A dataset of volatile organic compounds (VOCs) concentrations collected near a petroleum refinery area

Abstract

Air pollution is a significant environmental and public health concern, as prolonged exposure to air pollutants can lead to respiratory and cardiovascular diseases, as well as an increased risk of cancer. Highly industrialized urban areas are characterized by the presence of a wide variety of pollutants in the air. Industrial areas with petroleum refining and petrochemical facilities emit large amounts of volatile organic compounds (VOCs) into the atmosphere, which promotes the formation of tropospheric ozone (O₃) and fine particles through photochemical processes generated in the troposphere. Furthermore, the combustion of fuels such as gasoline and diesel is a significant source of VOC emissions. VOCs represent the most hazardous class of air pollutants. Therefore, in this paper, we present a dataset containing real-time measurements of VOCs, including hexane, benzene, toluene, methane, acetone, and ammonium, as well as a calculation of the total VOCs. Furthermore, the dataset is enriched with meteorological parameters, such as temperature, relative humidity, barometric pressure, wind direction, and wind speed, all of which are measured within the same time window as the pollutants. The data were collected at an air quality monitoring station equipped with a low-cost sensor (LCS), located near a petroleum refining plant in Cadereyta, Nuevo León, Mexico, from March 1 to May 31, 2024. The dataset serves as a valuable resource for identifying air pollution trends in the area and for examining the correlation between VOC emissions and human health in regions associated with the petroleum refining industry.

Specifications Table

Subject	Earth & Environmental Sciences
Specific subject area	Air pollution, low-cost sensor, air quality monitoring network, data analytics
Type of data	Table, Figure Raw, Analysed
Data collection	Data was collected using an air quality monitoring station equipped with low-cost sensors installed 9 m above ground level. Total volatile organic compounds (TVOC) were recorded using an SGP30 sensor. Temperature, relative humidity, and barometric pressure were measured with a Bosch BME280 sensor. The presence of toluene, hexane, methane, ammonium, and acetone in the air was measured using laboratory calibrated MQ series sensors. All sensors were configured to automatically collect and measure the pollutants every 3 min, with real-time transmission to a repository hosted in private cloud storage via an internet service.
Data source location	• Institution: Autonomous University of Tamaulipas • City/Town/Region: Cadereyta, Nuevo León • Country: México • Latitude and longitude: 25.60094132860872, −99.99562936220205
Data accessibility	Repository name: Dataset of Volatile Organic Compounds (VOC) at ground level near to petrochemical plant Data identification number: 10.17632/d2h7zrdzn8.3 Direct URL to data: https://data.mendeley.com/datasets/d2h7zrdzn8/3
Related research article	None

1. Value of the Data

• •The VOC dataset collected using low-cost sensors (LCS) can be regarded as the first available online with data from northeastern Mexico.
• •High levels of air pollutant concentrations have led to an increase in respiratory and cardiovascular diseases [1,2]. Furthermore, the compounds benzene, toluene, ethylbenzene, and xylene have been identified as having potential carcinogenic effects [[3], [4], [5], [6]]. In this sense, the dataset enables the determination of air pollution levels by providing the concentration levels of volatile organic compounds (VOCs) at the neighborhood level, which are affected by emissions from a petrochemical industrial area.
• •The dataset includes attributes with values for total VOC concentrations and the five most significant compounds associated with the petroleum refining and petrochemical industries.
• •The dataset can support air quality analysts in examining the influence of meteorological parameters on VOC air pollution.
• •This dataset can serve as a foundation for studies exploring the relationship between VOC air pollution and human health in regions linked to the petroleum refining industry.

2. Background

Air pollution in metropolitan urban areas is characterized by the presence of various pollutants, including particulate matter (PM_2.5 and PM₁₀), nitrogen oxides (NOx), tropospheric ozone (O₃), sulfur dioxide (SO₂), and volatile organic compounds (VOCs). VOCs are composed primarily of hydrogen and carbon atoms and include a diverse group of >99 compounds, classified by their chemical properties as alkanes, alkenes, aromatics, and oxygenates, among others [7]. The primary sources of VOC emissions include industrial processes and vehicular emissions (light and heavy transport) [[8], [9], [10]]. In the industrial sector, processes including fossil fuel combustion, petroleum refining, petrochemicals, industrial wood processing, alcoholic beverage production, and the food industry are sources of VOC emissions [[11], [12], [13], [14]].

On the other hand, vehicle emissions are a source of benzene, toluene, ethylbenzene, and xylene [9]. The VOCs have high volatility at room temperature (20 °C), evaporation capacity in the presence of sunlight and high chemical reactivity, which favors their interaction with other atmospheric pollutants and their participation in photochemical processes for the formation of O₃ [8,[15], [16], [17]]. Additionally, urban areas can be affected by pollution emitted from industrial zones because the climatic and topographic characteristics of each region influence the dispersion and movement (vertical and horizontal) of the pollution plume, transporting pollutants to other parts of the city, particularly in the distribution of VOC concentrations [18,19].

