Abstract
This dataset includes high resolution, detailed end use data from a net-zero occupied home that demonstrates zero-carbon living and transportation capacity. The house is located in Davis, California, U.S., and the dataset includes full year data from 2020 with 1 minute time resolution. The data has been monitored with more than 230 sensors installed in the house, and there are total 332 channels available. The data includes detailed end use electricity data (e.g., HVAC system, lighting, plug load including major appliances), building's interior thermal conditions (e.g., indoor air temperatures in multiple rooms and relative humidity), HVAC system operation data (e.g., soil temperatures around ground bores and supply water temperatures), on-site power generation system data (e.g., PV power supply and PV surface temperatures) and etc. The original dataset from the house has been curated, and the data has been carefully reviewed for quality check. The data quality check revealed there are 156 minutes of data were missing in the month of April, and around 1,404 minutes of data was missing in August. The data gap was filled with linear interpolation in case the gap is less than continuous 6 hours. Otherwise, the data is filled with -9999. The data curation has been processed using the Tsdat framework (https://github.com/tsdat/tsdat). In addition, a semantic description for the dataset was generated by leveraging the Brick (https://brickschema.org/). The final curated and processed data as well as raw data are currently available through https://bbd.labworks.org/ds/bbd/hshus.
Keywords: Smart Home, Energy Efficiency, Sustainable Home, Electrification, Renewable Energy, Ground Source Heat Pump, Net-zero home
Specifications Table
| Subject |
Energy
|
| Specific subject area | This dataset includes rich data from an occupied net-zero residential smart home. Corresponding specific subject areas could be
|
| Type of data | The data types include temperature, relative humidity, solar radiation, flowrate, pressure, status, power, and energy consumption. The details are shown in section “Data description”. |
| How the data were acquired | The data were collected by multiple kinds of sensors and recorded in multiple data loggers. The details of the sensors and loggers are shown in section “Experimental design, materials, and methods”. |
| Data format |
|
| Description of data collection | This dataset was collected from a fully occupied home with normal operation condition. The dataset includes full-year data for 2020, and the data resolution is 1 minute. The original data was stored at 6-month interval. The target home had been occupied with one family (three occupants) from Jan 1 to Apr 6, 2020. Due to the impact of Covid-19 pandemic, they left the city (Davis, CA) on Apr 6. The home was then un-occupied from Apr 7 to Aug 21, 2020. It was finally occupied with another two residents for the remainder of the year. Efforts had been paid to find short-term occupants to fill the vacant time, but the house was kept unoccupied for several months due to the pandemic. |
| Data source location |
|
| Data accessibility | Repository name: Benchmark Datasets of Building Environmental Conditions and Occupancy Parameters Data identification number: 10.17041/1856495 Direct URL to data: https://bbd.labworks.org/ds/bbd/hshus <Peer Review Only> Website: https://bbd.labworks.org/ Email: HondaSmartHomeUser@gmail.com Password: HondaSmartHome1! |
| Related research article | There is no research article available at this moment. |
Value of the Data
This paper contains data from a zero-carbon residential home, called Honda Smart Home, in California, USA. It is a 2-story single family residence with detached garage. Honda Smart Home includes multiple advanced systems, such as adaptive circadian lighting, passive solar design, radiant geothermal heating & cooling, and pre-cooling system.
The building has a living area of 180 m2 (1,944 ft2). This house serves a living laboratory for implementing, demonstrating and measuring the impact of various energy efficient technologies and practices for creating net zero homes. The home may deploy a particular technology for a limited period of time for measuring the performance and impact of that technology or combination of technologies.
The high-resolution data (i.e., 1-minute) allows researchers and data users to investigate the performance of individual energy efficient technologies and overall highly efficient and/or net zero energy homes.
The usefulness of such datasets to scientific community includes:
-
•
The data set can establish the ground truth of a building's operation.
