POLICY & REGULATION · 8 min read
Sri Lanka’s Reservoirs Have Nearly Halved Since March. The Electricity Tariff Has Not Changed Yet
On 3 August 2026, the Public Utilities Commission of Sri Lanka left electricity tariffs unchanged for the third quarter, citing a projected cost-revenue gap of Rs 417 million — 0.3% of supply cost. Between 28 March and 31 July, storage in the Irrigation Department’s 73 major reservoirs fell from 76.74% of capacity to about 40%. The drought is already affecting the electricity system. It has not yet been reflected in prices. The Q4 decision, due in the quarter the Meteorology Department expects El Niño to peak, is where the two lines meet.
By the ESGNexus Editorial Team · August 2026 · Estimated reading time: 8 minutes
KEY TAKEAWAYS
|
The Q3 tariff decision looks, on its face, like a non-event: no change, a rounding-error gap, business as usual. Read against the water data—it is something else. It is a regulator holding the price steady in the exact quarter the Minister of Environment identified as the severe-drought window, using a finite buffer and a revenue base that the tariff itself does not generate. This piece is about the cost channel of the 2026 drought — the electricity bill — and is a companion to our earlier map of corporate El Niño exposure across sectors, not a repeat.
What PUCSL Decided — and What Held the Price Flat
The Commission’s statement, as reported on 3 and 4 August, followed a review of the National System Operator’s third-quarter cost submission. Its arithmetic is worth setting out in full, because the tariff was held flat by two factors unrelated to the cost of generating a unit of electricity in July.
| Line item (Q3 2026, PUCSL estimate) | Rs million |
| Estimated Q3 supply cost | 159,274 |
| Balance carried forward from Q1 (Rs 30,118 mn less Rs 3,366 mn coal-related costs excluded from tariffs) | 26,753 |
| Total estimated Q3 cost | 186,027 |
| Revenue at the existing tariff | 156,244 |
| Total projected revenue, including 2025 profit share, government consumer subsidies and the NSO allocation | 185,610 |
| Gap: cost over revenue | 417 (0.3%) |
Source: PUCSL statement as reported by Lanka Business Online, 3 August 2026, and Daily FT, 4 August 2026. These are corroborated press reports of the Commission’s statement; the decision document itself is not published on pucsl.gov.lk. The component figures reconcile: 159,274 + 26,753 = 186,027; 186,027 − 185,610 = 417.
Two things stand out. First, the tariff alone does not cover costs. Under the existing structure, revenue is Rs 156,244 million against total costs of Rs 186,027 million. The gap is covered by Rs 29,366 million in revenue from outside the tariff — a share of 2025 profits, government subsidies for consumers, and an allocation for the National System Operator’s operations. Daily FT reported component figures for these three items that do not sum to the reported total; we have therefore stated the components without assigning a figure to them.
Second, outlets describe the carried-forward balance differently. Daily FT and Daily Mirror call the original Rs 30,118 million a first-quarter revenue surplus; Lanka Business Online calls it a deficit. The Commission’s arithmetic adds the adjusted Rs 26,753 million to costs, which is consistent with a deficit being recovered rather than a surplus being spent. We use the neutral term “carried-forward balance” and note that the sign is unconfirmed. Either way, it is a one-off item that will not recur at this size in Q4.
The Reservoirs Tell a Different Story
While the tariff held, the water did not. The Irrigation Department monitors 73 major reservoirs for irrigation, drinking water, and electricity. Its readings in 2026, as reported, show a steady decline that accelerated once the south-west monsoon underperformed.
| Date | Reading | Scope and source |
| 28 March 2026 | 76.74% of capacity | 73 major reservoirs, Irrigation Department, reported by Daily Mirror |
| Early July 2026 | About 65% nationally; hydropower reservoirs 60% | National storage across all managing agencies, Water Management Secretariat of the Mahaweli Authority at a government media briefing, reported by Daily Mirror, 6 July |
| 14 July 2026 | 48% of capacity | 80 reservoirs under the Irrigation Department (a wider universe than the 73-reservoir series), Director of Irrigation (Water Management), reported by Newswire |
| 26 July 2026 | Victoria active storage 35.14% | Single reservoir, reported by Ada Derana |
| 31 July 2026 | About 40% of capacity | 73 major reservoirs, Irrigation Department Director (Hydrology and Disaster Management), reported by Newswire, Lanka News Web and Sunday Times, 31 July–2 August |
Source: Daily Mirror, 28 March and 6 July 2026; Newswire, 14 and 31 July 2026; Sunday Times, 2 August 2026; Ada Derana, 26 July 2026. The March and July 73-reservoir readings come from the same department and the same reservoir universe, so the comparison is like-for-like. The 80-reservoir and national readings are included for texture and are not directly comparable to the 73-reservoir series.
