MCP-iBAT: field measurements on twin full-scale test cells with a PCM coating in Reunion Island (2020–2026)

Description

This dataset contains almost six years (June 2020 – March 2026) of 10-min field measurements on two identical full-scale test benches located in two contrasting tropical climates of Reunion Island (France): Saint-Pierre (coastal lowlands, 60 m a.s.l., 21.3408° S, 55.4911° E) and Le Tampon (mid-altitude, 670 m a.s.l., 21.27° S, 55.50° E). Each bench (5.86 × 2.88 × 2.57 m) is made of lightweight steel–polyurethane sandwich panels and divided into two twin, unoccupied and closed cells. In cell B, the inner surface of the sun-exposed north wall (6.86 m²) is coated with a commercial plaster containing a microencapsulated paraffin phase change material (PCM, Enerciel®, 3–5 mm). Cell A, without PCM, is the reference. Short periods recorded before the coating was applied (June 2020 at Saint-Pierre, July–August 2020 at Le Tampon) are included. Measured variables: - inner surface temperatures of all walls, ceiling and floor, and inner and outer surface temperatures of the north walls (type T thermocouples); - indoor air temperature at two heights (thermocouples) and, at Saint-Pierre until September 2023, indoor air temperature and relative humidity (Testo 175 H1); - at Saint-Pierre from July 2021, outer surface temperatures of the other walls, floor and roof, and air temperature in the ceiling void; - conductive heat flux density at the inner surface of the walls and ceiling (Hukseflux HFP01SC heat flux plates); - on-site weather: air temperature and relative humidity, global and diffuse horizontal irradiance, long-wave irradiance, wind speed and direction, rainfall, atmospheric pressure. Files: - MCPiBAT_SaintPierre_20200612-20260308.csv and MCPiBAT_LeTampon_20200724-20260307.csv: one file per site, one row every 10 min (UTC time stamps), with a Configuration column (NoPCM / PCM), an AirConditioning flag (short air-conditioning test at Saint-Pierre, 26–29 July 2021) and harmonised, self-explanatory column names (e.g. q_B_NorthWall_in_FLUX6096); - MCPiBAT_variable_dictionary.csv: unit, description, sensor, original identifier, period of availability and statistics of each variable; - README.txt: description of the files and of the column naming. The data are 10-min averages of 1-min records; invalid channels and values were removed and no gap-filling was applied. The dataset can be used to study the in-situ behaviour and ageing of a PCM-enhanced envelope under tropical conditions, to identify phase-change parameters from field measurements, and to calibrate and validate building energy simulation models. The data were collected within the MCP-iBAT project (PIMENT laboratory, University of Reunion Island) and are described in detail in a Data in Brief article.

Steps to reproduce

1. Experimental setup: two identical prefabricated test benches (Saint-Pierre and Le Tampon, Reunion Island), each divided into two twin cells (A: reference; B: microencapsulated PCM coating on the inner face of the north wall). The cells are unoccupied, closed and free-running (no internal gains; no heating or cooling, except an air-conditioning test at Saint-Pierre on 26–29 July 2021, flagged in the AirConditioning column). 2. Sensors: type T thermocouples made and calibrated in the laboratory (±0.1 °C) on the walls, ceiling, floor, roof and on a vertical measurement line at the centre of each cell (about 0.80 m and 1.60 m above the floor); Hukseflux HFP01SC-05 heat flux plates fixed on the inner surfaces with aluminium tape and a thin layer of thermal paste; Testo 175 H1 temperature/humidity loggers; on-site weather stations of the PIMENT laboratory. 3. Acquisition: Campbell Scientific CR3000 data loggers, scan rate 0.1 Hz, 1-min averages, stored in the database of the PIMENT laboratory. 4. Processing: records extracted from the database, averaged over 10 min (mean of the ten 1-min records), merged with the weather data on a common UTC time base and gathered into one file per site with a continuous 10-min time base. Column names were harmonised (<Q>_<Cell>_<Element>_<Position>_<Sensor>; weather: Met_<Quantity>). 5. Filtering (Python, pandas): removal of invalid or undocumented channels (sensors with constant values, data-logger and weather-station housekeeping channels, faulty weather channels); removal of error codes (−100), values outside physical bounds, isolated spikes (deviation from the 1-h moving median) and flat lines (identical values for ≥ 24 h, or ≥ 6 h for outdoor air temperature and humidity). About 15,300 values (0.05 %) were removed. No gap-filling or other correction. 6. Use: read the CSV files with any tool (comma separator, decimal point, empty field = missing value). Convert Time_UTC to local time with UTC+4. Use MCPiBAT_variable_dictionary.csv to find units and the period of availability of each variable. The raw export and the processing script are available from the authors on request.

Institutions

• University of Reunion Island

  Réunion, Saint-Denis
• Université de Montpellier

  Occitanie, Montpellier

Categories

Funders

• Conseil Régional de La Réunion

  Réunion, Saint-Denis

  Grant ID: GURDTI/20181212-0002116

Licence