-
•
Correlations in the data can suggest least-cost pathways to accomplish tasks like nonintrusive load monitoring, virtual sensing, building energy model calibration, forecasting, benchmarking, control optimization, fault detection, and many others.
-
•
Processing lower-resolution versions of this data can help us understand the achievable accuracy and precision of derived conclusions when sensor data is not nearly as ubiquitous.
-
•
Researchers, facility managers, sensors and controls platform developers can benefit from this data.
1. Data Description
The Honda Smart Home has diverse measurements, including HVAC system, individual lighting and plug loads, water, renewable power, batteries, plug-in EV etc. Data were recorded on a minutely basis from a range of sensors that monitored electricity usage, equipment performance, indoor conditions, and outdoor conditions.
There are more than 300 measured data channels. Based on the physical features, these data can be categorized into different types, like temperature, flowrate, and power, as listed in Table 1. The examples of data variables are also listed.
Table 1.
Data types and example of data variables.
| Data type (unit) | Example of data variables |
|---|---|
| Temperature (degF) | Outdoor air temperature |
| Indoor air temperatures | |
| Ceiling panel supply air temperature | |
| Cold water supply temperature | |
| Desuperheater supply/return temperature | |
| Surface temperature of walls/ceilings/slabs/insulations | |
| Temperature inside walls | |
| Heat pump chilled/cooling water temperature | |
| Dry bore temperature | |
| Wet bore temperature | |
| Soil temperature | |
| Relative humidity (%) | Outdoor air relative humidity |
| Indoor air relative humidity | |
| ceiling panel supply air relative humidity | |
| Air relative humidity in wall cavity | |
| Solar radiation (W/m2) | Incident Solar Radiation |
| Flowrate (Gal) | Cold water flowrate |
| Heat pump refrigerant flowrates | |
| Heat pump chilled/cooling water flowrate | |
| Ceiling panel flowrate | |
| Domestic hot water flowrate | |
| Sink flowrate | |
| Shower flowrate | |
| Clothes washer water flowrate | |
| Pressure (hPa) | Outdoor barometric pressure |
| Status | Fan status |
| Status of window actuator | |
| Lighting status | |
| Heat pump heating/cooling status | |
| Ceiling loop valve opening | |
| Power (KW or Btu/h) | Plug power |
| PV supply power | |
| Appliance power | |
| Fan power | |
| HVAC power | |
| Water heater power | |
| Battery supply/sink power | |
| Grid supply/sink power | |
| Lighting power | |
| Delivered heating/cooling energy from heat pump | |
| Heat pump energy consumption | |
| Delivered energy from floor | |
| Delivered energy from ceiling panel | |
| Energy consumption (kWh) | HVAC Inverter |
| Lighting | |
| Appliance | |
For exhibition purpose, the following figure (Fig. 1) shows several processed time-series data of typical summer and winter month (i.e., Aug and Feb) in the year of 2020. TOA means the measured outdoor dry bulb temperature, TMB refers to the measured air temperature in the master bedroom, TKit means the measured air temperature in the kitchen. It is worth noticing that the data logger did erroneous re-zeroing on Feb 17,2020, so that some variables, like TMB and TKit, were not recorded correctly on that day, as shown in Fig. 1.
Fig. 1.
Example data – Outdoor temperature, room temperatures for master bedroom and kitchen.
2. Experimental Design, Materials and Methods
2.1. Data collection
All the measured data in this target home were collected by multiple kinds of sensors, as listed in Table 2. The sensors were connected with multiple data loggers and all the measured data were then transferred to the data loggers, as listed in Table 3.
Table 2.
