A compiled dataset of ready-mix concrete Environmental Product Declarations for life cycle assessment

The carbon footprint of a concrete structure is directly affected by the selected concrete mixture proportions. To better understand the influence of different concrete mixtures, data was collected from Environmental Product Declarations (EPDs). Data from 39,213 U.S.A. ready-mix concrete EPDs was obtained from public repositories provided by the American Society for Testing and Materials and the National Ready Mixed Concrete Association. The EPDs in .pdf format were analyzed using a custom Python script to extract useful information for building designers, sustainability practitioners, and researchers including: life cycle assessment (LCA) midpoints (Global Warming Potential, Ozone Depletion Potential, Acidification Potential, Photochemical Ozone Creation Potential, Abiotic Depletion, Total Waste Disposed, and Consumption of Freshwater), concrete strength classes, declared unit, concrete curing time, production components, concrete manufacturers’ company and plant locations, and additional LCA information. Both the dataset and an example of the Python script used to extract the information from the EPDs are provided. This dataset enables users to quickly assess the environmental impacts (including the Global Warming Potential) of different concrete mixtures without the need for extensive data collection and analysis. In summary, this dataset provides environmental information about concrete mixtures to aid civil engineering and architectural researchers, sustainability consultants, building engineering practitioners, and environmental policymakers to make sustainability-informed decisions when specifying concrete in the U.S.A.


a b s t r a c t
The carbon footprint of a concrete structure is directly affected by the selected concrete mixture proportions.To better understand the influence of different concrete mixtures, data was collected from Environmental Product Declarations (EPDs).Data from 39,213 U.S.A. ready-mix concrete EPDs was obtained from public repositories provided by the American Society for Testing and Materials and the National Ready Mixed Concrete Association.The EPDs in .pdfformat were analyzed using a custom Python script to extract useful information for building designers, sustainability practitioners, and researchers including: life cycle assessment (LCA) midpoints (Global Warming Potential, Ozone Depletion Potential, Acidification Potential, Photochemical Ozone Creation Potential, Abiotic Depletion, Total Waste Disposed, and Consumption of Freshwater), concrete strength classes, declared unit, concrete curing time, production components, concrete manufacturers' company and plant locations, and additional LCA information.Both the dataset and an example of the Python script used to extract the information from the EPDs are provided.This dataset enables users to quickly assess the environmental impacts (including the Global Warming Poten-tial) of different concrete mixtures without the need for extensive data collection and analysis.In summary, this dataset provides environmental information about concrete mixtures to aid civil engineering and architectural researchers, sustainability consultants, building engineering practitioners, and environmental policymakers to make sustainabilityinformed decisions when specifying concrete in the U. .pyfile (a snippet of Python code to extract information from an EPD).

Data collection
The data recorded in the .csvfile was extracted from 39,213 EPDs using a custom Python script (see the .pyfile).Reported LCA midpoints (e.g., Global Warming Potential and Ozone Depletion) were recorded for each concrete mixture and were subdivided by LCA stage (A1-A3, A1, A2, A3).Additionally, the strength of the concrete, the declared unit, product description, and LCI information were collected.The ready-mix concrete producer's company, plant, and plant locations were also extracted from the EPDs.Lastly, the internet links to access the EPDs are provided.

Data source location
The database contains information extracted from EPDs provided online by the American Society for • The dataset contains additional information extracted from the ready-mix concrete EPDs including the mixture description, concrete compressive strength, declared unit, product components, the Life Cycle Inventory products and sources, and the street locations of the ready-mix concrete plants.The locations of the ready-mix concrete producers, including each ready-mix concrete plant, are provided to help consultants identify concrete mixtures that are near their project site.• This dataset can be reused to further understand the environmental effects of the production of different concrete mixtures, which benefits building sustainability consultants, building practitioners, architects, civil and structural engineers, concrete plant manufac-turers, structural concrete institutes (e.g., the American Concrete Institute, Structural Engineering Institute) and policymakers.Statistical analyses can be conducted to provide current baseline values of LCA midpoints for concrete mixtures of various compressive strength classes.Lastly, ready-mix concrete manufacturers can compare nationally the environmental footprint of their concrete mixtures to other ready-mix concrete manufacturers, encouraging the development of more sustainable concrete production.

