs of increasing renewable energy shares under a range of greenhouse gas caps and carbon prices. The study finds that incremental costs increase linearly from zero to 80% RE and then escalate moderately. The study concludes that this cost escalation is not a sufficient reason to avoid renewables targets of 100%. [61]
OnSSET
| Project | OnSSET |
|---|---|
| Host | KTH Royal Institute of Technology |
| Status | active |
| Scope/type | cost-effective electrification |
| Code license | MIT |
| Website | www |
| Mailing list | groups |
| Repository | github |
| Documentation | onsset-manual |
| Datasets | energydata |
OnSSET is the OpeN Source Spatial Electrification Toolkit. OnSSET is being developed by the division of Energy Systems, KTH Royal Institute of Technology, Stockholm, Sweden. The software is used to examine areas not served by grid-based electricity and identify the technology options and investment requirements that will provide least-cost access to electricity services. OnSSET is designed to support the United Nations' SDG 7: the provision of affordable, reliable, sustainable, and modern energy for all. The toolkit is known as OnSSET and was released on 26 November 2016. OnSSET does not ship with data, but suitable datasets are available from energydata.info. The project maintains a website and runs a mailing list. [63] [64] [65]
OnSSET can estimate, analyze, and visualize the most cost-effective electrification access options, be they conventional grid, mini-grid, or stand-alone. [66] The toolkit supports a range of conventional and renewable energy technologies, including photovoltaics, wind turbines, and small hydro generation. As of 2017 [update] , bioenergy and hybrid technologies, such as wind-diesel, are being added.
OnSSET utilizes energy and geographic information, the latter may include settlement size and location, existing and planned transmission and generation infrastructure, economic activity, renewable energy resources, roading networks, and nighttime lighting needs. The GIS information can be supported using the proprietary ArcGIS package or an open source equivalent such as GRASS or QGIS. [67] OnSSET has been applied to microgrids using a three‑tier analysis starting with settlement archetypes. [68]
OnSSET has been used for case studies in Afghanistan, [69] Bolivia, [68] [70] Cameroon, [71] Ethiopia, [66] [72] Malawi, [73] Nigeria, [66] [74] [75] and Tanzania. [67] OnSSET has also been applied in India, Kenya, and Zimbabwe. In addition, continental studies have been carried out for Sub-Saharan Africa and Latin America. [76] A 4‑way GIS ‑based study set in Nigeria reported that OnSSET offered the best set of capabilities. [77]
OnSSET results have contributed to the IEA World Energy Outlook reports for 2014 [78] and 2015, [79] the World Bank Global Tracking Framework report in 2015, [80] and the IEA Africa Energy Outlook report in 2019. [81] OnSSET also forms part of the nascent GEP platform. [82]
pandapower
| Project | pandapower |
|---|---|
| Host |
|
| Status | active |
| Scope/type | automated power system analysis |
| Code license | BSD-new |
| Website | www |
| Repository | github |
| Python package | pypi |
| Documentation | pandapower |
| Discussion | forum |
pandapower is a power system analysis and optimization program being jointly developed by the Energy Management and Power System Operation research group, University of Kassel and the Department for Distribution System Operation, Fraunhofer Institute for Energy Economics and Energy System Technology (IEE), both of Kassel, Germany. The codebase is hosted on GitHub and is also available as a package. The project maintains a website, an emailing list, and online documentation. pandapower is written in Python. It uses the pandas library for data manipulation and analysis and the PYPOWER library [83] to solve for power flow. Unlike some open source power system tools, pandapower does not depend on proprietary platforms like MATLAB.
pandapower supports the automated analysis and optimization of distribution and transmission networks. This allows a large of number of scenarios to be explored, based on different future grid configurations and technologies. pandapower offers a collection of power system elements, including: lines, 2-winding transformers, 3-winding transformers, and ward-equivalents. It also contains a switch model that allows the modeling of ideal bus-bus switches as well as bus-line/bus-trafo switches. The software supports topological searching. The network itself can be plotted, with or without geographical information, using the matplotlib and plotly libraries.
