In the U.S. alone, 2.7 trillion gallons of treated water are lost every year. For utilities, that translates to more than $6.4 billion in unrealized revenue annually, according to Bluefield Research.
Those figures are easy to skim past. It's easy not to talk about water loss.
But inside a utility, these metrics show up in very tangible ways. Pumps run longer than they should. Chemicals are used to treat water that never reaches a customer. System capacity is consumed without generating revenue. And over time, infrastructure appears to underperform without a clear explanation.
At that scale, non-revenue water is not just a technical issue. It is one of the clearest indicators of how well a utility understands its own system.
Non-revenue water is a measurement problem before it is a loss problem
Non-revenue water (NRW) is often described simply as water that is lost from the system.
In reality, it is broader than that.
NRW includes all water that is produced but does not generate revenue, whether it is actually physically lost, inaccurately measured, or intentionally unbilled for any number of reasons. It is typically calculated as the difference between system input volume and billed consumption, expressed as a percentage of total production.
That simplicity is part of the problem.
Because once everything is grouped into a single number, it becomes difficult to distinguish what is actually happening inside the system.
Some of that water is leaking from pipes. Some of it is reaching customers but not being billed correctly. Some of it is being used legitimately but not tracked as revenue. Each of those conditions requires a different response.
Without separating them, utilities are left managing a symptom rather than a cause.
The difference between losing water and losing revenue
At its core, NRW is made up of two primary categories.
The first is real losses. These are the physical leaks that occur throughout the system, from distribution mains to service connections and storage infrastructure. This is water that has already been treated and pressurized, and then never reaches a customer. The cost of those losses is not just the water itself, but the energy, chemicals, and capacity required to produce it.
The second is apparent losses. This is water that does reach the customer but is not properly measured or billed. In many systems, aging meters, data handling issues, or billing inaccuracies quietly erode revenue year after year. From a financial standpoint, these losses can be just as significant as leakage.
There is also a third, smaller category: authorized but unbilled consumption. This includes activities such as firefighting or system flushing. These uses are necessary, but they still contribute to the overall gap between production and revenue.
Understanding how these categories interact is what separates a meaningful NRW analysis from a surface-level metric.
Why the NRW percentage alone can be misleading
Most utilities track NRW as a percentage.
But that number, on its own, often creates more confusion than clarity.
A system reporting 10% NRW could be highly efficient, or it could be masking significant leakage behind high consumption. A system reporting 25% NRW could be struggling operationally, or simply operating across long transmission distances with low customer density.
The percentage does not account for system layout, pressure conditions, infrastructure age, or the cost of producing water.
Which leads to a more useful way to think about NRW: It is not exactly a performance score. It is more like a signal. Without context, it can easily point utilities in the wrong direction.
The water audit: Turning a percentage into insight
To move beyond that limitation, utilities rely on the American Water Works Association’s water audit methodology. (Download AWWA's free water audit software, too.)
A water audit breaks the system into a structured balance. It accounts for how much water enters the system, how much is billed, and where losses occur across both real and apparent categories.
More importantly, it forces utilities to evaluate the quality of their own data.
In many cases, the most important outcome of an audit is not the NRW percentage itself, but the realization that the underlying data is incomplete or unreliable.
For many utilities, the first step in reducing NRW is simply improving visibility.
The economic level of leakage
This shift is captured in the concept of the Economic Level of Leakage (ELL).
ELL represents the point at which the cost of reducing water loss further is equal to the value of the water saved. Below that point, additional investment may not be justified. Above it, utilities are effectively losing money by not acting.
This reframes NRW from a technical target into a financial and operational decision.
In water-scarce regions, where supply is constrained and water has high value, aggressive leak reduction is often warranted. In regions where water is abundant and production costs are low, pushing NRW to extremely low levels may deliver limited return.
This is why a 10% NRW system can be either best-in-class or underperforming.
The number alone does not tell the story.
The industry knows about NRW. It still struggles to act.
If NRW is so well understood, why does it persist at scale?
Part of the answer is not technical. It is structural.
Writing in AWWA’s Opflow, George Kunkel observes that while utilities often emphasize customer conservation, the same urgency is not consistently applied to reducing system losses. That imbalance reflects a broader issue across the sector.
NRW management remains fragmented. Regulatory expectations vary widely, and there is no consistent standard for how utilities assess or act on water loss. As a result, progress tends to depend more on local priorities than industry-wide practice.
At the same time, most water systems in the U.S. are small. These utilities often operate with limited staff and constrained budgets, making it difficult to implement advanced leak detection programs or deploy new technologies at scale.
The tools exist. Advanced metering, acoustic monitoring, district metering, and pressure management have all proven effective.
But adoption remains uneven.
In many cases, meaningful progress has not been driven by optimization, but by crisis. Prolonged drought in states like California and Georgia has forced utilities and regulators to take NRW more seriously, particularly when water loss directly threatens supply.
Without that pressure, NRW often remains a secondary priority.
That dynamic is beginning to shift.
As climate variability increases and water supply becomes less predictable, minimizing NRW is becoming less about efficiency and more about resilience.
Where utilities can start
For utilities looking to improve NRW, the starting point is not a specific percentage target. It is understanding.
A structured water audit provides visibility into where losses are occurring. From there, improving meter accuracy often delivers the fastest financial return, particularly in systems with aging infrastructure. Targeted leak detection can then focus on areas where risk is highest, rather than spreading resources too thin.
Over time, tracking NRW as a trend, rather than a snapshot, allows utilities to see how system performance is evolving and where intervention is most effective.
The bottom line
Non-revenue water is often framed as a problem to solve. At this scale, it is better understood as a signal to interpret.
A signal of infrastructure condition. Of data quality. Of operational discipline. And of how prepared a utility is to manage long-term system performance.
The utilities that benefit most are the ones that understand what their NRW actually represents. Because ultimately, NRW is about how well a utility understands the system it operates.
And in a sector facing growing infrastructure pressure, that understanding is becoming one of the most valuable assets a utility can have.
