Digital Tools for HVAC-Design, Operation and Efficiency Management

Abstract

The project BIM4EEB aims also to develop digital tools to support the design, procurement, installation, post-renovation operation, user feedback and profiling of building automation systems for HVAC. This helps supporting decision making, interaction with tenants and owners during the design, construction, and post-renovation operation phases. The development of the tools will be underpinned by a sound methodological approach. Work will include considerations of interoperability with Smart City technology of automation systems for HVAC. Specific objectives will be related to the development of the following software tools:

• A software component supporting the automatic generation of the layout for control systems emphasising on user preferences and including constraint checking of BAC-topologies against selected building codes. Data and information stored in BIM models are used to generate the initial recommendations and constraints and to deliver the final installation instructions.

• A software component allowing the seamless specification and evaluation of user comfort and systems performance. The underpinning information model will merge data sources from BIM (dimensional data) and BAC (factual data).

• An energy-refurbishment assessment tool, for bridging the gap between commercial simulators and the BIM management system.

• A user-profiling component allowing to compare expectations of tenants and owners regarding comfort and systems’ performance against monitored parameters. The results of this software component can be used in the pre- and post-renovation phases to update the content of BIM systems and thus to improve their accuracy and to reduce efforts for data acquisition and verification.

5.1 Constrains Checking, Performance Evaluation and Data Management in Building Renovation Processes

In any renovation project time is an important factor, the ability to have an early indication of a building’s requirements, in terms of quantities and positions, of HVAC, lighting and other equipment, can significantly reduce costs and time. In addition, performance evaluation and efficient management of data is essential to the monitoring of any project. As part of BIM4EEB, the BIMcpd Platform was developed, and it contained features to help address this.

The constraint checking module specifically focused on providing an overview of the locations of diffusers, ducting, lamps, sockets, and electrical conduits. Users were able to upload 2D images or 3D models, create zones and select products from a catalogue database. The recommended positions for diffusers, lamps and sockets were calculated, with Dijkstra’s algorithm being applied after this to find the shortest path to connect these back to a central point. This offers huge potential to combine with other applications such as computational fluid dynamics (CFD), VR/AR and quantity surveying, to carry out further analysis on the recommendations (Fig. 5.1).

Fig. 5.1

Constraint checking module: model viewer

To ensure that the renovation process delivers the savings (energy, carbon emissions and cost reduction) data needs to be gathered, stored, processed, and analysed. In BIMcpd, two modules were used for this: (a) data management and (b) performance evaluation. Data management consisted of two options for the users, file upload or the API which was connected to the BIM Management System (BIMMS). Once uploaded/connected, the user mapped the data to the BIMcpd structured database, which ensured that the data being uploaded to the database included essential metadata (such as unit type for temperature). The structed database is core to the performance evaluation module and guaranteed that the information displayed in this tool was accurate (Fig. 5.2).

Fig. 5.2

Performance evaluation module

The performance evaluation tool consisted of several features and was designed with the end-user in mind. Firstly, the user did not have to choose which way to display the data, BIMcpd used an algorithmic approach to determine which way was best. Users could query the database without any experience of database querying languages by simply choosing from a series of drop-down lists. Once displayed, they could apply one of many outlier detection algorithms included in the tool to identify, and if they choose, remove them, and reload the visualisation (Fig. 5.3).

Fig. 5.3

M&V change point regression

Measurement and verification (M&V) is the industry standard for energy efficiency projects, BIMcpd contained a full-featured M&V tool, which enabled users to create a baseline (simple linear, change-point or multi-variate regression) and then once the renovation works have been completed (i.e. application of the energy conservation measures or ECMs), re-evaluate by creating a reporting period and comparing this with the baseline to determine if the expected reductions/savings were achieved.

In any M&V project, non-routine adjustments can be made, due to changes in circumstances between the baseline and reporting periods, BIMcpd contains this option also. Therefore, BIMcpd can help reduce time, costs, and maximise data use in renovation projects.

5.2 BIM-Assisted Energy Refurbishment Assessment Tool

The BIMeaser (BIM Early-Stage Energy Scenario) tool is a tool that supports the energy related decision-making process in the early design stage of the renovation process. The tool enables the assessment of several energy refurbishment design options enabling architects and engineers to provide solutions that best fit to the client requirements while optimizing the energy use and comfortable indoor climate conditions for the occupants. The intended use event of the BIMeaser is the collaborative design session, where building designers (architect, structural, HVAC, electricity) discuss about the expected energy saving measures of the renovated building.