From 76.74% to about 40% is a fall of nearly half in four months. It is not a hydropower-only measure — the 73 reservoirs serve agriculture and drinking water first, and in a drought, those uses take priority over generation. But that is precisely why it matters for electricity: when storage is low, the remaining water is diverted away from the turbines. The Water Management Secretariat said in early July that hydropower reservoirs specifically stood at 60%; no later hydro-specific figure has been published, which is itself a gap.
Why the Reservoir Curve Becomes an Electricity Cost
The link between water and price is not speculative. The Institute of Policy Studies set it out in April: in a good rainfall year, hydropower can supply 40–45% of national electricity; when the monsoons underperform, as they do in some El Niño years, drinking and irrigation needs are prioritised, and the Ceylon Electricity Board ramps up thermal generation. In 2022, thermal plants supplied roughly 47% of total generation, with oil-fired plants absorbing the shock of falling hydro output. In 2025, solar contributed 9.5% to the grid—growing, but not yet large enough to offset a hydro shortfall.
That substitution is the cost channel. Hydro is the cheapest generation in the system; thermal — coal and, at the margin, oil — is the most expensive. Every percentage point of demand that shifts from one to the other moves the supply-cost line that PUCSL prices against. The regulator’s quarterly mechanism is designed to pass exactly that kind of change through to the tariff. What kept Q3 flat was not that the cost line stayed put, but that a one-off balance and non-tariff revenue absorbed the difference. Neither is a renewable resource.
What the Regulator Has Not Told Us
The single most useful number in this story is not published. PUCSL’s Q3 estimate must be based on an assumed generation mix—how much hydro, how much coal, how much oil—for July through September. That assumption determines whether the flat decision already prices in a drought or assumes a normal hydro contribution. Neither the Commission’s statement, as reported, nor its website discloses the assumption. The Commission’s tariff-communications page lists the National System Operator’s bulk supply submission for Q3 as received on 6 April 2026, four months before the decision was announced; the decision itself is not posted.
Two consequences follow. If the Q3 assumption was normal hydro, the drought that arrived in July has not been costed at all, and Q4 inherits both the drought and the gap. If the assumption already reflected low hydro, the Rs 417 million gap is the drought-inclusive figure, and the buffer that absorbed it is what has been spent. A reader cannot tell which. That is a disclosure gap in a regulatory process that affects every business’s electricity bill in the country, and ESGNexus will keep asking about it.
A second gap is data frequency. CEB publishes generation statistics annually. Nothing in the public domain allows a CFO to track the hydro share of generation month by month throughout this drought. Corporate treasurers are being asked to budget for a cost driver they cannot observe.
The Q4 Decision Lands in the El Niño Peak
As reported by Xinhua on 31 July, the Minister of Environment said the country should expect severe dry conditions from July to September and heavy rainfall from October to December, peaking around November. The Director General of the Department of Meteorology, Ajith Wijemanna, has put the probability of the current El Niño strengthening into a strong event in October–December at 81%, citing global climate monitoring, and has said the event could become extremely strong. The Irrigation Department has warned that the combination of August heat and El Niño conditions will drive significant evaporation losses from the remaining storage.
| Until rainfall returns, the water available in reservoirs and other sources must be used carefully. — L.S. Suriyabandara, Director (Hydrology and Disaster Management), Irrigation Department, as reported by Newswire, 31 July 2026 |
Two aspects of the timing merit attention. The Q3 decision covered July to September but was announced on 3 August, a month into the quarter. Following the same pattern, the Q4 decision would arrive at or after the start of October, within the window the forecasts identify as the peak. And the Q4 estimate will be based on the reservoir position at the end of September — after a July and August when the Irrigation Department has told the public to treat it as a water-conservation emergency — with no equivalent of the Rs 26,753 million balance available a second time.
This is not a forecast that tariffs will rise. Heavy rain in October–December could quickly refill the hydro reservoirs, as the Water Management Secretariat noted, happened after last year’s cyclone. However, the decision on Q4 prices will be taken before that rain arrives, based on the numbers available at the end of the dry quarter. The exposure is the Q4 2026 and Q1 2027 energy budgets, and it is asymmetric: the buffer that absorbed the last decision has been used.