A 2016 publication evaluates the usefulness of the software by undertaking several case studies with major distribution system operators (DSO). These studies examine the integration of increasing levels of photovoltaics into existing distribution grids. The study concludes that being able to test a large number of detailed scenarios is essential for robust grid planning. Notwithstanding, issues of data availability and problem dimensionality will continue to present challenges. [84]
A 2018 paper describes the package and its design and provides an example case study. The article explains how users work with an element-based model (EBM) which is converted internally to a bus-branch model (BBM) for computation. The package supports power system simulation, optimal power flow calculations (cost information is required), state estimation (should the system characterization lacks fidelity), and graph-based network analysis. The case study shows how a few tens of lines of scripting can interface with pandapower to advance the design of a system subject to diverse operating requirements. The associated code is hosted on GitHub as jupyter notebooks. [85]
As of 2018 [update] , BNetzA, the German network regulator, is using pandapower for automated grid analysis. [86] Energy research institutes in Germany are also following the development of pandapower. [87] : 90
PowerMatcher
| Project | PowerMatcher |
|---|---|
| Host | Flexiblepower Alliance Network |
| Status | active |
| Scope/type | smart grid |
| Code license | Apache 2.0 |
| Website | flexiblepower |
| Repository | github |
The PowerMatcher software implements a smart grid coordination mechanism which balances distributed energy resources (DER) and flexible loads through autonomous bidding. The project is managed by the Flexiblepower Alliance Network (FAN) in Amsterdam, the Netherlands. The project maintains a website and the source code is hosted on GitHub. As of June 2016 [update] , existing datasets are not available. PowerMatcher is written in Java.
Each device in the smart grid system – whether a washing machine, a wind generator, or an industrial turbine – expresses its willingness to consume or produce electricity in the form of a bid. These bids are then collected and used to determine an equilibrium price. The PowerMatcher software thereby allows high shares of renewable energy to be integrated into existing electricity systems and should also avoid any local overloading in possibly aging distribution networks. [88]
Power TAC
| Project | Power TAC |
|---|---|
| Host | Erasmus Centre for Future Energy Business at the Rotterdam School of Management, Erasmus University |
| Status | active |
| Scope/type | automated retail electricity trading simulation |
| Code license | Apache 2.0 |
| Website | powertac |
Power TAC stands for Power Trading Agent Competition. Power TAC is an agent-based model simulating the performance of retail markets in an increasingly prosumer- and renewable-energy-influenced electricity landscape. The first version of the Power TAC project started in 2009, when the open source platform was released as an open-source multi-agent competitive gaming platform to simulate electricity retail market performance in smart grid scenarios. The inaugural annual tournament was held in Valencia, Spain in 2012.
Autonomous machine-learning trading agents, or 'brokers', compete directly with each other as profit-maximizing aggregators between wholesale markets and retail customers. Customer models represent households, small and large businesses, multi-residential buildings, wind parks, solar panel owners, electric vehicle owners, cold-storage warehouses, etc. Brokers aim at making profit through offering electricity tariffs to customers and trading electricity in the wholesale market, while carefully balancing supply and demand.
The competition is founded and orchestrated by Professors Wolfgang Ketter and John Collins and the platform software is developed collaboratively by researchers at the Rotterdam School of Management, Erasmus University Centre for Future Energy Business, the Institute for Energy Economics at the University of Cologne, and the Computer Science department at the University of Minnesota. The platform uses a variety of real-world data about weather, market prices and aggregate demand, and customer behavior. Broker agents are developed by research teams around the world and entered in annual tournaments. Data from those tournaments are publicly available and can be used to assess agent performance and interactions. The platform exploits competitive benchmarking to facilitate research into, among other topics, tariff design in retail electricity markets, bidding strategies in wholesale electricity markets, performance of markets as penetration of sustainable energy resources or electric vehicles is ramped up or down, effectiveness of machine learning approaches, and alternative policy approaches to market regulation. The software has contributed to research topics ranging from the use of electric vehicle fleets as virtual power plants to how an electricity customer decision support system (DSS) can be used to design effective demand response programs using methods such as dynamic pricing.
renpass
| Project | renpass |
|---|---|
| Host | University of Flensburg |
| Status | inactive |
| Scope/type | renewables pathways |
| Code license | GPLv3 |
| Website | github |
| Repository | github |
renpass is an acronym for Renewable Energy Pathways Simulation System. renpass is a simulation electricity model with high regional and temporal resolution, designed to capture existing systems and future systems with up to 100% renewable generation. The software is being developed by the Centre for Sustainable Energy Systems (CSES or ZNES), University of Flensburg, Germany. The project runs a website, from where the codebase can be download. renpass is written in R and links to a MySQL database. A PDF manual is available. [89] renpass is also described in a PhD thesis. [90] As of 2015 [update] , renpass is being extended as renpassG!S, based on oemof.
renpass is an electricity dispatch model which minimizes system costs for each time step (optimization) within the limits of a given infrastructure (simulation). Time steps are optionally 15 minutes or one hour. The method assumes perfect foresight. renpass supports the electricity systems found in Austria, Belgium, the Czech Republic, Denmark, Estonia, France, Finland, Germany, Latvia, Lithuania, Luxembourg, the Netherlands, Norway, Poland, Sweden, and Switzerland.