One of the important benefits of the BIMeaser tool is the faster energy modelling compared to the traditional approach. The energy modelling time reduction is −75% compared to the traditional manual description of the renovated building in the sophisticated energy simulation program.

5.2.1 The Collaboration in the Energy and Indoor Climate Design Process

BIM4EEB BIMeaser tool enables easy build-up of the “As-is” energy and indoor climate model of the building, applying the renovation scenarios and presenting the impact of each renovation scenario. The targeted user role is “Energy expert”, who is a separate consultant or a member of the design team. The simulated renovation scenario results will enrich the BIM Management (BIMMS) system content. These will be stored in RDF compatible format into the BIM Management system containing links to IFC model used in the simulation. The linking of OPR’s and the BIM model in the BIM Management system enables tracking of the building energy performance during the evolution of the renovation design, which is an important part of the performance-based design approach.

The OPR’s—e.g., operational energy cost, payback time of renovation and summer thermal comfort—are an essential part of the performance-based building design process, which assumes that design selections are validated against the OPR’s in each design stage before moving to a following design stage. The design team will handle the detailed technical energy selections affecting to the OPR’s using the tool as part of the collaborative work (Fig. 5.4).

Fig. 5.4

Overview of BIMeaser data exchange

5.2.2 Tool Operation, Outcomes, and Benefits

The BIMeaser tool and the supporting national renovation measure database was implemented as a web application. The commercial IDA Indoor Climate and Energy 4.8 simulator was connected as a back-end solver with the help of the available API. The technology readiness level of the tool and the API-integration is TRL6.

The intended use event of the BIMeaser is the collaborative design session, where building designers (architect, structural, HVAC, electricity) discuss about the expected energy saving measures of the renovated building. The BIMeaser is connected to the BIMMS, which contain the digital model of the building to be renovated. The digital model is imported to the BIMeaser from the BIMMS. The designers agree several renovation options, which are modelled as energy scenarios by using the supporting “drag and drop” functionality in the tool. The tool contains a national database for the “ready-made” renovation measures, which support the easy application and remembers the previously added measures for further use. Also new measures can be defined into the database if none of the existing ones are suitable. The simulation can be started and after the finalisation of the energy simulations, the values of the Owners Project Requirements (OPR’s) can be reviewed. Finally, the OPR’s can be published back to BIMMS for further use of other services (Fig. 5.5).

Fig. 5.5

An example of the OPR result summary in the BIMeaser tool

The main functionalities of the BIMeaser tool are:

• 1. Easy build-up of the “As-is” energy and indoor climate model of the building by using the BIM and linked data for accurate modelling in the early design stage (Concept design & Preliminary design), where the most important design selections are made according to the costs and performance.

• 2. Apply the renovation scenarios to the “As-is” -building. The BIMeaser tool enhances the collaborative work of the design team in the early stage of the design, which usually lacks the sophisticated indoor climate modelling tools. The indoor climate and energy design is a multi-domain challenge, and it should always be considered as a teamwork.

• 3. Present the impact of each renovation scenario in terms of Owners Project Requirements (OPR).

5.3 Energy-Related Behaviour Profiling Mechanism

One of the main innovations related to building management process is the incorporation of building occupants in the overall analysis. Building occupants are the active entities on the building environment and thus a thorough analysis of this interaction should be considered in any building related decision making. In this section, we present an innovative data driven framework towards the extraction of user-centred knowledge about the building environment. More specifically, the objective is to establish an occupant's context-aware behaviour modelling engine, which will continuously monitor and learn transparently the operational and behavioural patterns (i.e., the preferences) of building occupants, while on the same time interact with the occupants in an ambient manner to define user preferences and extract comfort levels, while taking into consideration also health boundaries. In addition, the overall framework will enable the delivery of Context-Aware Energy Behaviour Profiles, reflecting occupants’ energy behaviour as a function of multiple parameters, such as time, environmental context/conditions occupant comfort preferences and health/ hygienic constraints etc.

5.3.1 Occupancy and Behavioural Profiling Modelling Framework

The occupancy profiling modelling framework is targeting on extracting near real-time information about occupancy presence by processing data coming from different types of sensors in the building. On the other hand, comfort profiling is defined as a non-parametric model which consider the contextual conditions and user preferences to extract dynamically updated information about users’ comfort preferences. In more details, the details of the algorithmic framework towards the extraction of Occupancy and Behavioural Profiling are provided in the following.

