Madhya Marsyangdi

Page status
Working page
Updated
2026-05-13

Open in the interactive explorer and map

Madhya Marsyangdi (Middle Marsyangdi) is a 70 MW operating hydropower project on the Marsyangdi River — one of Nepal's established generation assets with a multi-decade operational track record.

Summary

Madhya Marsyangdi is a 70 MW peaking run-of-river hydropower project operating on the Marsyangdi River in Lamjung district, Gandaki Province. Commissioned in December 2008, it is the middle anchor of the ~170 MW Marsyangdi cascade and recorded 107.6% of design generation in FY 2081/82. The page tracks it as an operating-plant performance benchmark using the current NEA annual report source set.

Marsyangdi Cascade

Madhya Marsyangdi sits in the middle of one of Nepal's most productive hydropower corridors. The Marsyangdi River, a tributary of the Trishuli/Narayani system in the Gandaki basin, hosts a cascade of projects: Upper Marsyangdi A (50 MW), Madhya Marsyangdi (70 MW), and Marsyangdi Besi (50 MW, under construction). This ~170 MW cumulative capacity makes the Marsyangdi cascade a significant block of the national generation portfolio.

Operating History

As an operating plant, Madhya Marsyangdi provides a different lens from the pre-COD and under-construction projects that dominate the pipeline: actual generation data, real maintenance costs, and historical PLF performance under the standard PPA framework. Operating projects reveal whether the financial models in IPO prospectuses materialize in practice — and at what cost.

Transmission

The Marsyangdi valley's generation feeds into the national grid via the Marsyangdi corridor infrastructure, connecting to the east-west backbone. The Upper Marsyangdi 220 kV line provides the key evacuation route for this cascade.

Specifications

Parameter Value
Capacity 70 MW
Status Operating
Registry licence Operation
River Marsyangdi (Gandaki Basin)
District Lamjung
Province Gandaki Pradesh

Specifications (NEA FY 2081/82)

Middle Marsyangdi (MMHPS) is a 70 MW peaking run-of-river plant in Lamjung District, Gandaki Province — the middle anchor of the 170 MW Marsyangdi cascade. Commissioned in December 2008, it diverts flow from the Marsyangdi River originating in glacier-fed Tilicho Lake and can reserve flow for up to five hours of peaking, even during low discharge.

Parameter Value
Type Peaking run-of-river (PROR)
Units Francis
Commissioning December 2008
Peaking capacity 5 hours
Design generation 398 GWh/yr
FY 2081/82 generation 427.9 GWh
FY 2081/82 vs design 107.6%
Year-over-year change −3%
Cumulative lifetime generation 7,091.1 GWh
District Lamjung, Gandaki Province

In FY 2081/82, MMHPS completed overhaul of Unit 2, restored desander basins in-house, rebuilt spillway chute no. 1 with M60 concrete and steel lining, and installed a digital governor panel, 110V battery bank, and RTU for real-time Load Dispatch Centre communication.

The Marsyangdi Cascade — 170 MW in One Valley

MMHPS sits between Upper Marsyangdi A (50 MW, operating, upstream) and Marsyangdi Besi (50 MW, under construction, downstream) — a cumulative 170 MW in the Marsyangdi valley of the Gandaki basin. This is one of Nepal's most concentrated generation corridors. The Marsyangdi River, a glacier-fed tributary of the Narayani system, provides more reliable base flow than the monsoon-dominated rain-fed tributaries where most private IPPs are located.

The cascade's cumulative water use is increasingly significant as Marsyangdi Besi approaches commissioning. Environmental flow requirements and sediment management decisions at Upper Marsyangdi A directly affect MMHPS and the downstream plant. The Marsyangdi corridor connects to the national grid via the Upper Marsyangdi 220 kV transmission line, feeding the east-west backbone.

The Contrast That Matters — Design vs Actual

MMHPS's 107.6% of design generation (427.9 GWh against 398 GWh) is the highest performance ratio in the NEA large-plant fleet. Compare:

Plant MW Design (GWh) FY 2081/82 (GWh) % of Design
Middle Marsyangdi 70 398 427.9 107.6%
Kali Gandaki A 144 842 863.5 102.5%
Marsyangdi 69 462.5 451.8 97.0%
Chameliya 30 184.2 142.4 77.3%
Upper Trishuli 3A 60 489.8 356.0 72.7%

The divergence at the bottom of the table is Q-Design Discharge in action: the plants that underperform by 20–30% are the ones whose design hydrology assumptions were too aggressive. MMHPS's opposite result suggests the Tilicho-fed Marsyangdi basin has more reliable baseflow than the monsoonal rain-fed catchments where design overestimation is endemic — or that the 2008-era design hydrological study was appropriately conservative.

This is the structural pattern the Chilime page documents at 22 MW: conservative Q-design (Q65+) in a snow/glacier-fed basin produces actual generation above contracted energy. MMHPS demonstrates the same mechanism at 70 MW — and at three times the scale. See Q-Design Discharge for the full generation performance comparison across the RoR fleet.

Renewal Cycle — A Younger Plant with a Different Problem

At 16 years post-commissioning, MMHPS is entering the phase where routine overhauls begin to graduate toward major rehabilitation. The FY 2081/82 work scope — Unit 2 overhaul, desander basin restoration, spillway chute rebuild with M60 concrete and epoxy coating — is heavier maintenance than the plant would have required in its first decade.

The digital governor panel and RTU installation signals a shift from mechanical to digital control, the modernization trajectory that aging plants like Kali Gandaki A are now facing at greater cost and outage risk. MMHPS's relatively young age means this transition can be done incrementally rather than as an emergency rehab — but it also means it is a test case for what that incremental path looks like. Every NEA plant eventually follows it.

Sources

Limitations & Controversies

The current page is strongest on NEA annual-report and cascade-level context. It does not independently verify long-term maintenance cost trends, remaining useful life estimates, environmental flow compliance, or sediment-management records beyond the existing source set. FY 2081/82 generation data is a single-year snapshot; multi-year performance trends would require additional verification. The renewal-cycle discussion is forward-looking and not yet grounded in completed rehabilitation data.

See also