Open-access energy demand data for South and Southeast Asia

Coombes, Oliver G.; Hunt, Kieran M. R.; Bloomfield, Hannah C.

doi:10.5194/essd-2026-76

Preprints

https://doi.org/10.5194/essd-2026-76

Preprints

26 Feb 2026

| 26 Feb 2026

Status: this preprint is currently under review for the journal ESSD.

Open-access energy demand data for South and Southeast Asia

Oliver G. Coombes, Kieran M. R. Hunt, and Hannah C. Bloomfield

Abstract. Open-access electricity demand data are essential for meteorology–energy and climate–energy research, forecasting, and resilience planning. Yet in South and Southeast Asia (SASEA), such records are fragmented across sources, reported in inconsistent formats, and often difficult to find or access. This is a serious limitation in a region where electricity use and system stress are sensitive to monsoon variability, humid and dry heat, and other natural hazards.

In this paper, we present and describe a harmonised electricity demand dataset for twelve SASEA countries (Bangladesh, Bhutan, India, Malaysia, Myanmar, Nepal, Oman, Philippines, Singapore, Sri Lanka, Taiwan, and Thailand) at daily national resolution, spanning 2013–2025 with country-dependent coverage. We compiled raw data from national utilities, regulators, and international providers using reproducible retrieval workflows (e.g., APIs and automated scraping of PDFs/XLS/web portals). All records were standardised to megawatt-hours (MWh), and aligned to local-calendar daily totals (i.e., midnight-to-midnight in local standard time).

To support and encourage transparent downstream use, we also provide the raw extracted series, alongside harmonised daily aggregates, metadata, and our processing and scraping scripts. We also publish diagnostics quantifying completeness, gaps, and outliers flagged using a range of statistical methods. Independent validation against Ember monthly electricity demand shows strong agreement in temporal variability for most countries.

Our open dataset will enable regional and cross-country analysis of demand seasonality, growth and variability; evaluation of weather–demand sensitivity using reanalysis or forecasts; and event-based studies of disruption and recovery during extremes. We finish with a short case study application of our dataset and discussion on how it should and should not be used.

Received: 28 Jan 2026 – Discussion started: 26 Feb 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Oliver G. Coombes, Kieran M. R. Hunt, and Hannah C. Bloomfield

Status: final response (author comments only)

Subscribe to comment alert

RC1:
'Comment on essd-2026-76', Anonymous Referee #1, 23 Apr 2026
This paper presents a harmonized, open-access dataset of daily national electricity demand for 12 countries in South and Southeast Asia (SASEA), covering the period 2013–2025 with country-specific availability. The authors compile data from diverse sources, including national utilities, regulators, and international providers, using reproducible workflows such as APIs and automated scraping. The dataset is standardized to consistent units (MWh) and aligned to local daily totals. They also include diagnostic analyses of data completeness, gaps, and outliers, and validate the dataset against Ember monthly demand data. The resulting dataset offers a valuable resource for studies of electricity demand variability, weather–demand relationships, and energy system responses to extreme events in a data-sparse but climate-sensitive region. Considering the publication, the authors are asked to address the following general and specific comments.

General comments:
Fig 1 presents mean electricity demand (including per capita values), but the definition of the averaging period requires clarification. The authors should explicitly state the exact time range used to compute the mean for each country, and whether a consistent temporal window is applied across all countries. This is particularly important given the strong upward trends in electricity demand associated with economic growth across many SASEA countries. If the averaging periods differ between countries (e.g., due to data availability), the resulting means may not be directly comparable and could introduce systematic biases. The authors are therefore encouraged to (i) clarify the temporal coverage used in this figure, and (ii) discuss how differences in data availability and temporal trends may affect cross-country comparability. If applicable, it would be helpful to consider using a common overlapping period or otherwise justify the chosen approach.

Section 4.5 uses Ember electricity data as an independent validation source. The authors should clarify the underlying data sources of Ember and assess whether there is any overlap with the sources used in this study. If substantial overlap exists (e.g., IEA data), the validation cannot be considered fully independent and this limitation should be explicitly acknowledged and discussed.

Section 7.2, the description of data quality control could be strengthened. The authors are encouraged to provide a more systematic account of QC procedures (e.g., treatment of missing values, outlier detection, and consistency checks), as well as guidance for users on data reliability and appropriate use.

The manuscript would benefit from a clearer discussion of the potential applications of this dataset, in order to better highlight its value to the research community.