The optimization problem for each time step is to minimize the electricity supply cost using the existing power plant fleet for all regions. After this regional dispatch, the exchange between the regions is carried out and is restricted by the grid capacity. This latter problem is solved with a heuristic procedure rather than calculated deterministically. The input is the merit order, the marginal power plant, the excess energy (renewable energy that could be curtailed), and the excess demand (the demand that cannot be supplied) for each region. The exchange algorithm seeks the least cost for all regions, thus the target function is to minimize the total costs of all regions, given the existing grid infrastructure, storage, and generating capacities. The total cost is defined as the residual load multiplied by the price in each region, summed over all regions.
A 2012 study uses renpass to examine the feasibility of a 100% renewable electricity system for the Baltic Sea region (Denmark, Estonia, Finland, Germany, Latvia, Lithuania, Poland, and Sweden) in the year 2050. The base scenario presumes conservative renewable potentials and grid enhancements, a 20% drop in demand, a moderate uptake of storage options, and the deployment of biomass for flexible generation. The study finds that a 100% renewable electricity system is possible, albeit with occasional imports from abutting countries, and that biomass plays a key role in system stability. The costs for this transition are estimated at 50 €/MWh. [91] A 2014 study uses renpass to model Germany and its neighbors. [92] A 2014 thesis uses renpass to examine the benefits of both a new cable between Germany and Norway and new pumped storage capacity in Norway, given 100% renewable electricity systems in both countries. [93] Another 2014 study uses renpass to examine the German Energiewende , the transition to a sustainable energy system for Germany. The study also argues that the public trust needed to underpin such a transition can only be built through the use of transparent open source energy models. [94]
SciGRID
| Project | SciGRID |
|---|---|
| Host | Deutsches Zentrum für Luft- und Raumfahrt |
| Status | active |
| Scope/type | European transmission grid |
| Code license | Apache 2.0 |
| Website | www |
SciGRID, short for Scientific Grid, is an open source model of the German and European electricity transmission networks. The research project is managed by DLR Institute of Networked Energy Systems located in Oldenburg, Germany. The project maintains a website and an email newsletter. SciGRID is written in Python and uses a PostgreSQL database. The first release (v0.1) was made on 15 June 2015.
SciGRID aims to rectify the lack of open research data on the structure of electricity transmission networks within Europe. This lack of data frustrates attempts to build, characterise, and compare high resolution energy system models. SciGRID utilizes transmission network data available from the OpenStreetMap project, available under the Open Database License (ODbL), to automatically author transmission connections. SciGRID will not use data from closed sources. SciGRID can also mathematically decompose a given network into a simpler representation for use in energy models. [95] [96]
SIREN
| Project | SIREN |
|---|---|
| Host | Sustainable Energy Now |
| Status | active |
| Scope/type | renewable generation |
| Code license | AGPLv3 |
| Website | www |
| Repository | sourceforge |
SIREN stands for SEN Integrated Renewable Energy Network Toolkit. The project is run by Sustainable Energy Now, an NGO based in Perth, Australia. The project maintains a website. SIREN runs on Windows and the source code is hosted on SourceForge. The software is written in Python and uses the SAM model (System Advisor Model) from the US National Renewable Energy Laboratory to perform energy calculations. SIREN uses hourly datasets to model a given geographic region. Users can use the software to explore the location and scale of renewable energy sources to meet a specified electricity demand. SIREN utilizes a number of open or publicly available data sources: maps can be created from OpenStreetMap tiles and weather datasets can be created using NASA MERRA-2 satellite data. [d] [97]
A 2016 study using SIREN to analyze Western Australia's South-West Interconnected System (SWIS) finds that it can transition to 85% renewable energy (RE) for the same cost as new coal and gas. In addition, 11.1 million tonnes of CO2eq emissions would be avoided. The modeling assumes a carbon price of AUD $30/tCO2. Further scenarios examine the goal of 100% renewable generation. [98]
SWITCH
| Project | SWITCH |
|---|---|
| Host | University of Hawaiʻi |
| Status | active |
| Scope/type | optimal planning |
| Code license | Apache 2.0 |
| Website | switch-model |
| Repository | github |
SWITCH is a loose acronym for solar, wind, conventional and hydroelectric generation, and transmission. SWITCH is an optimal planning model for power systems with large shares of renewable energy. SWITCH is being developed by the Department of Electrical Engineering, University of Hawaiʻi at Mānoa, Hawaii, USA. The project runs a small website and hosts its codebase and datasets on GitHub. SWITCH is written in Pyomo, an optimization components library programmed in Python. It can use either the open source GLPK solver or the commercial CPLEX solver.