Data Description
This dataset compiles ready-mix concrete manufacturer, concrete mixture, and environmental information from 39,213 Environmental Product Declarations (EPDs) of various concrete mixtures currently available in the U.S.A.The ready-mix concrete EPDs were downloaded from public repositories provided by ASTM [1] and NRMCA [2] , with a custom Python script (see the accompanying .pyfile) employed to extract key information from the EPDs.The EPDs were followed according to ISO 14040 [4] , ISO 14044 [5] , and ISO 21930 [6] .Table 1 outlines the information obtained from the EPDs and presented in the dataset, including the ready-mix concrete manufacturer, ready-mix concrete plant location, general EPD information, engineering information regarding the concrete mixture proportions, LCA midpoints, and the Life Cycle Inventory (LCI) variability, and sources of LCI data as reported in the EPDs.
A key aim of this dataset is to provide high fidelity information of the environmental impacts for current concrete mixtures.Existing baseline reports such as the CLF 2021 [7] and NRMCA 2022 [8] reports exclude useful information, such as the complete list of product components in a concrete mixture and the primary sources of LCI data.Additionally, this dataset represents a broad distribution of ready-mix concrete EPDs currently available across the U.S.A. Ready-mix concrete EPDs from 37 different ready-mix concrete manufacturers and 389 unique ready-mix concrete plants were collected, as detailed in Table 2 .Fig. 1 shows the geographic distribution of EPDs.The dataset also represents a broad range of concrete compressive strength classes.Fig. 2 shows that though over half of the concrete mixtures have a design compressive strength between 27.6 MPa (40 0 0 psi) to 34.5 MPa (50 0 0 psi), several other strength classes are represented, including those below 17.2 MPa (2500 psi) and above 55.2MPa (80 0 0 psi).Because of the breadth of compressive strengths in the dataset, holistic assessments of strength against LCA mid-points, like Global Warming Potential, can be conducted, as demonstrated by Fig. 3 .Ad-   ditionally, the concrete mixtures also include up to 15 different product components: portland cement, alternative cements (slag cement, type 1 L cement, and hydraulic cement), aggregates (natural, crushed, and lightweight), batch water, admixtures (ASTM C494 and C260), fly ash, silica fume, fiber, glass pozzolan, and pigment.The percentage of concrete mixtures that include each component is shown in Fig. 4 .Note that a single concrete mixture commonly includes sev- eral product components (for example, natural aggregate, crushed aggregate, portland cement, batch water, slag cement, and chemical admixtures).
Two strategies to efficiently parse and analyze the large dataset include the filter feature in Excel and reading the dataset in Python or an equivalent coding language.In Excel or Python, users can view or filter concrete mixtures from a specific concrete plant, geographical location, concrete compressive strength class, or LCA midpoint.Python can further filter the dataset to identify EPD information including material components and LCI sources.Doing so can enable sustainability researchers and building practitioners to analyze the environmental performance of different concrete mixtures currently available in the U.S.A.A study by Anderson and Moncaster [9] is an example of an analysis that can be done with this data.

Experimental Design, Materials and Methods
The information in the ready-mix concrete EPD database is obtained from EPDs made available online from ASTM [1] and NRMCA [2] between February and April of 2023.The internet links to obtain the EPDs are provided in the dataset column: "EPD Source Link."The ready-mix concrete EPDs were downloaded as PDF files and put into subfolders based on the ready-mix concrete producer.The accompanying Python script (Con-crete_EPD_Information_Mining_Python_Script.py) was written to extract the relevant information from each ready-mix concrete EPD automatically, circumventing the need for manually reading, interpreting, and reporting the information for each EPD.The Python script was run for every ready-mix concrete producer to obtain the compiled EPD dataset, as illustrated in Fig. 5 .The Python script was created using Anaconda Navigator 2.4.1, using Jupyter Notebook with Python v. 3.8.The Python script employed the open-access library pdf plumber [10] to extract text from each EPD .pdffile.The provided .pyfile can be employed on any Python environment with Python v. 3.0 or newer.The user can define the desired information for extraction (e.g., the LCA midpoints, ready-mix concrete plant locations), and the Python script then extracts the information from the read .pdffile and saves each entry.Once the for loop completes, the recorded data is written as new .csvfile which can then be assembled to the EPD dataset and cleaned to match the syntax and numerical format across all ready-mix concrete producers.

Limitations
While the EPD database contains many entries, this is not an exhaustive list of concrete mixtures currently available.Second, the dataset includes ready-mix concrete produced only in the U.S.A.; no other countries are considered.Although no other countries are considered, a limitation with the analyzed EPDs is that European LCI audits are used for certain LCI items such as cleaning chemicals, indicating that the environmental performance for the items is similar within the U.S.A., but may contribute to the LCI variability and uncertainty.Third, the quality of the data and the amount of data available differs across concrete plant/ready-mix concrete producers.However, all EPDs are externally validated by a third-party organization.Lastly, EPDs have a five-year period that they are valid.Therefore, EPDs in the dataset may not be valid depending on the time that the future analysis is conducted.The earliest date that an EPD is not valid is April 3rd, 2024.Regarding the provided Python code, it should be noted that ready-mix concrete EPDs from different plants or manufacturers can differ from one another.Therefore, the provided code may need to be modified correctly extract all information from an EPD.

Fig. 1 .
Fig. 1.Distribution of ready-mix concrete plants with EPDs across the U.S.A. in the EPD dataset.

Fig. 2 .
Fig. 2. Breakdown of concrete compressive strength classes in the EPD dataset.

Fig. 3 .
Fig. 3. Concrete Compressive Strength and A1-A3 Global Warming Potential for all EPDs in the dataset.Note that this includes EPDs with different functional units (i.e., different curation times).

Fig. 4 .
Fig. 4. Breakdown of concrete mixture components.Note that the percentage of mixtures that contain any one product component is out of the entire dataset.

Table 1
The column headers in the EPD Dataset, grouped by category.