Annex 60 will develop and demonstrate new generation computational tools for building and community energy systems based on the non-proprietary Modelica modeling language and Functional Mockup Interface (FMI) standards. The anticipated outcomes are open-source, freely available, documented, validated and verified computational tools that allow buildings, building systems and community energy grids to be designed and operated as integrated, robust, performance based systems with low energy use and low peak power demand. The target audience is the building energy research community, design firms and energy service companies, equipment and tool manufacturers, as well as students in building energy-related sciences. Currently fragmented duplicative activities in modeling, simulation and optimization of building and community energy systems that are based on the Modelica and FMI standards will be coordinated. Tool-chains will be created and validated that link Building Information Models to energy modeling, building simulation to controls design tools, and design tools to operational tools. Invention and deployment of integrated energy-related systems and performance-based solutions for buildings and communities will be accelerated by extending, unifying and documenting existing Modelica libraries, and by providing technical capabilities to link existing building performance simulation tools with such libraries and with control systems through the Functional Mockup Interface standard. Demonstrations will include optimized design and operation of building and community energy systems.
Description of Technical Sector¶
Modeling and Simulation Needs for Very Low Energy Buildings and Community Systems¶
As buildings become increasingly integrated to reduce energy and peak power and to increase occupant health and productivity, new challenges are posed to engineers when using building simulation programs to support decision making during product development, building design, commissioning and operation. New requirements that were not yet recognized 20 to 40 years ago when the development of current building simulation programs started include:
- Model-based design of integrated building systems by design firms and of products by equipment and controls providers to optimize energy-efficiency and peak load, and to reduce time-to-market for components, systems and advanced control systems.
- Design processes based on a Building Information Model (BIM) which become increasingly used by design firms.
- Integrated design of building envelope, HVAC systems and control strategies by design firms.
- Model use to support operation, for control providers as part of an energy or smart-grid aware controller, for commissioning agents to provide a reference for the expected building operation, for fault detection and diagnostics providers to provide a reference model that can be used to classify fault signatures, and for urban planners and utility companies to develop design and operation strategies for energy grids with dispatchable distributed loads, generation and storage.
These applications require the integration of multiple domains (air-flow, thermodynamics, controls, indoor environmental quality, and electrical grid) and multiple disciplines (HVAC/energy consultant, architect, controls engineer, electrical engineer) that use a variety of tools that represent building systems across largely varying time scales from seconds to years and length scales from building components to urban districts. Next to the mentioned engineering consultancy services, the technology is as well important to the building simulation research community to build and deploy their R&D through standardized tools.
Figure 1 shows these new challenges that will, in Annex 60, be addressed through the use of a standardized modeling language, standardized Application Programming Interface (API) and standardized data models. In contrast to these new computational technologies, today’s building simulation programs were primarily designed as tools for energy analysis. Use of these programs during operation is cumbersome, as is their integration with models from other tools, or with models created by other team members or discipline experts. Furthermore, in today’s building simulation programs, operational sequences are highly idealized. This prevents these tools from being used for the verification of the proper design and implementation of control sequences. It also makes it difficult to use the tools to analyze existing buildings which may have unconventional systems and control logic, because the original systems may have been retrofitted and buildings may have been re-purposed. Furthermore, current building energy performance simulation tools are not designed for the overall investigation of the interaction between buildings, district heating, cooling and electricity systems at the same time, which is required to evaluate new energy concepts that exploit dispatchable, distributed energy storage, generation and loads.
Modeling and Simulation Technologies that will be leveraged in this Annex¶
Modeling and simulation technologies that will be leveraged in this Annex are based on the object-oriented modeling language Modelica. The Modelica Standard Library contains more than 1300 models and functions that are open-source, freely available and well documented. However, the Modelica Standard Library does not have models for building energy systems, nor was the development of such models part of three European projects which were conducted with 54 million Euros investments in the automotive, aerospace and chemical industry. Moreover, current Modelica developments in the building sector are not coordinated. Consequently, model libraries lack common interfaces and typically cannot be linked with each other.
The need to couple legacy code to Modelica is clearly recognized. The Functional Mockup Interface standardizes the application programming interface for inclusion of simulation models into other simulation programs. It also standardizes how different simulation programs can communicate with each other during run-time. Modelica and the Functional Mockup Interface (FMI) standard have been selected as they are industrial-strength, non-proprietary, industry-driven standards that allow technology transfer between the building performance simulation community and much larger dynamic modeling communities from controls, automotive, power-plant, electrical system and chemical plant modeling.
Rather than reinventing the wheel, a major focus of this Annex will be how these investments, that are orders of magnitudes larger than what is invested in any single building simulation program, can be leveraged and extended where needed in a way that coordinates the activities of several countries. The Annex project therefore does not focus on the development of a single tool. At the same time, with the FMI technology, existing building energy simulation tools can be integrated.