Currently, the technological solutions based on low-cost sensors (LCS) for monitoring air pollution and meteorological parameters are emerging as a viable alternative to the high implementation and maintenance costs associated with monitoring stations equipped with reference instruments, enabling increased granularity in air quality monitoring. LCS are primarily recognized for their low power consumption, compact size, local or remote storage of sensed data, and compatibility with Internet of Things (IoT) technologies [20]. Additionally, these sensors are distinguished by their ability to connect with other sensors, often managed by a microcontroller that controls multiple sensors simultaneously. This setup creates a monitoring station capable of handling real-time data transfer and supporting long-term continuous monitoring. Finally, LCS are characterized by their low cost (compared to reference instruments), as they can typically be purchased for under US$500 per unit.

3. Data Description

The dataset of volatile organic compounds (air pollutants) and meteorological parameters was collected at an air quality monitoring station in the city of Cadereyta, Nuevo León, located in northeastern Mexico (see Fig. 1), which is part of the Monterrey metropolitan area. The air quality monitoring station (AQMS-IoT-05) is situated in a residential area (in the northern part of the city) close to a large petroleum refining plant, surrounded by three wide avenues with heavy vehicular and cargo traffic.

Fig. 1.

Location of air quality monitoring station (blue map pointer) at Cadereyta, Mexico. The petrochemical plant is marked on the map with an orange bullet.

The Dataset_TVOC_v3.xlsx file contains raw data (2208 records) collected from March 1 to May 31, 2024. This file includes hourly mean values for the following attributes: Idsample (unique record identifier), Date, Hour, Hexane, CH4 (methane), Toluene, NH4 (ammonium), Acetone, and H2 (hydrogen), where the values are represented as the concentration of the polluting gas measured in parts per million (ppm). Moreover, the TVOC represents the estimated total concentration of various VOCs identified by the sensor, measured in ppm. The WD and WS attributes contain records of wind direction and wind speed, where the values represent degrees (°) and meters per second (m/s). The attributes T and RH contain recorded values for temperature ( °C) and the percentage of relative humidity (%), respectively. In the case of BP, it represents the barometric pressure in hPa, calculated by measuring absolute pressure. An example of the air pollutant records contained in the Dataset_TVOC_v3.xlsx file is shown in Table 1. The records are automatically assigned an identifier and sorted by the time and date they were stored in the database by the information system that manages the monitoring station's operation. Table 2 shows the continuation of the attributes included in the Dataset_TVOC_v3.xlsx file, displaying the attributes corresponding to the meteorological parameters considered in the monitoring.

Table 1.

Excerpt from the dataset showing only the air pollutant attributes.

Idsample	DATE	HOUR	HEXANE	CH4	TOLUENE	NH4	ACETONE	H2	TVOC
SE3_0030001	01/03/24	01:00	0.0978	0.0216	0.14	1.66	0.13	34.75	0.023
SE3_0030002	01/03/24	02:00	0.0978	0.0216	0.13	1.5	0.11	31.65	0.021
SE3_0030003	01/03/24	03:00	0.0978	0.0216	0.14	1.64	0.12	34.60	0.021
SE3_0030004	01/03/24	04:00	0.0978	0.0216	0.17	1.83	0.14	34.81	0.021
SE3_0030005	01/03/24	05:00	0.0978	0.0216	0.17	1.86	0.15	39.57	0.022
SE3_0030006	01/03/24	06:00	0.0978	0.0216	0.2	2.07	0.17	40.06	0.023
SE3_0030007	01/03/24	07:00	0.0978	0.0216	0.71	5.2	0.6	41.57	0.023
SE3_0030008	01/03/24	08:00	0.0978	0.0216	0.99	6.62	0.83	46.24	0.029

Table 2.

Excerpt from the dataset showing the attributes of the meteorological parameters.

Idsample	DATE	HOUR	WD	VS	T	RH	BP
SE3_0030001	01/03/24	01:00	176	1.111	13.37	74.76	980.85
SE3_0030002	01/03/24	02:00	186	0.5	12.6	85.05	980.47
SE3_0030003	01/03/24	03:00	210	0.417	12.34	91.28	979.89
SE3_0030004	01/03/24	04:00	280	0.417	12.01	91.42	979.51
SE3_0030005	01/03/24	05:00	247	0.583	11.12	88.67	979.58
SE3_0030006	01/03/24	06:00	250	0.472	10.75	91.17	979.78
SE3_0030007	01/03/24	07:00	263	0.472	10.73	94.07	980.46
SE3_0030008	01/03/24	08:00	178	0.5	11.65	90.48	981.01

