I've created 2 new public datasets which may be useful for this competition:\nhttps://www.kaggle.com/vgates/ds4g-puerto-rico-electricity-consumption-by-month\nhttps://www.kaggle.com/vgates/ds4g-puerto-rico-electricity-consumption-by-year
\n\nThe data comes from the U.S. Energy Information Agency (EIA).
\n\nI thought emissions factors might be more believable if it can be shown that emissions correlate with electricity consumption, which you can't do with just the one value from the WPPD data. In particular, Puerto Rico was hit by hurricanes Irma and Maria in September 2017, causing huge disruptions to its electric grid. This probably ought to be reflected in a drop in emissions that corresponds to the drop in electrical consumption seen in the month database.
\n\nAlso, the GPPD \"total estimated generation (GWh)\" appears to be wildly inaccurate. The sum for all plants is 486860 GWh, but Alberto Julian (@ajulian), in this kernel:\nhttps://www.kaggle.com/ajulian/capacity-factor-in-power-plants\nshowed that this is a ridiculous number. He corrects this value as best it can be done with the provided data, and gets 33914 GWh. But the CIA World Factbook:\nhttps://www.cia.gov/library/publications/the-world-factbook/geos/rq.html\nGives 20.95 billion kWh (2016 est.), which is higher than the EIA data I found, but is still much closer to the EIA data than either the raw or corrected numbers from the GPPD.
\n\nI am much more inclined to believe the EIA data than the supplied GPPD data for this.
", "messages": [ { "id": "768297", "postDate": "03/10/2020 16:04:22", "content": "I've created 2 new public datasets which may be useful for this competition:\nhttps://www.kaggle.com/vgates/ds4g-puerto-rico-electricity-consumption-by-month\nhttps://www.kaggle.com/vgates/ds4g-puerto-rico-electricity-consumption-by-year
\n\nThe data comes from the U.S. Energy Information Agency (EIA).
\n\nI thought emissions factors might be more believable if it can be shown that emissions correlate with electricity consumption, which you can't do with just the one value from the WPPD data. In particular, Puerto Rico was hit by hurricanes Irma and Maria in September 2017, causing huge disruptions to its electric grid. This probably ought to be reflected in a drop in emissions that corresponds to the drop in electrical consumption seen in the month database.
\n\nAlso, the GPPD \"total estimated generation (GWh)\" appears to be wildly inaccurate. The sum for all plants is 486860 GWh, but Alberto Julian (@ajulian), in this kernel:\nhttps://www.kaggle.com/ajulian/capacity-factor-in-power-plants\nshowed that this is a ridiculous number. He corrects this value as best it can be done with the provided data, and gets 33914 GWh. But the CIA World Factbook:\nhttps://www.cia.gov/library/publications/the-world-factbook/geos/rq.html\nGives 20.95 billion kWh (2016 est.), which is higher than the EIA data I found, but is still much closer to the EIA data than either the raw or corrected numbers from the GPPD.
\n\nI am much more inclined to believe the EIA data than the supplied GPPD data for this.
", "rawMarkdown": "I've created 2 new public datasets which may be useful for this competition:\nhttps://www.kaggle.com/vgates/ds4g-puerto-rico-electricity-consumption-by-month\nhttps://www.kaggle.com/vgates/ds4g-puerto-rico-electricity-consumption-by-year\n\nThe data comes from the U.S. Energy Information Agency (EIA).\n\nI thought emissions factors might be more believable if it can be shown that emissions correlate with electricity consumption, which you can't do with just the one value from the WPPD data. In particular, Puerto Rico was hit by hurricanes Irma and Maria in September 2017, causing huge disruptions to its electric grid. This probably ought to be reflected in a drop in emissions that corresponds to the drop in electrical consumption seen in the month database.\n\nAlso, the GPPD \"total estimated generation (GWh)\" appears to be wildly inaccurate. The sum for all plants is 486860 GWh, but Alberto Julian (@ajulian), in this kernel:\nhttps://www.kaggle.com/ajulian/capacity-factor-in-power-plants\nshowed that this is a ridiculous number. He corrects this value as best it can be done with the provided data, and gets 33914 GWh. But the CIA World Factbook:\nhttps://www.cia.gov/library/publications/the-world-factbook/geos/rq.html\nGives 20.95 billion kWh (2016 est.), which is higher than the EIA data I found, but is still much closer to the EIA data than either the raw or corrected numbers from the GPPD.\n\n I am much more inclined to believe the EIA data than the supplied GPPD data for this.", "votes": null }, { "id": "770155", "postDate": "03/12/2020 16:13:57", "content": "Thank you for sharing this public dataset in the forums and thank you for converting it to a more useful format (.csv instead of .pdf). https://www.eia.gov is a great place to find reputable external data!
