I have an op-ed in 獨立評論@天下, the Chinese-language opinion platform of the CommonWealth group, titled 《讓送水的管線也發電！台灣小水力與我們的生活》 — “Let the water pipes generate power too! Taiwan’s small hydro and our daily life”.

Starting point: Shalu

Can water generate electricity before it reaches the tap? The Shalu water distribution centre in Taichung already does it. Water descends from a regulating pond higher up, across a 76-metre drop. Energy previously dissipated in transport now drives a generating set: the same drinking-water infrastructure serves one more purpose.

Commissioned in February 2025 with Formosa Chemicals, the installation was estimated by the water utility at 4.8 million kWh a year, roughly the annual consumption of 1,300 households. That is a commissioning figure, to be checked against real operation — but it gives a domestic scale to an abstract notion.

In Yilan, the 安農萬富 micro plant puts the equipment in an entirely different setting: a watercourse. The Water Resources Agency reports about 750,000 kWh a year, with a fish pass. It is a verifiable case of generation coexisting with river life.

What small hydro has to demonstrate

That Shalu works does not mean a turbine suits every river. A supply main and an open watercourse raise different problems. The farmer will ask whether water still reaches the fields on time; the resident, how much is left at low water. Those questions need answers before construction, not after.

Priority should go to structures that already exist and need the least new civil works, with security of supply coming first. Small does not mean without effect: if a few extra kilowatt-hours cost too much in irrigation or river ecology, they are not worth having.

Taiwan does not lack leads: the Water Resources Agency has published 27 priority sites — existing weirs, headrace channels, pipelines. A list of sites to study is not a list of plants to build.

Two proposals

1. A public, readable specification for each site. Monthly flows, the volumes reserved as a priority for drinking water and irrigation, planned maintenance outages, grid connection terms. Publish what is measured, flag what is not. The operator can then cost the project; the resident can see whether generation touches the water they use. Seeing a strong flow in the rainy season is not enough to decide: that is judging a restaurant on its Sunday service.

2. Pool what can be pooled between neighbouring sites. A single managing authority can prepare joint tender documents, and operators share inspection teams and spare parts — because small output does not bring small maintenance costs: a breakdown still needs a technician, a part and a call-out. Local firms should be able to bid as a group. Worth piloting, without bundling in contested sites to make up the numbers.

Baseline surveys can be prepared once by the managing authority and opened to all bidders, rather than redone and paid for by each of them.

What residents actually ask

Less who wins the contract than who fixes the problems afterwards: who maintains it, what the area gets in return, who removes the installation at end of life. Without setting that down in advance, a deal that looks good today leaves rusting machinery tomorrow and a bill nobody wants to pay.

My conclusion: small hydro will not lower household bills and will not solve Taiwan’s supply problem on its own. Its value lies elsewhere — recovering energy currently lost, where the cost is reasonable and neither drinking water nor the environment suffers. Shalu shows that moving water and generating power can sit in the same facility.

→ Read the op-ed on 獨立評論@天下 (in Chinese)