Twelve Steps to Successful Measurement of RDII Reduction

It has often been said that the best way to learn how to do a thing is to ask experienced people what mistakes they have made and what corrective steps they took in subsequent projects.? After 40 years of experience in sewer-related projects as a regulator, an Owner and a consultant, I have seen many successful sewer rehabilitation projects as well as many others that did not fare well.??

It appears that there are essentially three fates for sewer rehabilitation projects that are intended to reduce Rainfall Dependent Infiltration Inflow (RDII) as shown in Figure 1.? From left to right the three fates are:

A) RDII is reduced and there are clear and measurable results;

B) There are subjective reasons to believe RDII was removed, but it cannot be quantified; and

C) There was no apparent reduction of RDII with symptoms remaining unchanged or actually worsening.

I have developed a list of 12 items that are the most common gateways to Category ?B? and ?C?.? This article addresses those projects falling into category ?B?.? Projects falling into this category would have been in category ?A? except for an adequate plan for demonstrating a reduction in RDII.? In many ways these ?B? projects conform to the adage that ?you can?t manage what you don?t measure?.? This article provides advice and guidelines to help you ?manage what you measure? and deliver Category ?A? projects.

Projects end up in the ?C? category for several reasons.? Shortcomings in the physical rehabilitation work contribute to the problem, including an inadequate job of locating the sources of RDII, piecemeal repair of sewers, repairing only public sewers and many other causes.? Similarly, projects may end up in the ?C? category because they failed in the initial diagnostic work needed to locate and quantify sources of RDII.? This list of 12 items is also a recipe for developing a suitable diagnostic effort.? I will discuss the first five items in detail.

Rainfall Strategy

The concept of measuring or modeling wet weather performance of sewer systems is fundamentally very simple.? It is the establishment of the relationship between rainfall (the independent variable) and the resulting RDII (dependent variable).? To determine if a sewer rehabilitation project was effective one has to compare that relationship before the work to the relationship after the work.? How hard could it possibly be?? It is hard enough that few attempt it.

Rainfall is at the top of the list because it ruins more rainfall-to-flow relationships than does flow data.? Mathematically, accuracy of rainfall is as important as accuracy of flow data, but many agencies treat it as an after thought or merely a cost item to be minimized. ?

The most important consideration is the density of the rain gauge (RG) network.? Recommendations vary from 1 RG/1 Mi2 to 1 RG/10 Mi2 so how is one to make a choice?? Try this experiment at home and you will learn the rule that ?the more intense the rainfall, the narrower the footprint?.? During the next big storm in your area, log in to your local NEXRAD radar and look at the hourly accumulation.?

In this example near our office, the green footprint of rainfall exceeding 0.5 inches is just 6 miles wide and the footprint of rainfall of 1.2 inches (yellow) is just 0.6 miles (1 km) wide.? Depending on the rain gauge density the measured rainfall could have been 0.5 or 1.2 inches for the hour.? This can make a huge difference in model calibration or RDII measurement.? In my view rain gauges should be no farther apart that 2 miles (4 mi2 per RG)


Duration of the study means that not enough data (dry days and storms) were collected to generate proper and statistically-valid rain-to-flow relationships.? The target should be twelve system-stressing storms.? A system-stressing storm is one in which flow at least doubled at most of the metering sites.? Smaller responses can easily get lost in the natural ?noise? in the data.

QA/QC Touchstones – KPIs

The term ?Key Performance Indicator? is commonly used in business enterprises as a tool to guide processes towards a goal.? The two Key Performance Indicators for RDII work are Q vs. i (RDII vs. Rainfall) plots and the depth-velocity scattergraph.? If data are organized properly and the sewer is not restricted, the Q vs. i plot should be linear.? If it?s not linear, a problem may exist in the supporting data. ?

The scattergraph is an ?Engineer?s Friend? because it reveals problems with the meter and defines the hydraulic conditions in the sewer, which is important to modelers.? Learn more and request a free scattergraph poster by following this link:

1.?? ?Flow Meter Depth Technology.
2.?? ?Size of Meter Basins.

Metering Depth Technology

I have looked at flow data from thousands of RDII metering sites and by far, the most common reason flow data are disqualified is because of a drifting pressure sensor.? Pressure sensors are prone to drift and the only rule I have been able to develop is that ?all pressure sensors drift at some time in their life, but one never knows when and by how much?.? Expensive ones drift less than cheaper ones, but all will suffer some day.
If a pressure sensor is the primary depth measurement, the user must be prepared to detect and correct drift and this is hard to do.? A sensor that drifts just ? inch per month is hard to detect, but over a two month period it can cause an error of over 100% in measured flow.

Pressure sensor technology is essentially a strain gauge while ultrasonic technology is based on a precise clock that does not drift.

Basin Size

Although Basin Size is number five on my list, it probably has the greatest effect on the overall success of RDII projects.? Small basins of around 10,000 LF or less of sewer bring these benefits:

  • Isolate measured RDII into the smallest fraction of the collection system. ?
  • Reduce the length of pipe that must be inspected
  • Reduce the magnitude of sewer repair/replacement
  • Reduce the overall project duration and cost
  • RDII reduction is easier to quantify because the reduction is a greater percentage of the average flow and is easier to measure

The reason that small basins bring such benefit is that RDII is not uniformly distributed throughout a collection system, as was assumed when these projects began in the late 1970?s.? In any set of non-uniformly-distributed data, the sample size determines how clearly one can understand the actual distribution.? In sewers, many analysts initially believed the age of the sewers would control the distribution of RDII (older pipes means more RDII), but they soon found that this is not a strong relationship. ?
The causes of sewer defects (RDII sources) are so complex that only actual measurement in small samples reveals the actual distribution and magnitude.? It has been demonstrated that RDII in collection systems can conform to the 80/20 Rule of Pareto?s Principle, if the basin size is small enough.? Application of the Rule would say the 80% of the total volume of RDII entering a collection system will enter in just 20% of the system.

Here is a graph of several recent RDII projects ADS has performed at different basin sizes.? Each data point on this graph shows the basin size of the study (X-axis) and the percentage of the system that produced 80% of the total RDII (Y-axis).? It is evident that the smaller basin-size studies have isolated 80% of total RDII volume into approximately 20% of the system.? Agencies that begin RDII reduction programs with this starting view can work faster and at less cost.?

The steps discussed in this article will produce a successful measurement of RDII severity or RDII reduction.? For more information on these twelve items look at the Webinar on the same topic recorded on 22 March 2011 at this link.

More information and technical papers are available at the ADS website.

Leave a Reply

Your email address will not be published. Required fields are marked *