The Water Money Is Going to the Right Problem. The Right Problem Is 17% of the Total.

 The system is working just not on the part that matters most.

There is a version of India's water story that is straightforward and true.

The country treats roughly 28% of the wastewater it generates daily. That gap between 72,000 million litres produced and the fraction actually cleaned before release is one of the cleaner infrastructure problems a government can be handed.

It has a measurable size. It has a known solution. It responds to capital.

The Ministry of Jal Shakti has tripled its budget allocation between FY19 and FY26, from Rs 31,000 crore to Rs 99,500 crore. A set of listed companies is building the plants, growing their order books, and in some cases earning margins that would look respectable in much less operationally complex businesses.

That version of the story is accurate.
It is also the version that stops too early.

The number that changes the picture is 83

Insert after this line:
“The number that changes the picture is 83.”

India uses approximately 740 billion cubic metres of water annually.

Agriculture takes 83% of that.
Industry, which generates 30% of GDP, uses 8%.

The treatment plants being built, the contracts being awarded, the order books being watched all of it addresses domestic and industrial water demand.

That is 17% of total use.

The remaining 83% is agriculture.

And it operates almost entirely outside the reach of any EPC contract that has been or could be awarded.

Paddy, sugarcane, and wheat together account for 80% of all irrigation water. Producing one kilogram of rice requires approximately 2,500 litres. One kilogram of sugar requires 3,000 litres.

India grows both at scale well beyond domestic nutritional requirements, much of it for export or domestic price support programs, with water priced at near zero in most state procurement frameworks.

The water embedded in those exports leaves the country without appearing in any balance sheet.

India's irrigation system runs at 38% efficiency. The global average sits at 50–60%.

That gap represents more lost water than the entire industrial sector uses.

Closing it would do more for India's water position than any volume of treatment plant construction.

None of the listed water companies has a product for that.

This is not a criticism.
It is a boundary.

The collision is happening in the same geography

Insert after this line:
“What makes this more than an accounting observation…”

A semiconductor fabrication plant consumes approximately 38 million litres of ultra-pure water per day.

That is the daily water consumption of 62,000 urban households.

A mid-sized data centre uses around 5 million litres daily.

These facilities are arriving in Gujarat, Uttar Pradesh, and Punjab states, already classified as water-stressed by India's Central Water Commission.

The geographic concentration is not accidental.

It reflects where land acquisition is manageable, where power infrastructure exists, and where policy support is most concentrated.

Water cost does not appear among them.

So the picture that emerges is this:

The government is funding water treatment capacity in states where industrial water demand is rising from industries the same government is incentivizing.

The treatment capacity addresses 17% of demand.
The industrial arrivals compound pressure on the 83% that treatment does not reach.

Nobody designed this collision.

It is what happens when two policy programs run in parallel without a shared accounting for water.

The projections that do not include this yet

NITI Aayog has estimated that by 2030, domestic water demand will exceed supply by 50 billion cubic metres.

By 2050, total demand could reach double the available supply.

The World Bank projects India's urban population could double to roughly 950 million.

These projections are built on current trends.
The industrial buildout is not fully inside them yet.

What the financials quietly reveal

6

Insert after this line:
“The financial picture inside the sector…”

In FY25, only Wabag converted its net profit into actual operating cash.

EMS reported a cash flow from operations to PAT ratio of 0.18.
Welspun and Enviro Infra reported negative operating cash flows.

Enviro Infra's EBITDA margins of 25–26% are genuine.

The cash behind them has not arrived yet.

It is sitting in receivables while government payment cycles run their course.

This is not mismanagement.

It is the operating condition of a sector whose primary customer pays slowly.

Every company here is doing two things at once:

• Building infrastructure
• Financing government working capital

Wabag has moved toward a different model.

It owns no construction machinery.
Its assets are technology and people.

It earns roughly 18% of revenue from Operations and Maintenance contracts.

These run up to 15 years, generating recurring cash flows at 30–35% EBITDA margins.

The shift toward O&M is real.
It is also slow.

Where desalination fits and where it doesn’t

Insert after this line:
“Desalination sits alongside all of this.”

India has 11,000 kilometres of coastline.

Desalinated water costs roughly 25% more than pipe-supplied freshwater.

Energy accounts for about one-third of the cost.

That makes long-term economics uncertain.

The technology exists.
Scale does not.

What this actually means

The water sector investment thesis is not wrong.

It is incomplete.

Government spending is real.
Order books are real.
The O&M transition is improving revenue quality.

But all of it applies to 17% of the problem.

The part that sits outside the thesis

The remaining 83%:

• agriculture
• irrigation efficiency
• crop pricing
• groundwater depletion

These require a different system.
Different incentives.
Different actors.

They are not addressable through EPC contracts.

The conclusion that holds both realities

India is building water treatment capacity and semiconductor fabs
in the same stressed geography
at the same time.

One is described as solving the water problem.
The other is not described in water terms at all.

Both descriptions are accurate.

Final thought

The sector is solving what it can measure.
The system is breaking where it does not look.

And both are happening together.