Table 3 shows a descriptive statistical analysis of each variable contained in the dataset. The descriptive analysis is divided by concentration by month, allowing for the identification of changes in air pollutants in relation to meteorological parameters. It is worth noting that these changes may also be attributed to local and neighboring emission sources. We observed that the mean concentrations of the hexane, methane, and toluene compounds increased with increasing temperature. This dataset includes 21 days (March 1–21) corresponding to the winter season, while the remaining records correspond to the spring season. These compounds are characterized by high volatility, reacting to increased temperature by converting more rapidly to vapor. The variability in the concentrations of hexane, methane, and toluene measured by the sensors is observed in the median calculated in Table 3. For example, median concentrations of hexane gas of 0.2 ppm were estimated in March, 0.51 ppm in April, and 1.15 ppm in May, representing a significant increase.

Table 3.

Descriptive statistics of air pollution and meteorological parameter variables.

Month	Variable	Mean	SD	Median	IQR	Min	25 %	75 %	Max
March	Hexane	0.15	0.15	0.13	0.12	0.01	0.06	0.18	1.54
	CH4	0.07	0.08	0.04	0.06	0.01	0.02	0.08	0.73
	Toluene	0.22	0.19	0.16	0.13	0.04	0.12	0.25	1.98
	NH4	2.12	1.16	1.75	1.07	0.65	1.41	2.48	10.86
	Acetone	0.19	0.16	0.14	0.12	0.04	0.1	0.22	1.64
	H2	42.31	4.42	42.22	5.93	26.73	39.28	45.22	60.21
	TVOC	0.03	0.01	0.03	0.01	0.01	0.02	0.03	0.21
	WD	175.14	92.51	151	135.24	1.00	104	239.24	357
	WS	1.3	0.93	1.07	1.33	0.03	0.5	1.83	5.33
	T	24.83	6.35	24.05	7.94	10.73	20.42	28.36	42.84
	RH	77.69	20.75	87.15	33.91	21.03	60.75	94.66	100
	BP	974.37	5.57	973.25	6.69	948.2	971.57	978.26	987.03
April	Hexane	0.76	0.77	0.51	0.91	0.01	0.27	1.18	8.28
	CH4	0.16	0.16	0.11	0.14	0.01	0.06	0.2	1.82
	Toluene	0.23	0.28	0.18	0.13	0.04	0.12	0.25	4.48
	NH4	2.19	1.45	1.91	0.98	0.62	1.46	2.44	19.52
	Acetone	0.2	0.23	0.15	0.1	0.03	0.11	0.21	3.64
	H2	40.43	4.76	40.62	6.86	24.44	36.98	43.84	52.8
	TVOC	0.03	0.01	0.02	0.01	0.01	0.02	0.03	0.21
	WD	152.27	80.39	126.45	75.53	3.00	103	178.52	360
	WS	1.82	1.29	1.49	1.56	0.06	0.89	2.44	7.92
	T	28.4	6.64	27.38	10.07	15.15	23.9	33.96	43.83
	RH	77.96	26.41	92.88	42.71	5.16	57.29	100	100
	BP	973.51	5.16	973.17	7.51	959.14	969.62	977.13	985.75
May	Hexane	1.27	0.77	1.15	0.44	0.27	0.94	1.38	8.95
	CH4	0.27	0.24	0.2	0.22	0.02	0.12	0.34	1.89
	Toluene	0.39	0.5	0.27	0.24	0.09	0.19	0.43	8.8
	NH4	3.18	2.19	2.58	1.61	1.17	2.03	3.64	31.7
	Acetone	0.33	0.41	0.23	0.2	0.08	0.17	0.37	7.04
	H2	32.57	11.93	37.92	17.55	1.05	23.5	41.05	50.96
	TVOC	0.03	0.02	0.03	0.01	0.01	0.02	0.04	0.24
	WD	124.21	64	108	58	1.00	92	150	358
	WS	1.73	1.13	1.49	1.59	0.08	0.94	2.53	5.31
	T	34.14	6.24	32.9	10.72	21.29	28.74	39.45	49.88
	RH	80.2	20.78	85.28	34.18	27.45	65.82	100	100
	BP	966.82	28.39	968.61	4.08	399.7	966.52	970.6	982.85

Fig. 2 shows a time series comparison of the methane concentration values, illustrating hourly concentrations across the three months included in the dataset. This graph visualizes the variability and dispersion of the data during this period, highlighting outliers found in April, and with the highest concentrations occurring in May. The peak values were recorded on the 9th, with high concentrations also occurring on the 13th, 15th, 18th, 21st, 30th, and 31st In April, the highest outlier was registered on the second day of the month.

Fig. 2.

Comparison of monthly records of the hourly concentration of the pollutant methane. Source: Author. Dataset: Dataset_TVOC_v3.xlsx.

