Vehicle-grid integration is a relatively new field of R&D with many open questions for all stakeholders, including those in the automotive market, the electricity market, policy makers, and end users of PEVs. V2G-Sim provides systematic quantitative methods to address the uncertainties and barriers facing vehicle-grid integration for all of these stakeholders.

The long term success of PEVs and of vehicle-grid integration depends on the collective efforts of researchers, engineers, policymakers, and end users amongst each group of stakeholders. V2G-Sim provides a common platform for all of these stakeholders to quantitatively understand the challenges facing PEVs and vehicle-grid integration, systematically develop solutions, and determine the effectiveness of a given approach to ensure that optimal solutions are identified prior to expensive deployment projects. V2G-Sim incorporates the following key features to the benefit of users in each group of stakeholders.

Understanding the impacts of vehicle-grid integration on vehicles, on the grid, and for end users requires stakeholders to simultaneously consider many interconnected variables. For instance, in the interconnected transportation and electricity market of the future, when and where a person drives their car will impact the operation of the electricity grid. V2G-Sim provides stakeholders with a well developed, detailed, and validated platform which they can leverage towards their individual application. Electricity utilities, for example, can understand distribution systems impacts without having to develop their own models for how people drive and how vehicles consume energy and charge. Automakers can study their individual powertrains and batteries without having to develop their own drive cycles, driver behavior models, or grid integration models. V2G-Sim includes automated methods to initialize each validated sub-model, allowing users to quickly produce results without substantial time investment in developing and validating models.

The simplicity in using V2G-Sim lies in the included databases of vehicle usage profiles by drivers, drive cycles, powertrain models, charging models, and battery degradation models. However, each component within V2G-Sim is fully customizable. Users can create and run their own vehicle powertrain models, charging control models, battery models, etc. The statistical input methods built into V2G-Sim allow users to quickly generate vehicle usage profiles for small or large fleets of vehicles (from 1 or 2 vehicles, to 1 or 2 million vehicles) to study vehicle-grid integration. The input statistics can be easily changed to allow users to quantify how individual vehicles or aggregate grid impacts will change under any scenario (e.g. vehicles charging in the evening at home only vs. vehicles that charge at home and at work). Individual vehicles in a simulation can be modeled as any powertrain or vehicle type and can be instructed to charge or discharge to the electricity grid in different ways. Users can design and implement customized models that consider driver behavior, pricing and economics, utility rate structures, charger availability, or any other external factor.

Depending on the research and development objectives for a given V2G-Sim user, detailed models can be activated or deactivated and replaced with built-in simplified models that are calibrated against detailed models. These simplified models enable substantial computational efficiency. For instance, V2G-Sim can predict the vehicle-grid energy interactions for 24 hours of operation of ~130,000 vehicles within ~1.5 hours on a single CPU core of a laptop computer. For each of these ~130,000 vehicles, the vehicle-grid interactions and SOC profile is resolved for each individual vehicle within this computing time. Users can choose to output all data to results files with second-by-second time steps, or choose to output at fewer time steps to limit the file size of the output results.

This computational efficiency enables users to carry out large numbers of parameter sweeps for simulations with hundreds of thousands of vehicles, enabling research and development on vehicle-grid integration at an unprecedented scale. Alternatively, these fast simulations can be used in providing coarse answers to a given question prior to activating detailed models.

V2G-Sim is inherently scalable in the number of vehicles modeled. Users can initialize and run models of a few hundred vehicles just as easily as simulations of a few hundreds of thousands of vehicles. The granularity of spatial resolution of vehicle-grid impacts is also scalable based on the data provided to resolve vehicle locations, for instance V2G-Sim can support spatial resolution at the state or city level just as easily as at the neighborhood level.

As the number of vehicles simulated increases, or the spatial granularity in desired results increases, the computational time for a V2G-Sim simulation also increases. Although significant effort has been dedicated to enabling computational efficiency, some users may seek simulations of large numbers of vehicles with all of the detailed models activated. For these specialized applications, LBNL is working towards releasing a parallel processing version of V2G-Sim which can be run on multiple cores of a single computer or on a high performance computing cluster. A pilot project is underway to test and deploy V2G-Sim on the National Energy Research Scientific Computing Cluster (NERSC).

V2G-Sim’s systematic and quantitative methodology provides predictive information to address the uncertainties of many stakeholders. Some key examples are:

Reliability of the electricity grid requires careful control to ensure that adequate supply is always available to match electricity demand. A critical function of an electricity system operator is to forecast how much electricity demand there will be at different times of the day. As adoption of plug-in vehicles (PEVs) grows, electricity system operators must reliably forecast the power demand from charging PEVs. Forecasting PEV charging, however, requires knowledge of what types of PEVs people drive, how people will use their cars, how far they will travel, at what times, etc. Each of these factors can be difficult to predict and includes uncertainties, as each individual uses their car differently and has the freedom to make unexpected trips. V2G-Sim considers all of these factors to forecast PEV loads on a second-by-second timescale and also quantify the uncertainty in the load forecasts.

Additionally, PEVs can provide a valuable service to the electricity grid by using their large batteries as energy storage for the grid. For PEV batteries to be used as part of the electricity market, many PEVs must be part of an aggregation system that issues control signals instructing how each vehicle should charge or discharge to respond to electricity grid requirements. The aggregator bids some capacity from the collection of EVs into the electricity market for different times of the day. The aggregator must not only ensure that it has adequate capacity to satisfy its bids amongst all the EVs in its system, but must also ensure that each PEV is charged when it needs to be. To reliably perform this service, the aggregator needs dependable forecasts of the charging needs of each PEV, the flexibility of each PEV to delay charging or discharge to provide a grid service. V2G-Sim provides this information along with uncertainty estimates enabling aggregators to plan their bids and understand the risks associated with each bid.

Vehicle batteries degrade with time and usage. A perceived challenge facing the use of vehicle batteries for providing grid services is that these grid services will lead to accelerated degradation of vehicle batteries. V2G-Sim’s coupled powertrain models, charging models, and battery degradation models enable automakers and battery manufacturers to understand battery degradation from driving and from grid services. V2G-Sim users can quantify the battery degradation that results from driving vehicles versus the degradation from any grid service that is modeled using the customizable managed charging and discharging control models.

Although V2G-Sim includes built-in capacity fade models, batteries experience many different forms of degradation and each battery chemistry and design degrades differently. V2G-Sim enables researchers, automakers, and battery manufacturers to apply their own degradation models specific to their own cell chemistry and design. V2G-Sim can then be used to study degradation for a wide variety of driving patterns, in a variety of vehicle powertrain types, providing a variety of different grid services.

The limited energy storage capabilities of today’s EV batteries lead to range anxiety amongst drivers. Before choosing to purchase an EV, drivers need to know whether the EV will meet their individual needs, and whether they must invest in specialized charging stations at their homes or workplaces. V2G-Sim coupled with the upcoming V2G-Sim mobile app allows drivers to measure their trips to create customized drive cycles for several weeks of their mobility. Coupled with the V2G-Sim server, the mobile app enables drivers to predict the battery state-of-charge profile for any EV they are considering purchasing to identify whether a given EV model will meet their needs. On the mobile app, drivers can choose the locations where their EV will be charged and the type of charger the EV will be plugged into to identify whether simple 120 V wall outlets satisfy their needs or whether they must invest in dedicated charging stations for locations where the EV will be parked.