Electricity supply reliability
Monthly power-interruption duration and frequency across utility service areas.
- Records
- 4,110
- Vintage
- Monthly observations in the report linked as 2024-Q3; conflicting internal headers are retained · Monthly observations in the report linked as 2024-Q4; conflicting internal headers are retained · Monthly observations in the report linked as 2025-Q1; conflicting internal headers are retained · Monthly observations in the report linked as 2025-Q2; conflicting internal headers are retained · Monthly observations in the report linked as 2025-Q3; conflicting internal headers are retained · Monthly observations in the report linked as 2025-Q4; conflicting internal headers are retained · Monthly observations in the report linked as 2026-Q1; conflicting internal headers are retained
- Last checked
- 16 October 2026
- Published by
- GovernmentPublic Utilities Commission of Sri Lanka
What this dataset contains
PUCSL’s quarterly workbooks publish monthly System Average Interruption Duration Index (SAIDI) and System Average Interruption Frequency Index (SAIFI) values for CEB’s four distribution divisions and LECO. They include utility-province and service-area detail as well as published division totals. This series covers July 2024 to March 2026. The 2024 Q4 CEB sheets separate interruption causes; their total columns are used here.
How it is used here
Track the published CEB division indices over time, compare verified months within an operating geography, or query utility-area observations. Inspect the unit, reported month, reported year and period-status fields before combining reports.
What it cannot tell you
SAIDI measures duration per customer and SAIFI interruptions per customer, not total outage minutes or total incidents. The workbooks include planned outages, breakdowns and transmission failures; licensees use different feeder and customer-count assumptions, and some assumptions omit unreported faults or use planned durations. Read the original Assumptions worksheets when comparing areas. Utility geographies differ from administrative boundaries. Provincial and division indices overlap area observations and cannot be summed or averaged without customer weights. Raw numeric cell values, displayed tokens and cached formulas are preserved; formulas are not recalculated. Hasalaka is labelled Proposed and its source zeros do not establish operating reliability. The LECO sheet linked as 2025 Q1 prints 2024 as its duration year and September in a frequency column; all 42 LECO observations from that file remain undated. Its duration unit is hours per customer, while the other duration sections specify minutes. No silent date correction or unit conversion is made.