Fig. 3 illustrates the relationship between air pollutants and meteorological variables, where the Spearman correlation coefficient determines the strength of the association between the variables. This graph highlights the strong associations between NH4 and toluene, as well as between acetone and toluene, and between acetone and NH4, which are observed in both the general correlation analysis and the analysis by month. Moreover, the density graph for each variable, shown in Fig. 3, illustrates the distribution of the variables. In this graph, the variables are grouped by month to facilitate comparisons among them. Furthermore, the scatter plots display the relationship between two continuous variables, with each observation represented as a point.

Fig. 3.

Scatter, density, and correlation plots were generated using the values recorded by the AQMS-IoT-05 monitoring station. This diagram groups them by month to compare the values of all variables over the three months. Source: Author. Dataset: Dataset_TVOC_v3.xlsx.

Fig. 4 displays a heatmap, which provides a visual representation of temperature data for the entire month of May 2024. This graph intuitively identifies the low and high temperature values and how they change throughout the day, distinguishing between daytime and nighttime. On the 24th and 26th, temperatures exceeding 45 °C are recorded between 1:00 pm. and 4:00 pm.

Fig. 4.

Heatmap plot of the temperature attribute for May 2024. Source: Author. Dataset: Dataset_TVOC_v3.xlsx.

4. Experimental Design, Materials and Methods

The monitoring station includes a low-cost SGP30 sensor (LCS) to measure TVOC concentrations within a range of 0 to 60,000 ppb. This sensor detects a wide variety of VOC gases with a calibrated output and high accuracy. Additionally, a function was configured in the software code deployed on the microcontroller to manage the SGP30 sensor, thereby improving measurement accuracy by compensating for ambient humidity with an external sensor that calculates the relative humidity percentage. Temperature, relative humidity, and barometric pressure were measured using the Bosch BME280 sensor, which is known for its high performance and accuracy. The sensor has a temperature measurement range of −40 °C to 85 °C with an accuracy of ±1.0 °C. The relative humidity measurement has a tolerance of ±1 % with a response time of 1 s. The pressure sensor has a sensitivity error of ±0.25 %. Therefore, this LCS is characterized by exceptional precision in measuring meteorological factors. The gases hexane, CH4, toluene, NH4, and acetone were detected and measured using independent MQ series low-cost sensors. The MQ LCS were calibrated in a laboratory using a vacuum chamber, with a controlled interior temperature of 20 °C and a relative humidity of 55 % (±5 %). Additionally, an electrical resistor with the required capacity for each pollutant was installed, as specified in the technical data sheet provided by the manufacturer.

The monitoring equipment was installed on the roof of a monitoring station at a height of 9 m above ground level, with a 2.4 m separation between the roof and the sensors, and no obstacles in the surroundings (e.g., trees, building walls, or shrub barriers) to ensure accurate measurement of the air pollutants present in the study area. In the designed technological solution, each sensor was managed by a microcontroller capable of transmitting data in real time via a wireless link to an internet service, collecting data every three minutes. The gas sensors were equipped with relative humidity and temperature sensors to adjust their measurements. Three devices are installed for each pollutant to prevent failures or data loss during the study period. The data transmitted by microcontrollers was received by a web service available in a private cloud over the internet. A set of business rules was established to validate the consistency of the collected data. Additionally, another set of business rules calculates the hourly average of the collected data for each attribute, automatically saving each calculated average value in a separate database, also available in the same repository within the private cloud. This feature enables an exclusive operation database and a separate database for the data analysis process, without affecting the system's operation.

Limitations

The dataset represents the concentration of diverse air pollutants, specifically volatile organic compounds (VOCs), over a specific period in Cadereyta, Mexico, which poses a geographic limitation. Additionally, considering the city's vast area, the industrial sector within its boundaries, and the mountainous terrain, it is recommended to install more monitoring stations in other regions while maintaining the same neighborhood-level approach, in order to establish a dense air quality monitoring network.

Ethics Statement

The authors have read and follow the ethical requirements for publication in Data in Brief and confirming that the current work does not involve human subjects, animal experiments, or any data collected from social media platforms.

CRediT Author Statement

Jailene Marlen Jaramillo-Perez: Conceptualization, Methodology, Software, Validation, Data curation, Formal analysis, Investigation, Writing – original draft, Visualization. Bárbara A. Macías-Hernández: Conceptualization, Methodology, Formal analysis, Investigation, Writing – original draft, Writing – review & editing, Supervision, Project administration, Funding acquisition. Edgar Tello-Leal: Conceptualization, Validation, Data curation, Formal analysis, Resources, Writing – original draft, Writing – review & editing, Visualization, Supervision, Project administration, Funding acquisition.

Data Availability

• Mendeley DataDataset of Volatile Organic Compounds (VOC) at ground level near to petrochemical plant (Original data).