Three-quarters of Tanzania's water loss isn't leaking pipes. The regulator said so itself.
By Frank Guo · Technology & Product Leadership, addanode
EWURA's Water Utilities Performance Review Report for FY2024/25 carries a bad headline number: non-revenue water jumped from 36.8% to 42.3% in a single year. But the sentence that should reorganise every utility's response sits quietly in the NRW section: «The high level of NRW is mainly due to apparent losses, which contributed 75% of the NRW.» Three-quarters of the loss is water that reached someone and was never billed — metering error, unauthorised use, billing gaps. You do not fix that with a pipe crew.
Which edition this reads. This analysis is based on the full PDF of EWURA's Water Utilities Performance Review Report for FY2024/25, published on ewura.go.tz and downloadable directly (linked below), cross-checked against the FY2023/24 edition for trend. EWURA publishes annually; this page will be refreshed within 30 days of the next edition appearing.
One year, 5.5 points, TZS 182.6 billion.
36.8% → 42.3%
Sector NRW in one year — after the previous year had recorded a 0.4-point improvement. The service benchmark is 20% or below. The trend did not slow; it reversed and accelerated.
TZS 182.6 billion
EWURA's own estimate of revenue lost to NRW above the 20% benchmark in FY2024/25 — the excess loss alone, not the total.
75% apparent losses
The regulator's attribution: three-quarters of NRW is commercial — metering inaccuracy, unauthorised consumption and billing gaps — not physical leakage.
For scale: the same report celebrates revenue collection growing 14%, from TZS 419.2 billion to TZS 479.1 billion — an increase of about TZS 59.9 billion. Set that against the TZS 182.6 billion lost above the benchmark and the arithmetic is uncomfortable: the sector's excess water loss costs roughly three years of its revenue growth, every year. Growth is real, and it is being outrun by the leak in the billing system.
Apparent losses are a measurement problem wearing a pipeline costume.
Almost every conversation about African water losses defaults to infrastructure: old pipes, deferred maintenance, bursts. Sometimes that is the story. EWURA's attribution says that in Tanzania, mostly, it is not. Apparent losses are water that was produced, delivered and consumed — and then vanished commercially:
- Under-registering meters — customers billed for less than they draw, silently, for years. The report notes the average metering ratio actually declined by 1.1% in the year, below the 100% benchmark.
- Unauthorised consumption — connections the billing system has never heard of.
- Billing and data gaps — consumption measured at the meter that never becomes an invoice.
The intervention set for apparent losses is entirely different from the one for physical losses. Pressure management and leak crews do nothing here. What works is reconciliation: zone inlet metering compared against the sum of customer billing inside the zone, meter verification against an independent reference, and consumption profiles that flag the connection registering a fraction of its neighbours. Every one of those is a measurement-and-data exercise — the argument set out in non-revenue water in Tanzania and, for the prepaid schemes now spreading on the LUKU model, in prepaid water meter verification.
The cross-country pattern is now three for three. Kenya: 97% metering ratio, 48% losses (WASREB analysed). Zambia: metering up 75.5%→79%, NRW up anyway (NWASCO analysed). Tanzania: metering ratio falling while 75% of a growing loss is commercial. Three regulators, three datasets, one conclusion: counting customers is not the same as measuring the system.
Growth in connections, stasis in service.
- Water connections grew 9% to 1,812,712 and sewerage connections 4.4% to 62,314 — genuine expansion.
- Average service hours stayed at 14 per day, unchanged year on year, against a best practice of 24. Expansion is reaching more people without yet reaching them for longer.
- Operational expenses exceeded revenue at a significant number of WSSAs — the report's own wording — meaning O&M cost recovery remains out of reach for much of the sector even as collections grow.
- Rankings: Moshi WSSA emerged overall best in Cluster I, Nzega in Cluster II and Biharamulo in Cluster III, with Iringa and Utete taking category honours. Per the discipline we apply to every regulator report: the least performers are named in the original document, and we leave them there.