What This Means for a CFO on Monday Morning
1. Stress-test Q4 and Q1 energy lines against a thermal-heavy quarter, not Q3. One-off items drove the Q3 tariff. Budget against the supply-cost line, and use the 2022 pattern — thermal at roughly 47% of generation — as the downside case, not a tail risk.
2. Separate the tariff risk from the outage risk. A flat tariff and a stable supply are distinct issues. The 2016, 2019, and 2022 episodes prompted emergency procurement and load management. Site-level continuity plans should be reviewed against reservoir readings, not against the last tariff notice.
3. Ask PUCSL, in writing, which hydro share the Q3 estimate assumed. Industry associations and large consumers have standing to ask. The answer indicates whether the drought has been priced in or deferred. ESGNexus has asked the same question and will publish the response.
4. For SLFRS S2 filers: this is now a physical-risk disclosure item. Drought-driven generation-cost volatility is a current-period climate-related risk with a quantifiable transmission channel. First-wave filers preparing FY2025 and FY2025/26 disclosures have a live, Sri Lanka-specific example that does not need to be borrowed from a global scenario set.
5. Log your own energy-intensity baseline this quarter. If Q4 brings a tariff move, the companies able to show what changed and why will be those that recorded consumption and cost per unit of output beforehand.
What ESGNexus Will Track Next
This article has a shelf life. Its warning expires when the Q4 2026 tariff decision is announced — after that, it becomes a post-mortem on whether the regulator’s Q3 arithmetic priced the drought or deferred it. Until then, ESGNexus will log each Irrigation Department reservoir reading as reported, track whether PUCSL publishes the Q3 decision document or the assumed generation mix, and read the Q4 decision against both. The energy-cost exposure identified here will be carried into the Sustainability Index profiles of first-wave SLFRS filers in energy-intensive sectors as they publish their first mandatory-year reports.
Sources & Further Reading
Daily Mirror, “Electricity tariffs to remain unchanged in third quarter”, 3 August 2026 — dailymirror.lk
Lanka Business Online, “Sri Lanka Electricity Tariffs to Remain Unchanged for Q3 2026”, 3 August 2026 — lankabusinessonline.com
Daily FT, “No electricity tariff changes in 3Q: PUCSL”, 4 August 2026 — ft.lk
Sunday Times (Times Online), “PUCSL says no electricity tariff revision for third quarter of 2026”, 3 August 2026 — sundaytimes.lk
PUCSL, “Communications on Electricity Tariff Revision” (Bulk Supply Tariff Submission for 2026 Quarter 3, received 6 April 2026) — pucsl.gov.lk
Daily Mirror, “Hydro power safe for now, but uncertainty remains: Mahaweli and Irrigation Dept.”, 28 March 2026 — dailymirror.lk
Daily Mirror, “Sri Lanka braces for El Niño impacts due to weather volatility warnings”, 6 July 2026 — dailymirror.lk
Newswire, “Reservoirs at 48% capacity, drinking water unaffected — Irrigation Director”, 14 July 2026 — newswire.lk
Newswire, “Irrigation Department urges public to conserve water amid dry weather”, 31 July 2026 — newswire.lk
Sunday Times, “Urgent call for responsible water use as hot weather makes reservoirs run dry”, 2 August 2026 — Sunday Times.lk
Ada Derana, “Dry spell intensifies across Sri Lanka amid El Niño warning”, 26 July 2026 — adaderana.lk
Xinhua, “Sri Lanka warns of stronger El Niño impact, possible drought and heavy rains”, 31 July 2026 — english.news.cn
Institute of Policy Studies of Sri Lanka (Erandathie Pathiraja), “Powering through Droughts and Crises: Redesigning Sri Lanka’s Energy Resilience”, Talking Economics, 28 April 2026 — ips.lk
| About ESGNexus ESGNexus is Sri Lanka’s independent platform for ESG, CSR, and sustainability intelligence. We track company-level ESG performance, regulatory developments, and sustainability data across Sri Lanka’s listed companies, large unlisted corporates, and state-owned enterprises. All editorial content is independently produced. Sponsored content is clearly labelled. |
Data disclaimer: Information in this article is sourced from publicly available documents. ESGNexus does not independently verify company disclosures. Errors and omissions excepted.