Sensor list.
| Sensor type | Sensor manufactory | Sensor model | Number |
|---|---|---|---|
| Thermocouple (Immersion probe) | Omega | TQSS-18U-6 | 13 |
| Thermocouple (Twisted/Shielded) | Omega | EEPP-T-20-TWSH-SLE | 75 |
| Thermistor | Omega | 33 | |
| Power meter | Schneider Electric | 44 | |
| Wattnode | WNB-3D-240P | 1 | |
| Flow meter | Omega | FTB431 | 20 |
| Onicon | F1300-1” brass | 5 | |
| Dwyer | WMT2-A-C-03 | 2 | |
| RH | Setra | SRH1-2P-W-2C-T5-N/ SRH1-2P-D-2C-T0-N |
6 |
| RTD | Setra | SRH1-2P-W-2C-T5-N/ 2651R25WD2BT1C | 5 |
| Pyranometer | Licor | Li 200Sz | 1 |
| Anemometer | Vaisala | 2 | |
Table 3.
Logger list.
| Logger manufactory | Logger model | Number | |
|---|---|---|---|
| dataTaker | DT85 Series | 74 | |
| CEM20 #1 | 40 | ||
| CEM20 #2 | 36 | ||
| CEM20 #3 | 40 | ||
| Vaisala | WXT 520 | 7 | |
| Schneider Electric | Powerlink MVP BCPMA | 43 | |
| Dent | PoweScout 3 | 9 | |
| Cantara | AMX | 93 | |
| Advantech | ADAM 4150 | 21 | |
One whole year measured data (from Jan 1, 2020 to Dec 31, 2020) recorded in 12 monthly data files were downloaded from its official website and formed as raw data. All the systems were with normal operation conditions and the data resolution is 1-minute.
2.2. Data processing
The raw data were then checked, curated, and integrated into one single whole-year data file (called “honda_latest_curated_20220322.csv”) and presented in this paper.
A complete list of variables (channels) in the dataset is listed in the raw data files. This variable list introduces the description for each variable (channel), including the subsystem, measurement location, and measured parameters. The missing data in the raw data files were filled by using a programmatical approach. The approach that handles the raw data is introduced as following:
-
1.
We used the Tsdat (https://tsdat.readthedocs.io/en/latest/), an open-source Python framework for pocessing and standardizing time-series data [1]. We created a data pipeline configuration for Tsdat and executed the pipeline with the raw data. The execution of the data pipeline generates a curated data.
-
2.
For all the columns, linear interpolation of missing data gaps was done for the data gap with the gap of less than continuous 6 hours. All the rows that contain linearly interpolated values were also marked so that the data user can track if the value is either actually coming from the sensor or linearly interpolated by data curator.
-
3.
For the data gaps longer than continuous 6 hours, the value of -9999 was filled.
We also developed a Brick model, which is a semantic metadata description for the dataset [2]. Complex relationships across locations, sensors, equipment, and data labels are represented in a graph using standard terms and concepts provided by the Brick schema (https://brickschema.org/). The representation is saved as an RDF TTL (Terse Triple Language, pronounced Turtle) formatted file. Fig. 2 shows a part of the developed Brick model and a visualization of the developed model.
Fig. 2.
Developed Brick ontology and its visualization.
Ethics Statements
As part of the Occupancy Agreement, all occupants expressly agree and understand that the house is monitored and that the data is recorded, and that the house uses prototype systems for HVAC, Lighting, and Energy Management.
CRediT Author Statement
Borui Cui: Data curation, Writing – original draft preparation; Sangkeun Lee: Data processing, Tsdat pipeline configuration, Brick ontology development; Michael Koenig: Original data monitoring; Piljae Im: Funding acquire, supervision, Writing – original draft preparation; Mahabir Bhandari: Writing – review & editing.
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Data Availability
References
- 1.Lansing Carina, et al. OCEANS 2021: San Diego–Porto. IEEE; 2021. Tsdat: An Open-Source Data Standardization Framework for Marine Energy and Beyond. [Google Scholar]
- 2.Balaji Bharathan, et al. Proceedings of the 3rd ACM International Conference on Systems for Energy-Efficient Built Environments. 2016. Brick: Towards a unified metadata schema for buildings. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.