Specific comments:
1) On page 4 Fig 1: Abbreviations (e.g., OMN, IND, LKA) are not defined at first occurrence. Please ensure that all abbreviations are clearly defined when first introduced in the text or figures.
2) On page 5 Table 1: The use of the term “country” may not be strictly appropriate in all cases (e.g., Taiwan). The authors are encouraged to revise the terminology throughout the manuscript accordingly.
3) On page 7 Table 2: Table 2 should be relocated to Section 3.3, rather than being placed in Section 4.
4) On page 9 Fig. 2: The current UpSet plot is not easy to interpret. The authors are encouraged to improve the figure design or consider an alternative visualization that more clearly communicates the overlap between outlier detection methods.
5) On page 18 Fig. 9: It is recommended that the colors in the legend be consistent with those of the corresponding curves in the Fig. 9.
Citation: https://doi.org/10.5194/essd-2026-76-RC1
RC2: 'Comment on essd-2026-76', Anonymous Referee #2, 02 Jun 2026

The authors have constructed a very interesting electricity demand dataset with important potential applications. I also appreciate the substantial effort required to harmonize multiple databases and data sources. However, I think several aspects of the manuscript remain unclear and need further clarification.
1. L96: “Generation was taken as a proxy for demand because grid-scale electricity storage remains minimal across SASEA.”

This assumption needs to be explained in more detail. In particular, the authors should discuss the possible impacts of electricity storage, imports/exports, transmission losses, and different reporting boundaries across countries. This issue is also relevant to the comparison with Ember, where the authors mention that discrepancies may be caused by different definitions or boundaries. For example, the MAPE for Malaysia reaches 29.6%, which seems quite high, but the manuscript does not provide a detailed explanation. A country-by-country clarification of what each original series represents would be very helpful.
2. L124: “Daily values deviating more than ±0 standard deviations…”

Please check this statement. If “±0 standard deviations” is not a typo, the outlier detection method would be difficult to understand.
3. Figure 9

The final subplot in Figure 9 is unclear for me.
4. Daily-resolution analysis

The daily resolution is one of the main strengths of this study, but the manuscript does not clearly show daily time-series curves. Most current analyses are based on monthly, seasonal, or weekly averages. I suggest adding representative daily demand curves for the countries, which would help readers assess whether the data is reasonable and how missing values, outliers, holidays, and special events appear in the dataset.
In addition, the authors could consider adding more daily-scale analysis, for example highlighting holiday periods or the COVID-19 period. For a large country such as India, it may also be interesting to compare the daily demand series with other independent daily activity indicators, such as Carbon Monitor daily emissions, if appropriate.
5. Interpretation of Figure 8

The patterns in Figure 8 deserve further discussion. Some countries seem to show their lowest demand during the middle of the week, which is somewhat counterintuitive. The authors should examine whether this pattern is related to holidays, missing data, reporting conventions, or other country-specific factors.
For the monthly patterns, the authors could also consider using ERA5 or other meteorological data to explore the relationship between electricity demand and temperature. This would strengthen the climate–energy relevance of the dataset.
6. Dataset updates

Please discuss whether and how this dataset will be updated in the future. For example, India appears to be updated only until 2024, while other countries have more recent records. It would be useful to clarify whether this is due to source availability, data-access limitations, or the current processing workflow.
Additional minor comments:

1. Some figures need further improvement. In some cases, the font size is too large(Figure 2), while in others the text is too close to the axes. Please improve the readability and consistency of the figures.

2. Since Taiwan is included in the dataset, I suggest using a more neutral term such as “countries/regions” rather than referring to all study units simply as “countries”.

Citation: https://doi.org/10.5194/essd-2026-76-RC2

Oliver G. Coombes, Kieran M. R. Hunt, and Hannah C. Bloomfield

Data sets

SASEA Electricity Demand Data for 2013-2025 Oliver Coombes, Kieran Hunt, Hannah Bloomfield https://doi.org/10.5281/zenodo.17175212

Model code and software

South and Southeast Asia (SASEA) Electricity Demand Data for 2013–2025 Oliver Coombes https://github.com/C4ntasaur/sasea-demand

Oliver G. Coombes, Kieran M. R. Hunt, and Hannah C. Bloomfield

Viewed

Total article views: 758 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
387	337	34	758	34	41

HTML: 387
PDF: 337
XML: 34
Total: 758
BibTeX: 34
EndNote: 41

Views and downloads (calculated since 26 Feb 2026)

Month	HTML	PDF	XML	Total
Feb 2026	93	23	4	120
Mar 2026	122	83	18	223
Apr 2026	96	82	4	182
May 2026	66	136	5	207
Jun 2026	10	13	3	26

Cumulative views and downloads (calculated since 26 Feb 2026)

Month	HTML	PDF	XML	Total
Feb 2026	93	23	4	120
Mar 2026	122	83	18	223
Apr 2026	96	82	4	182
May 2026	66	136	5	207
Jun 2026	10	13	3	26

Viewed (geographical distribution)

Total article views: 754 (including HTML, PDF, and XML) Thereof 754 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 09 Jun 2026

Short summary

We assembled and openly shared daily electricity use records for twelve countries in South and Southeast Asia from 2013 to 2025. Data were gathered from official reports and agency websites, converted to a common format and time zone, and checked for missing days and extreme values. Our published dataset largely matches an independent monthly source. The dataset will support future studies of demand growth, weather effects, and electricity system planning and resilience.


Total:	0
HTML:	0
PDF:	0
XML:	0