SWITCH is a power system model, focused on renewables integration. It can identify which generator and transmission projects to build in order to satisfy electricity demand at the lowest cost over a several-year period while also reducing CO2 emissions. SWITCH utilizes multi-stage stochastic linear optimization with the objective of minimizing the present value of the cost of power plants, transmission capacity, fuel usage, and an arbitrary per-tonne CO2 charge (to represent either a carbon tax or a certificate price), over the course of a multi-year investment period. It has two major sets of decision variables. First, at the start of each investment period, SWITCH selects how much generation capacity to build in each of several geographic load zones, how much power transfer capability to add between these zones, and whether to operate existing generation capacity during the investment period or to temporarily mothball it to avoid fixed operation and maintenance costs. Second, for a set of sample days within each investment period, SWITCH makes hourly decisions about how much power to generate from each dispatchable power plant, store at each pumped hydro facility, or transfer along each transmission interconnector. The system must also ensure enough generation and transmission capacity to provide a planning reserve margin of 15% above the load forecasts. For each sampled hour, SWITCH uses electricity demand and renewable power production based on actual measurements, so that the weather-driven correlations between these elements remain intact.
Following the optimization phase, SWITCH is used in a second phase to test the proposed investment plan against a more complete set of weather conditions and to add backstop generation capacity so that the planning reserve margin is always met. Finally, in a third phase, the costs are calculated by freezing the investment plan and operating the proposed power system over a full set of weather conditions.
A 2012 paper uses California from 2012 to 2027 as a case study for SWITCH. The study finds that there is no ceiling on the amount of wind and solar power that could be used and that these resources could potentially reduce emissions by 90% or more (relative to 1990 levels) without reducing reliability or severely raising costs. Furthermore, policies that encourage electricity customers to shift demand to times when renewable power is most abundant (for example, though the well-timed charging of electric vehicles) could achieve radical emission reductions at moderate cost. [99]
SWITCH was used more recently to underpin consensus-based power system planning in Hawaii. [100] The model is also being applied in Chile, Mexico, and elsewhere. [101]
Major version 2.0 was released in late‑2018. [101] An investigation that year favorably compared SWITCH with the proprietary General Electric MAPS model using Hawaii as a case study. [102]
URBS
| Project | URBS |
|---|---|
| Host | Technical University of Munich |
| Status | active |
| Scope/type | distributed energy systems |
| Code license | GPLv3 |
| Repository | github |
URBS, Latin for city, is a linear programming model for exploring capacity expansion and unit commitment problems and is particularly suited to distributed energy systems (DES). It is being developed by the Institute for Renewable and Sustainable Energy Systems, Technical University of Munich, Germany. The codebase is hosted on GitHub. URBS is written in Python and uses the Pyomo optimization packages.
URBS classes as an energy modeling framework and attempts to minimize the total discounted cost of the system. A particular model selects from a set of technologies to meet a predetermined electricity demand. It uses a time resolution of one hour and the spatial resolution is model-defined. The decision variables are the capacities for the production, storage, and transport of electricity and the time scheduling for their operation. [103] : 11–14
The software has been used to explore cost-optimal extensions to the European transmission grid using projected wind and solar capacities for 2020. A 2012 study, using high spatial and technological resolutions, found variable renewable energy (VRE) additions cause lower revenues for conventional power plants and that grid extensions redistribute and alleviate this effect. [104] The software has also been used to explore energy systems spanning Europe, the Middle East, and North Africa (EUMENA) [103] and Indonesia, Malaysia, and Singapore. [105]