Differences between Modelica and Legacy Simulators used in the Buildings Industry¶
An important difference between Modelica and related languages that have been used in the buildings industry, for example EES, NMF, Smile and MATLAB/Simulink, is that Modelica has orders of magnitude higher investments, it is an open-source language with both open-source as well as commercial modeling and simulation environments (https://www.modelica.org/tools), and it is well-posed to become a de-facto standard for modeling of dynamic, engineered systems. In Modelica, equations are typically encapsulated into models that are then used to schematically define the system architecture. From this high-level language, executable code is automatically generated.
The anticipated outcome of this Annex are open-source, freely available, documented, validated and verified computational tools that allow building and community energy grids to be designed and operated as integrated, robust, performance based systems with low energy use and low peak power demand. Targeting the building energy research community, design firms and energy service companies, equipment and tool manufacturers, as well as students in building energy-related sciences, currently fragmented duplicative activities in modeling and simulation of building and community energy systems will be coordinated taking advantage of Modelica and FMI standards.
Tools and libraries will be validated and their usability tested through their use in projects that support product development, building design, building integration in community energy grids, and through the real-time application of models. Validations will involve comparisons between simulations, analytical solutions and real-scale experiments (synergistic effects to existing Annex projects) and inter-model comparisons using existing standards such as ANSI/ASHRAE Standard 140 or VDI 6020 Standard, as well as process-related verifications.
In terms of the multi-domain approach, the scope is not restricted to specific building types. However, country-specific subtasks will restrict the scope, where appropriate, to specific building types, and/or to a specific modeling level such as a city quarter.
Multi-disciplinary challenges that will be addressed in this project include research and development of the following:
- A common open-source infrastructure of models that allow sharing and deploying to market the contributions of currently uncoordinated activities in Modelica-based building simulation.
- Tool-chains that link object-oriented CAD systems, building design tools and controls design tools with Modelica models, and that allow the deployment of these models to real-time systems in support of building commissioning, building controls and fault detection and diagnostics.
The proposed end-products will be in Subtask 1 modeling libraries, tools for co-simulation and for processes based on a Building Information Model, in Subtask 2 case study reports that illustrates and explains the use of these technologies, and in Subtask 3 a guidebook. In particular, the deliverables will be:
- Subtask 1: Technology Development
- Activity 1.1: Validated, with respect to accuracy and usability, documented models that can be used by designers, manufacturers, control providers, researchers and students with multiple open-source and commercial Modelica simulation environments that use a standardized model format, such as JModelica, OpenModelica, Dymola, SimulationX, MapleSim, Mosilab, MathModelica or AMESim.
- Activity 1.2: New algorithms, implemented in existing building simulation programs and in co-simulation middleware, that allow efficient co-simulation and model-exchange through the Functional Mockup Interface standard.
- Activity 1.3: Interfaces that allow designers to configure Modelica models from a Building Information Model compatible CAD system.
- Activity 1.4: Python packages that automate the workflow of developing and using Modelica and FMI models and co-simulators.
- Subtask 2: Validation and Demonstration:
- Activity 2.1: Case studies that demonstrate to designers the co-design of building energy and control systems under consideration of system dynamics (energy storage and controls), and under uncertainty and variability.
- Activity 2.2: Case studies that demonstrate to urban planners and utilities the integration of buildings into a community-level energy grid.
- Activity 2.3: Software and, for designers, control providers and for students, case studies that demonstrate how to use models to assist in the operation of buildings.
- Subtask 3: Dissemination
- A guidebook that demonstrates how technologies of Annex 60 can be used in applications that are beyond the capabilities of traditional building simulation programs. Applications include rapid virtual prototyping, design of local and supervisory controls, and deployment of models in support of commissioning and operation. Dissemination will also include special tracks at conferences.
Thus, the ultimate goal of this Annex is to leverage, further develop as needed, deploy and demonstrate the use of modern modeling, simulation and analysis techniques that result in:
- Accelerated invention and deployment of integrated systems and performance-based solutions at the building and community level, at a reduced technical risk for early adopters, through a model-based design and performance verification process.
- Reuse of models across the whole building life cycle to ensure realization of design intent and persistence of energy savings, peak demand reduction and comfort through proper operation.
- Increase of system-level performance through more effective collaboration facilitated by interoperable models and simulators for the currently vertically separated disciplines: controls, thermal systems, daylighting, electrical systems and air quality.
Compared to conventional building simulation programs, the end-product will thus provide the means to reduce energy consumption and peak power demand through:
- Activity 1.1: Rapid prototyping of new equipment, systems and controls that accelerates time-to-market for innovative low energy systems. Engineers in leading edge design firms are not restricted to design, test the performance, and size systems for which implementations in conventional building simulation programs already exist. Rather, they themselves can add models of innovative technologies, which may not yet have deep enough market penetration to have implementations in conventional building simulation programs.
- Activity 1.2: Standardized APIs that contribute to simulator interoperability, which allow designers to link different simulators for run-time data exchange in order to conduct integrated analysis of buildings in which control loops are closed across domains that are simulated by different tools. An example of this is the control of an active facade in which Modelica computes thermal comfort and Radiance computes glare, and both are input signals to a controller that actuates the shade. During operation, standardization of simulator interfaces facilitates the use of models to monitor and verify conformance to the design intent as represented by the model.
