About 1670 plants have capacity factors above 100%

[duncangeere]

Issue Type

(mark with x between brackets)

• Proposed Data Correction
• New Data Source

Countries

['DZA', 'AGO', 'ARG', 'AUS', 'AUT', 'AZE', 'BEL', 'BEN', 'BRA',
'CMR', 'CAN', 'CHL', 'CHN', 'COL', 'CIV', 'DNK', 'DOM', 'ECU',
'EGY', 'FIN', 'GAB', 'DEU', 'GRC', 'HND', 'ISL', 'IND', 'IDN',
'ITA', 'JAM', 'JPN', 'JOR', 'KAZ', 'LBN', 'LBY', 'MYS', 'MEX',
'MDA', 'MNG', 'NZL', 'NER', 'PAK', 'PAN', 'POL', 'ROU', 'SVK',
'SVN', 'ZAF', 'KOR', 'ESP', 'SWE', 'TWN', 'TZA', 'THA', 'TUR',
'TKM', 'GBR', 'USA', 'UZB']

Affected plant(s)

See full list in this csv.

Database field(s)

capacity, estimated generation

Description

Firstly, thanks for curating such a fantastic dataset. It's proving extremely helpful in an (unfortunately confidential, for now) project that I'm working on.

However, I noticed during my analysis that a small percentage of plants have a mismatch between estimated generation and nameplate capacity that means they have a capacity factor of greater than 100%, i.e. they're more than 100% efficient.

I'm using the following formula to calculate it, which I've double-checked:
(estimated_generation_gwh * 1000) / (capacity_mw * 365 * 24)

Unfortunately I don't have any recommendations for fixing this, or data to correct with, but I wanted to highlight this as it may help you pinpoint a source of systematic error somewhere. Let me know if I can help with any extra info.

Source Information

None

Data Provider

(Select one or more with x between brackets)

• Official Government Data
• Utility/Producer Data
• Non-profit/Independent Group Data
• Unknown Quality

Data Format

(Select one or more with x between brackets)

• Text on web page
• Structured web page (table or regular format for data)
• Machine-readable format (Excel, CSV, XML, ...)
• Human readable document (PDF, Word, ...)

Data Location

(insert URL(s) or source of information)
N/A

Additional Info

N/A

[Whiax]

Issue Type

(mark with x between brackets)

• [ x ] Proposed Data Correction
• New Data Source

Countries

United States of America

Affected plant(s)

WRI1026808

Database field(s)

estimated_generation_gwh

Description

In some fields, the estimated generation is unplausible, some corrections to the model should be made. Let's take two examples: The capacity in Mw for the three gorge dam in China (WRI1000452) is 22.5gw for an estimated generation of 92k gwh, which is plausible because 22.5 * 365 * 24 = 200k and 92k < 200k. For AES Corp plant in Puerto Rico (WRI1026808), the capacity is 0.4gw and the estimated generation is 450k gwh, which is unplausible because 0.436524=3.5k gwh and 450k > 3.5k.
This problem makes the estimated generation unusable.
Other examples: PCA-Valdosta Mill in USA (USA0060084), real generation between 2015 and 2017 is ~300gwh, estimation is 18k gwh. Elizabethtown Power LLC capacity: 34.7mw, estimation: 34kgwh, plausible (cap36524) < 300gwh.

Possible corrections involve thresholding the estimate by the maximum possible value, calculating the average utilization rate of these plants by fuel type and capacity and reinferring their estimate, ...

Source Information

Me crosschecking with plausible informations

Data Provider

• [ x ] Official Government Data
• Utility/Producer Data
• Non-profit/Independent Group Data
• [ x ] Unknown Quality

Data Format

• Text on web page
• [ x ] Structured web page (table or regular format for data)
• Machine-readable format (Excel, CSV, XML, ...)
• Human readable document (PDF, Word, ...)

Data Location

https://www.fluor.com/projects/solid-fueled-power-aes-puerto-rico-epc

[MichaelTiemannOSC]

As the previous poster said, the math estimating generation for WRI1026808 is incorrect. The source link for that powerplant clearly leads to a nameplate capacity of 450MW, so unclear how the estimate can be off by a factor of 1000. Looking forward to seeing a fix in the next version!