AI Analysis: Natural gas supply and disposition, monthly

Executive Summary

Statistics Canada's monthly natural gas supply and disposition dataset (Table 25100086) covers 844 records across 14 Canadian regions and 25 supply/disposition categories for January 2026, revealing a heavily right-skewed distribution where storage inventories and gross withdrawals dominate reported volumes. The dataset exhibits extreme variability, with values ranging from -4.4 million to 26.6 million units and a standard deviation over three times the mean, driven by a small number of very large outliers. Outlier analysis flagged 75 records (15.8% of data points) as statistically anomalous, underscoring the structural diversity across Canada's natural gas supply chain.

Key Findings

• The dataset contains 844 records spanning 14 geographic regions and 25 supply/disposition categories, with 474 non-null data points available for statistical analysis.
• The distribution is strongly right-skewed, with a mean of 883,621 dramatically exceeding the median of 29,252 — indicating a small number of very large values dominate the dataset.
• Values range from a minimum of -4,417,244 to a maximum of 26,575,205, producing an enormous spread of ~31 million units with a standard deviation of approximately 2.88 million.
• Opening inventory (storage), Closing inventory (storage), and Gross withdrawals are the top three categories by average volume, confirming that storage dynamics are the dominant feature of Canada's natural gas system.
• 75 outliers (15.8% of 474 records) were identified using the IQR method, with bounds set at -541,221 (lower) and 906,275 (upper), reflecting the dataset's high structural variability.
• 23 negative values are present in the dataset, likely representing inventory adjustments, line losses, or net withdrawals rather than data errors.
• Data is expressed across four units of measure — Cubic metres, Gigajoules, Canadian dollars, and Customers — contributing to the wide value range and requiring unit-aware interpretation of cross-category comparisons.