- Activity 1.3: BIM to Modelica translators that provide a means to specify in a BIM not only the envelope and HVAC components, but also their control sequences. During design, this allows improving the effectiveness of controls to reduce energy and peak demand using a BIM-based process. During commissioning, the actual implemented control sequences can be verified versus an executable representation of the design intent, as stored in the BIM. Furthermore, building performance models can be updated more efficiently in terms of following model changes within an integrated planning process.
- Activity 2.1: The quantification of risk, the co-design of HVAC and control sequences and sizing that takes into account thermal storage (where equipment size is the solution to an optimal control problem) allows proper sizing of systems, thereby avoiding inefficient operation due to operating equipment at low part-load efficiencies.
- Activity 2.2: Peak load reduction and load shifting at the community level that allows reducing the need for electrical reserve capacity generators, which typically have high CO2 emissions and large capital costs.
- Activity 2.3: The use of models to augment monitoring, control and fault detection and diagnostics methods. This promises to detect a degradation of equipment efficiency over time because measured performance can be compared to expected performance at the current operating conditions. Furthermore, use of models during operation allows operational sequences to be optimized in real-time to reduce energy or cost, subject to dynamic pricing.
Means and Technical Approach¶
The Annex objectives shall be achieved by the participants collaboratively working on the Subtasks described below and as summarized in Figure 2. Subtasks 1 will develop the required software technology. Significant work already exists and will be leveraged. Subtask 2 is focused on validation, verification, demonstration and deployment of the developed software technology in the context of whole building and community energy performance design and operation. Due to the work that already exists for Subtasks 1, all Subtasks will start concurrently. Subtask 3 will develop a guidebook, organize special tracks at professional conferences, and ensure effective collaboration with professional organizations
Subtasks 1 and 2 are further organized into the following activities.
Subtask 1 - Technology Development¶
Activity 1.1 Modelica model libraries will develop free open-source libraries of Modelica models for building and community energy systems with associated documentation for new and experienced users. This will be accomplished through the further development and validation of existing libraries. An outcome will be comprehensive free open-source libraries that provide the modeling infrastructure for the overall Annex as well as for the buildings industry.
Activity 1.2 Co-simulation and model exchange through Functional Mockup Units will develop co-simulation and model-exchange interfaces in legacy building energy simulation programs and further develop middle-ware for co-simulation and model exchange. The non-proprietary Functional Mockup Interface standard will be leveraged. The outcome will be interfaces in legacy simulators and middleware that allow coupling Modelica models with legacy simulation programs, such as for computational fluid dynamics or daylighting for which equation-based models may not exist or may not be suited.
Activity 1.3 Building Information Models will develop BIM to Modelica translators for individual buildings, and through integration of geographical information systems, for community energy systems. This will be accomplished through the use of existing standards for exchanging energy calculation data, and through extending standards such as IFC as appropriate. This capability will facilitate the construction of whole building Modelica models, it will integrate energy performance simulation, especially with respect to Modelica, with the developments of BIM-based tools that are ongoing outside of this Annex, and provide a path for a next-generation BIM that also specifies control sequences.
Activity 1.4 Workflow automation tools will develop free open-source Python packages to automate the workflow of developing and using Modelica and FMI models and co-simulators. The Python packages will assist developers in unit tests and checking libraries with conformance to coding guidelines, users in pre-processing and post-processing batches of simulation, such as for parametric studies or for uncertainty propagation, and in the integration of Modelica or FMI models with optimization packages for design optimization and Model Predictive Control. This activity is meant to avoid duplication of work that would otherwise be conducted in the different activities of Subtask 2.
Subtask 2 - Validation and Demonstration¶
Activity 2.1 Design of building systems will demonstrate how to co-design energy and control systems for buildings and how to size systems under consideration of diurnal weather patterns, energy storage and time-varying electricity prices of a smart grid. The approach is to formulate the problem as an optimal control problem, possibly with stochastic input, and solve the problem using existing nonlinear programming algorithms. This will demonstrate how new requirements of smart grids affect the design and operation of buildings, how state-of-the-art computer algorithms can help in designing very low energy buildings and ultimately lead to correctly-sized systems that operate efficiently across a larger range of part-load conditions.
Activity 2.2 Design of district energy systems will validate and use tools from Subtask 1 to design district energy systems and to develop control algorithms for smart grid integration of clusters of buildings, e.g., an office park, campus or residential neighborhood. This will be accomplished by integration of models for thermal, electrical and control systems at different levels of fidelity depending on the scale of the district energy system. This capability will assist utilities and control providers in the development of control strategies and rate structures for energy grids with a large share of distributed renewable energy generation.
Activity 2.3 Model use during operation will use control models from Activity 1.1 and FMI export programs from Activity 1.2 to deploy energy-aware control algorithms and models for monitoring to experimental facilities, building automation and energy management systems. The integration of these algorithms and models will be done through Functional Mockup Units which provide a simulator and control-system independent software interface. This will allow model use in hardware-in-the-loop experimentation, during building commissioning for functional testing, and during building operation to ensure energy-minimizing and smart-grid responsive control that continuously monitors performance relative to design intent.