- The report's stakeholder recommendation names the fix in as many words: intensify NRW reduction through targeted investment in infrastructure and smart technologies. Given the 75% attribution, the second half of that sentence is where the leverage is.
The FY2024/25 numbers on one table
| Indicator | FY2023/24 | FY2024/25 | Reading |
|---|---|---|---|
| Non-revenue water | 36.8% | 42.3% | Down 5.5 points in a year; benchmark ≤20% |
| Share of loss that is commercial | — | 75% | "Apparent losses" — a measurement problem, not a pipe problem |
| Value of NRW above benchmark | — | TZS 182.6 billion | About three years of revenue growth |
| Revenue | — | TZS 479.1 billion (+14%) | Growth in connections, stasis in service |
| Metering ratio | — | declined 1.1 points | Fewer customers metered while losses rose |
| Hours of supply | 14 / day | 14 / day | Unchanged |
Three practical readings for a Tanzanian WSSA.
- Diagnose before you dig. If the national attribution holds even approximately at your utility, most of your recoverable revenue sits in commercial loss — findable by reconciling zone input against zone billing, at a fraction of the cost of network rehabilitation. A zone whose night flow is modest but whose billing gap is large has an apparent-loss problem, and now you know which streets to audit rather than which pipes to replace.
- Meter verification is revenue work. A declining metering ratio plus 75% apparent losses means the meter stock itself is part of the leak. Independent verification — comparing a sample of customer meters against a reference measurement — turns "our meters are probably fine" into a number, and usually into a replacement list that pays for itself.
- The record has to be continuous, because the regulator's isn't. EWURA reviews performance annually, in arrears. TOCEMS signals where enforcement is heading — continuous and real-time (see NEMC compliance in Tanzania). A WSSA with continuous zone data doesn't just find losses faster; it walks into tariff reviews and performance reviews with evidence instead of estimates.
Frequently asked questions
What is Tanzania's non-revenue water rate?
Per EWURA's Water Utilities Performance Review Report for FY2024/25, sector NRW was 42.3%, up from 36.8% in FY2023/24 — a 5.5-point deterioration in one year, against a service benchmark of 20% or below. Earlier editions had reported figures around 33%, so the widely quoted "one-third" number is now two editions out of date.
What are apparent losses, and why do they dominate in Tanzania?
Apparent (commercial) losses are water that was delivered and consumed but never correctly billed — under-registering meters, unauthorised connections and billing gaps — as opposed to real losses through leaks and bursts. EWURA attributes 75% of Tanzania's NRW to apparent losses. They dominate where metering accuracy and billing data lag network growth, and they are corrected through measurement and reconciliation rather than through pipe replacement.
How much does NRW cost Tanzanian utilities?
EWURA estimates revenue lost to NRW in excess of the 20% benchmark at about TZS 182.6 billion in FY2024/25. Total sector revenue collection in the same year was TZS 479.1 billion, up 14% — so the excess loss equals roughly three years of the sector's revenue growth.
Which Tanzanian water utilities performed best?
In the FY2024/25 ranking, Moshi WSSA was overall best in Cluster I, Nzega WSSA in Cluster II and Biharamulo WSSA in Cluster III, with Iringa WSSA and Utete WSSA leading their ranking categories.
Why did NRW rise if connections and revenue both grew?
Because the three measure different things. Connections grew 9% and collections 14% — but production grew into a network whose commercial controls lagged: the metering ratio declined 1.1%, and three-quarters of the loss is attributed to metering, unauthorised use and billing gaps. Growth without reconciliation widens the gap between water delivered and water billed, which is precisely what the numbers show.
Where can I read the report itself?
EWURA publishes the Water Utilities Performance Review Reports on ewura.go.tz under Publications → Performance Reports, as directly downloadable PDFs — including the FY2024/25 edition this analysis reads and prior years back to FY2019/20. The primary-sources list below links them.
75% of the loss is commercial. Reconciliation is how you find it.
Tell an engineer your zone layout and billing volumes. You'll get a first-zone reconciliation proposal — input metering against billed consumption, sized for the budget you have.