\n\nAs per the potentially erroneous value in the GPPD dataset, for now we encourage you to account for and correct any potential errors or anomalies with the data -- and to describe what you are doing and why you are doing it. An important part of this competition will be defending the validity of your model within the markdown cells of the public notebooks that are used as your submission.
", "rawMarkdown": "Thank you for sharing this public dataset in the forums and thank you for converting it to a more useful format (.csv instead of .pdf). https://www.eia.gov is a great place to find reputable external data! \n\nAs per the potentially erroneous value in the GPPD dataset, for now we encourage you to account for and correct any potential errors or anomalies with the data -- and to describe what you are doing and why you are doing it. An important part of this competition will be defending the validity of your model within the markdown cells of the public notebooks that are used as your submission.", "votes": null } ], "comments": [ { "id": 770155, "author_name": "paultimothymooney", "author_url": "", "post_date": "03/12/2020 16:13:57", "content": "Thank you for sharing this public dataset in the forums and thank you for converting it to a more useful format (.csv instead of .pdf). https://www.eia.gov is a great place to find reputable external data!
\n\nAs per the potentially erroneous value in the GPPD dataset, for now we encourage you to account for and correct any potential errors or anomalies with the data -- and to describe what you are doing and why you are doing it. An important part of this competition will be defending the validity of your model within the markdown cells of the public notebooks that are used as your submission.
", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "768297": "I've created 2 new public datasets which may be useful for this competition:\nhttps://www.kaggle.com/vgates/ds4g-puerto-rico-electricity-consumption-by-month\nhttps://www.kaggle.com/vgates/ds4g-puerto-rico-electricity-consumption-by-year\n\nThe data comes from the U.S. Energy Information Agency (EIA).\n\nI thought emissions factors might be more believable if it can be shown that emissions correlate with electricity consumption, which you can't do with just the one value from the WPPD data. In particular, Puerto Rico was hit by hurricanes Irma and Maria in September 2017, causing huge disruptions to its electric grid. This probably ought to be reflected in a drop in emissions that corresponds to the drop in electrical consumption seen in the month database.\n\nAlso, the GPPD \"total estimated generation (GWh)\" appears to be wildly inaccurate. The sum for all plants is 486860 GWh, but Alberto Julian (@ajulian), in this kernel:\nhttps://www.kaggle.com/ajulian/capacity-factor-in-power-plants\nshowed that this is a ridiculous number. He corrects this value as best it can be done with the provided data, and gets 33914 GWh. But the CIA World Factbook:\nhttps://www.cia.gov/library/publications/the-world-factbook/geos/rq.html\nGives 20.95 billion kWh (2016 est.), which is higher than the EIA data I found, but is still much closer to the EIA data than either the raw or corrected numbers from the GPPD.\n\n I am much more inclined to believe the EIA data than the supplied GPPD data for this.", "770155": "Thank you for sharing this public dataset in the forums and thank you for converting it to a more useful format (.csv instead of .pdf). https://www.eia.gov is a great place to find reputable external data! \n\nAs per the potentially erroneous value in the GPPD dataset, for now we encourage you to account for and correct any potential errors or anomalies with the data -- and to describe what you are doing and why you are doing it. An important part of this competition will be defending the validity of your model within the markdown cells of the public notebooks that are used as your submission." }, "source": "meta" }