Cut Dhaka WASA non-revenue water below 25 percent before the aquifer forces the choice

Diagnosis

Dhaka's piped water system loses more than a quarter of what it produces before any customer is billed. The curated assessment puts non-revenue water above 25 percent, water that is pumped, treated, and pressurized but never reaches a paying meter because of physical leaks, illegal connections, and unmetered or under-billed use. Every liter lost is a liter of energy, chemicals, and groundwater spent for no revenue and no service.

That waste matters now because the system leans heavily on groundwater, and the note flags aquifer overdraw: the water table is being drawn down faster than it recharges. Leakage and overdraw compound each other. The more the network loses, the more must be pumped to maintain pressure, and the more the aquifer is mined. A utility that loses over 25 percent of its output is effectively running extra wells to fill holes in its own pipes. Left unaddressed, this is not a billing nuisance, it is a slow-moving supply crisis: deeper wells, higher pumping costs, land subsidence risk, and a thinner buffer for dry seasons. The lead responsible body in the registry is the Ministry of Power, Energy and Mineral Resources (MoPEMR), and the energy intensity of pumping that wasted water is precisely why the energy authority has standing here.

Recommended actions

1. Establish a non-revenue water baseline and a binding reduction target. Owner: MoPEMR, through a service-level directive to the utility. Mechanism: a standing water-balance audit (production metering, district-metered-area mapping, and an annual non-revenue water statement). Observable signal: a published, audited non-revenue water figure exists and is repeated quarterly on the same definition, replacing the current "above 25 percent" estimate with a measured number.
2. Carve the network into district metered areas and pressure-manage them. Owner: MoPEMR oversight, utility execution. Mechanism: bulk meters at zone boundaries plus pressure reduction at night, funded as a dedicated capital budget line rather than absorbed into general operations. Observable signal: night-flow readings fall in each completed zone, and the zones with the worst losses are sequenced first.
3. Convert pumping reduction into a verified energy saving. Owner: MoPEMR with the Sustainable and Renewable Energy Development Authority on efficiency, and the Bangladesh Energy Regulatory Commission on tariff treatment. Mechanism: meter the electricity drawn by water pumping and tie leakage cuts to reduced pump runtime. Observable signal: pumping energy per unit of water billed declines as zones are fixed.
4. Cap and license groundwater abstraction to protect the aquifer. Owner: MoPEMR coordinating with the relevant water and groundwater authorities. Mechanism: an abstraction permit regime with monitored well draw, so that leakage savings translate into reduced pumping rather than expanded supply. Observable signal: measured well draw flattens or falls even as service holds.
5. Make new connections legal and metered by default. Owner: utility under MoPEMR directive. Mechanism: a connection regularization drive that converts illegal taps into metered accounts instead of disconnecting demand into the informal market. Observable signal: metered customer count rises faster than produced volume, shrinking the unbilled share.

Sequencing (first 12 months)

Start with measurement: stand up the water balance and the district-metered-area map. Without a defensible baseline, every later claim of savings is contestable and every budget request is unfunded. In parallel, meter pumping energy, because that link is what gives MoPEMR a clean mandate and a hard number to manage against. Once two or three worst-loss zones are metered, begin pressure management there. The baseline unlocks the target, the target unlocks the capital line, and the early-zone results unlock the case for scaling to the rest of the network.

Risks and constraints

The binding constraints are fiscal and political. Capital for metering and zone works competes with more visible projects, so the energy-saving framing is what keeps it fundable. Illegal connections have constituencies; regularization that adds meters is more durable than enforcement that adds grievances. Aquifer caps bite only if abstraction is actually monitored, and a registry that lists an energy ministry as lead for a water utility signals a coordination seam that must be closed by explicit directive, not left to assumption.

Bottom line

Dhaka loses over a quarter of its water before billing while drawing down its aquifer, and the two problems feed each other through the pump. Measure the loss, ring-fence the network into metered zones, and convert leakage cuts into verified energy and groundwater savings, in that order.