Singapore: thermal generation doesn't account for oil and biomass, and configuration wrongly assumes all Steam Turbine generation is renewable

[q--]

When did this happen?

The unknown='renewable' part since 2020, the rest even longer (possibly from the beginning).

What zones are affected?

SG

What is the problem?

• Since 2020, the SG parser wrongly assumes all Steam Turbine generation is renewable Added singapore to 'unknown' fossil free exceptions #2423 (comment). This accounts for 2% of total generation (ElectricityMaps 2023)
• The SG parser assumes certain generation types imply natural gas. They do not.
  • All CCGT/COGEN/TRIGEN and Gas Turbine generation is mapped to gas, even though CCGT (Closed Cycle Gas Turbines), Cogeneration, and other Gas Turbines can run on oil as well (as is the case in e.g. South Africa). (Or biogas.)

The actual energy generation breakdown according to Our World in Data

For comparison, ElectricityMap incorrectly reports 96.2% gas and 2.65% 'unknown' generation for the same year. (As well as solar.)

Proposed solution

Map all non-solar generation to 'unknown' and add a custom carbon intensity override, like was done with e.g. El Salvador (#6412) and other countries.

[jarek]

The SG parser assumes certain generation types imply natural gas. They do not.

All CCGT/COGEN/TRIGEN and Gas Turbine generation is mapped to gas, even though CCGT (Closed Cycle Gas Turbines), Cogeneration, and other Gas Turbines can run on oil as well (as is the case in e.g. South Africa). (Or biogas.)

Is there any evidence to indicate that they run on oil in Singapore? Certainly I would be surprised if a small country with well-developed infrastructure reported "gas turbine" generation but actually fired it with oil.

The actual energy generation breakdown according to Our World in Data
[chart showing gas at 93.79% in 2022, oil at 3.80%, 'bioenergy' at 0.69%, and solar at 1.72%]

Should we use Our World in Data data over operator-provided data in this case?

https://www.ema.gov.sg/resources/statistics/plant-mix-for-electricity-generation with linked PDF or XLS https://www.ema.gov.sg/content/dam/corporate/resources/statistics/files/excel/Plant-mix.xls.coredownload.xls gives following averages for 2022: "combined-cycle plants" 95.61%, solar 1.17% (though noting that it's estimated), "others" 3.08%, and "steam plants" 0.14% averaged over year (though it went to 0 in September 2022 and stayed at 0)

https://www.ema.gov.sg/resources/statistics/fuel-mix-for-electricity-generation showing "the contribution of the various fuel types towards gross electricity generation" gives "natural gas" as source for 92% of generation in 2022, "petroleum products" as 2.57%, and "others" as 5.41% (so assuming solar is around 1.4%, that's about 4% of "others" unaccounted for).

I wonder if some of the differences is due to local generation at big industry locations?

[yujia21]

A bit more of a breakdown for 2022 here: https://www.ema.gov.sg/resources/singapore-energy-statistics/chapter2 that specifies that Coal is 1% and others 4.4%.

[q--]

Is there any evidence to indicate that they run on oil in Singapore? Certainly I would be surprised if a small country with well-developed infrastructure reported "gas turbine" generation but actually fired it with oil.

To clarify, gas turbine is a type of engine. The word 'gas' doesn't mean 'natural gas' here, but that the working fluid is in gaseous form. Quoting ChatGPT:

In a gas turbine, air is compressed, mixed with fuel, and then ignited to create a high-temperature, high-pressure gas. This high-energy gas is then expanded through a series of turbine blades, which are connected to a shaft. As the gas expands through the turbine, it drives the turbine blades and, in turn, the shaft. The rotating shaft is used to perform work, such as turning an electrical generator or propelling an aircraft.

See also https://en.wikipedia.org/wiki/Combined_cycle_power_plant#Fuel_for_combined_cycle_power_plants

[q--]

A bit more of a breakdown for 2022 here: https://www.ema.gov.sg/resources/singapore-energy-statistics/chapter2 that specifies that Coal is 1% and others 4.4%.

Interestingly, most of the 'other' generation is further specified in the Excel files one can download at the right of the page.

(This is from the "Tidy Data (XLSX, 4 MB)" file, tab T2.1. Note that it also has e.g. capacity data.)

Unfortunately it's only biomass and municipal waste, but it's still a significant part of 'other' generation.

[VIKTORVAV99]

@mathilde-daugy do you have some insights or opinions here?
Not sure if we should change CCGT/COGEN/TRIGEN to unknown (from gas) or not?

[mathilde-daugy]

@VIKTORVAV99 I will investigate this in more detail tomorrow. Right off the bat I would say that we should not make any hasty changes because our parser is a live parser and we won't be able to backfill the data :)

After a first read, I do think that we should aggregate all non solar production and adjust the EF!