Which Performance Indicator Whats Wrong with Percentages

Because water utilities are of different sizes, with different characteristics, comparisons of performance in water loss management need to be made in terms other than volume per year. Traditionally, several different performance indicators are used by North American utilities to compare water losses—percent of system input volume or the metered water ratio, and "per mile of mains per day" appear to be the most common. But are these reliable indicators for comparing performance?

Why do some countries use "per property per day," or "per service connection per day," or "per kilometer of systems (mains + services length) per day?" The IWA Task Force on Water Losses, with nominated representation from the AWWA, has been considering best practice internationally, and their conclusions8 strongly suggest that there are more reliable and meaningful performance indicators than "percent of system input" and "per mile of mains."

In emphasizing the importance of the correct choice of measuring units, another example from history is useful. Two thousand years ago, in the first century A.D., Julius Frontinius Sextus, then water commissioner for Rome, was spending the whole of his professional career trying (and failing) to achieve a meaningful balance between the quantities of water entering and leaving the aqueducts, which served the city. Failure was not due to lack of diligence on his part—he was simply using the wrong measures. The accepted Roman method was to compare only areas of flow; because they did not take velocity of flow into account also, their calculations could never be reliable for management purposes.

Expressing losses as a percentage is not the best way to compare loss-management performance, as systems with lower demands or successful customer side conservation programs will never be able to compete with those with larger demands. Instead the volume of loss per service connection per day should be used.

Because per capita consumption in North America is so high compared to most other countries, the common practice of expressing water losses as a percent of system input volume tends to produce lower figures than would be the case in the other countries. This gives a false impression of true performance when comparisons of performance are made with other countries with lower per capita consumption.

The same problem occurs when comparisons are made between North American utilities with a high consumption base and North American utilities with a low consumption base. Data of 1996 showed that 51 water supply systems in California had density of connections varying from 24 to 155 per mile, with an average of 75 per mile. The average metered consumption per connection varied from 136 to 2200 gal/service conn/d, with an average of some 600 gal/service conn/d. Suppose that each of these water utilities was achieving real losses of 60 gal/service conn/d, which is around three times the unavoidable annual real losses (21 gal/service conn/d) for a system with 75 conn/mi, pressure of 70 psi and customer meters 50 ft from the curb stop. Table 7.3 and Fig. 7.4 show that the percent real losses for various systems in California would vary from less than 3% to almost 30%, a tenfold range, depending upon their average consumption per connection, even if all of them had exactly the same actual leakage management performance of 60 gal/service conn/d.

Based on the average consumption of 600 gal/service conn/d, a target of 10% real losses or less might seem reasonable. However, from the above figures it can be shown that

• For utilities with low consumption per service connection it would be a quite unrealistic target, being almost equal to the unavoidable annual real losses.

• For utilities with high consumption it would represent real losses of around 11 times the unavoidable annual real losses.

If Table 7.3 and Fig. 7.4 were not in themselves sufficient to demonstrate the problem of using percentages for comparisons of performance in managing real losses, there would be further serious disadvantages.

• Where a utility exports water, the percentage real losses will be lower if the exported volumes are included in the calculation, and higher if they are excluded.

• The problem of expressing water losses in percentage terms is compounded when demand management measures (customer side conservation) to reduce per capita


Consumption in gals/service line/d

Real Losses in gal/service line/d

System Input in gal/ service line/d

Real Losses as % of System Input Volume

























Table 7.3 How Percent Real Losses Vary with Consumption, for Real Losses of 60 gal/service conn/d

Table 7.3 How Percent Real Losses Vary with Consumption, for Real Losses of 60 gal/service conn/d a.

3 10

200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 Average consumption in U.S. gals/service line/d

Figure 7.4 How percent real losses vary with consumption, for real losses of 60 gals/service line/d.

3 10

200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 Average consumption in U.S. gals/service line/d

Figure 7.4 How percent real losses vary with consumption, for real losses of 60 gals/service line/d.

consumption (pcc) are applied—as the pcc goes down, the percent water losses goes up. Not a great incentive to demand management in its widest sense, simply because of the choice of an inappropriate performance indicator!

Technical committees worldwide (Germany, United Kingdom, South Africa) have recognized these paradoxes of using percentages, but perhaps most significantly the England and Wales Economic Regulator [Office of Water Services(OFWAT)] also recognized it and stopped publishing water losses statistics in percentage terms in 1998. Water system managers who unquestioningly accept percentages as a valid measure of technical performance in management of water losses should consider if they are falling into the same trap as Julius Frontinius Sextus, 2000 years ago—using a simple, but inappropriate, measure to draw inappropriate conclusions.

Dealing With Bronchitis

Dealing With Bronchitis

If you're wanting to know more about dealing with bronchitis... Then this may be the most important letter you'll ever read! You are About To Read The Most Important Information That Is Available To You Today, You Will Achieve A Better Understanding About Bronchitis! It doesn't matter if you've never had bronchitis before or never known anyone who has, This guide will tell you everything you need to know, without spending too much brainpower!

Get My Free Ebook

Post a comment