Tanzania made water prepaid, LUKU-style. Now make it provable.
No African country has embraced prepaid water like Tanzania: roughly a third of urban utilities and over 1,400 villages vend water the way TANESCO vends electricity — tokens and taps, topped up over M-Pesa, TigoPesa and Airtel Money. The collection problem is being solved. The remaining question is physical: does the water dispensed match the tokens sold, is every kiosk and dispenser actually up, and are the meters honest? addanode is the independent verification layer — we don't sell prepaid meters, we prove the fleet.
The gaps a national prepaid wave leaves open.
Vends are recorded; water isn't
The vending platform knows every token sold down to the shilingi. Only an independent flow measurement knows what physically left the pipe. A persistent gap between the two is a bypass, a drifting meter, a passing valve or a software fault — four different fixes, all invisible while the vending system audits itself.
Village dispensers fail in silence
A prepaid tap in a village that's jammed, dry or powerless stops serving — and the operator learns about it from complaints travelling by road. Uptime telemetry per dispensing point (activity, tank level, power, tamper) turns a dead tap into a same-day ticket, protecting both revenue and the community's patience with prepayment.
Meter fleets drift as they age
Thousands of prepaid meters, installed in waves, wear at different rates — and a meter under-registering vends water free. Zone reconciliation (independent inlet metering versus summed vends) exposes drift in aggregate and ranks which clusters deserve testing, so replacement budgets follow evidence.
Fitted to Tanzania's prepaid reality.
- LUKU literacy is the foundation — we build on it. Because TANESCO normalised tokens and USSD top-ups years ago, Tanzanian water prepayment skipped the education curve. What the electricity years also taught customers is to argue with meters; independent verification is how a water scheme wins those arguments with data.
- Mobile money is the rail; we watch the water. Top-ups ride M-Pesa, TigoPesa and Airtel Money into your vending platform. We deliberately stay off the payment side: reconciliation uses the platform's vended-volume exports against our measured flow, keeping the verification independent.
- Village sites are the hard case — and the common one. Solar/battery power, tamper-alarmed enclosures, LoRaWAN or 4G backhaul and local buffering are the standard build, because COWSO-run dispensers don't come with mains power or fibre.
- Utilities and villages need different dashboards, same platform. An …UWASA managing prepaid zones sees zone reconciliation and meter-fleet health; a district council or scheme operator sees per-village dispenser uptime and delivered litres. Both run on addaNet, both export for the regulator and funders.
Adoption figures come from published Tanzanian sector reporting. The verification pattern is our Southern African metering build, adapted to LUKU-style water vending.
A verification layer for a prepaid scheme, in four moves.
1 · Independent metering
Bulk flow at zone inlets and above dispenser clusters — independent of the prepaid hardware, minutes of plumbing per point.
2 · Point telemetry
Dispenser activity, tank level, power and tamper state per kiosk or village tap; solar-powered, buffered.
3 · Reconcile
Measured flow vs vended volumes per zone and point, daily; alerts on gaps, dead points, night draw and tamper events.
4 · Evidence out
Delivered-litres and uptime per point, exportable — for the utility board, the council, EWURA conversations and funders.
A pilot — one prepaid zone or a cluster of village dispensers — is typically live in 2–4 weeks, beside NRW monitoring and water telemetry on one dashboard.
Tanzania's prepaid wave, and where verification fits
| Fact | Gap it leaves | Verification move |
|---|---|---|
| 23 of 70 urban utilities and ~1,400 villages on prepaid meters | Collections fixed; losses still 42.3% nationally | Bulk-vs-sum-of-units per zone |
| Tokens via M-Pesa, TigoPesa, Airtel Money — "LUKU" language borrowed from electricity | Vend records exist; delivered-volume records mostly do not | Continuous delivered volume against vends |
| Village schemes far from technicians | Failed meters and offline token systems go unnoticed | Tank level + heartbeat on every scheme |
| Disputes over over-billing and faulty meters | Only the meter's own reading exists | An independent zone count that has no side |
Prepaid operators we build for in Tanzania.
Urban WSSAs running prepaid zones — utilities in the national prepaid wave whose meters and vends deserve an independent check · Village schemes and their councils — COWSO-run dispensers where delivered-litres evidence protects both revenue and funding · Development-funded programmes — donors and implementers who need verified dispensing data, not estimates, across hundreds of points · Private and institutional suppliers — compound and estate operators vending water where the buy-vs-sell balance is the margin.
Prepaid water verification in Tanzania — common questions
Do you sell prepaid water meters in Tanzania?
No — Tanzania has active prepaid vendors already. We're the layer they can't be: an independent measurement with no meter sales to defend, reconciling what was vended against what physically flowed. Vendor-neutrality is what makes the numbers settle arguments instead of starting them.
How does flow-vs-vend reconciliation work?
An independent flow sensor upstream measures delivered water; your vending platform exports vended volumes; the dashboard compares them per point per day. Theft, meter drift, leaks and software faults each leave a different signature in the gap — and each gets a different fix.
Can you monitor village dispensers with no power or network?
That's the standard case: solar/battery nodes, LoRaWAN or 4G backhaul (satellite where truly remote), local buffering, tamper-alarmed enclosures. A healthy point needs no visits; a failing one raises a ticket the same day.
We already get reports from our vending platform. Why add this?
Because those reports describe transactions, not water. They cannot see a bypass, an under-registering meter or a valve passing overnight — the exact places prepaid schemes leak value. Independent measurement is the difference between an audited scheme and a self-reported one.
What does a verification pilot cost?
A pilot on one zone or a dispenser cluster is quoted as a written cost band after a free remote review of your scheme list and vending platform. It typically pays for itself with the first bypass, drifting meter batch or recovered dead-point revenue it finds.
Does this connect to our NRW programme?
Directly. Prepaid vends versus zone inlet flow is the cleanest NRW split available: collection losses are already squeezed out by prepayment, so the residual gap is physical loss and bypass — see NRW monitoring in Tanzania. One dashboard carries both.
Prepaid solved your collections. Let's verify the water.
Tell us about your zones, dispensers and vending platform. We'll propose a verification pilot that makes the whole scheme provable.