Occupancy diversity profiles: The scope of this module is to enable the extraction of accurate occupancy profiles. To exploit the full set of data available, a twostep approach is adopted. At first, initial profiles are defined by an expert and consist of the configuration occupancy file for a building zone. Information about typical occupancy profiles is available at software libraries of professional BIM software. Then, user defined information is incorporated in the analysis. First, the end users of the building may be provided with a tool (user application) to fine tune the accuracy of profiling information. Then and if sensor data available, occupancy diversity profiles are extracted by incorporating actual measurements in the analysis. The overall workflow for the extraction of occupancy profiles is presented in the following figure (Fig. 5.6).

Fig. 5.6

Occupancy analytics process

Comfort analytics engine: The process includes data acquisition of sensorial data from different end points. More specifically, indoor temperature, humidity etc.… data are tracked from sensorial equipment installed in building environment (Fig. 5.7). Outdoor environmental conditions can be available either via outdoor sensors installed or via weather service available in public.

Fig. 5.7

Indoor environmental data—temperature timeseries

In addition to contextual conditions, the comfort analytics engine handle information (control actions) available from the different controllable devices in building premises (i.e., HVAC) as a means of user’s settings associated to its comfort mode. This requires special equipment to be installed in building premises. If no special IoT equipment is available, the building occupants can express their comfort settings against the environmental conditions through an intuitive user interface (UI) developed in the project. A Likert scale approach is considered to get end users’ feedback.

Then by applying ML based techniques (MultinomialNB, GaussianNB or regression techniques) we can extract the level of preference or non-preference of an occupant or group under specific contextual conditions. Among the different classifiers examined, the Random Forest classifier is performing the best results (70.7% accuracy level) for the definition of comfort/discomfort values. The results of the statistical analysis over the data reveals the typical thermal comfort profile for the user of the zone, as a probabilistic function of being at a thermal comfort state under different environmental conditions (Fig. 5.8).

Fig. 5.8

Thermal comfort profiling analysis—distribution curve

5.3.2 Energy Behavioural Profiling Framework

The occupant’s behavioural profiles should be further complemented by energy behaviour profiles with the aim to deliver occupants energy behaviour profiles, reflecting multiple parameters such as time, environment conditions, energy costs, occupant comfort preferences etc.… More specifically, indoor temperature and humidity data, energy consumption, sub metering consumption are tracked from sensorial equipment installed in building environment (Fig. 5.9). Outdoor environmental conditions can be also available either via outdoor sensors installed or via weather service available in public.

Fig. 5.9

Energy consumption—per month data

Then, different ML based techniques (MLP neural net regression Least Absolute Shrinkage Selector Operator linear regression) are considered to correlate energy data with environmental and operational characteristics towards the extract of information about typical energy profiles. This information may be further utilized for the extraction of accurate energy forecasts or energy simulations at the building environment. More specifically, the following correlations are derived from the analysis:

Correlation versus time. Typical month, week and daily profiles are derived from the analysis showing the periodicity of the energy consumption over the time (Fig. 5.10).

Fig. 5.10

Energy consumption—month distribution

Energy consumption versus env conditions. Environmental conditions are mainly associated with heating/cooling operation and thus the correlation with energy consumption should be depicted in the following, we show the accuracy of an energy modelling framework that consider environmental conditions (indoor and outdoor) in the analysis and following a long period training process (Fig. 5.11).

Fig. 5.11

Energy consumption versus indoor conditions

Energy consumption versus Device Type. It is evident that the consumption of the different building devices directly impacts total consumption (as total consumption is the aggregation of device level information). The disaggregation of consumption to the different device types is key information to better understand consumption patterns. Thus, correlation of device level consumption with total consumption is required to extract device level consumption patterns.

We presented in brief the different components and processes that consist of the Occupants Behavioural Profiling framework. This holistic framework is split into different modules to reflect the different layers of the analysis, namely occupancy, comfort, health and energy behaviour. For each of them, sub processes are defined by applying different separate methods and analytics techniques. This microservices based approach ensures a high level of modularity of the overall Occupants Behavioural Profiling framework. The engine is fully adaptable to the demo site needs and requirements and thus fine tuning of the model may apply to meet each project specific objectives and data availability.