News | May 1, 1998

The Eight Types of Sewer Hydraulics: Part 1. Basis for Analysis

by Patrick L. Stevens

This installment, the first of three, introduces the basis for, and the value of ,using scattergraphs for identifying sewer problems.

Scattergraphs of depth-velocity (V) and depth-flow (Q) data from open-channel flowmeters can tell operators and rehabilitation engineers a lot about a sewer's hydraulic performance. Scattergraphs have not been widely used in the U.S.--either as a diagnostic or engineering tool, possibly because until recently, sensors had higher drift and lower resolution than needed for useful graphs. (Low resolution and drifting sensors will produce wide and unrepeatable patterns.)

Every sewer has a hydraulic performance signature. However, most sewers fall into one of eight hydraulic categories.

    1. Normal open channel flow

    2. Silt or obstacles
    3. Bottlenecks
    4. SSO downstream
    5. SSO upstream
    6. Temporary blockage
    7. CSO or dams
    8. Variable downstream conditions - siphons, pump stations, etc.

Although many more types of characteristic patterns are to be found in sanitary and combined sewer and sewer structures, these eight types are the most common.

Nearly all open-channel flow-measuring devices perform by sensing both the flow's depth and velocity. These data usually are collected once every 1 to 15 minutes for use in calculating the flow rate. These rates traditionally are displayed as a time-series hydrograph. A scattergraph provides not only an excellent quality-control flow-monitor-data tool but insight into the hydraulic performance of the sewer.

Methodology

During the last century, several hydraulic engineers developed equations (pipe curves) to describe the relationship between the depth of an open-channel gravity flow to the velocity of that flow. For a given depth of flow there is a unique and predictable velocity (and flow rate). Figure 1 is a common pipe curve described by Robert Manning in 1890. The plot of paired depth and velocity readings over several days should form a pattern similar to this pipe curve. Patterns that deviate indicate that either the pipe curve is invalid, the meter is malfunctioning, or the pipe's hydraulics are not normal.

Figure 1: Robert Manning's 1890 Pipe Curve.



Since most open-channel flowmeters simultaneously measure both depth and velocity, every pair of readings should fall on a theoretical pipe curve ( Figure 2). The Manning equation requires the pipe curve to begin at the origin on this graph. More important, the collection of depth and velocity data points should trace out the pipe curve for the sewer it is measuring. If the sensors are not subject to drift and have sufficient resolution, the data points collected over several months for all depths should trace the curve for the entire depth range.

Figure 2: Paired readings should follow curve.



Experience, however, shows that finding sewers that produce scattergraphs conforming this closely to hydraulic theory is unusual. Although the pattern traced by the data points look quite similar to the theoretical Manning pipe curve, they do not line up exactly--that is they are scattered about the theoretical curve..

In this article, all Manning pipe curves pass through the field calibrations. A Manning curve drawn superimposed in this manner seldom lines up closely with the data points.

Figure 3 shows a scattergraph of data and is an example of normal open channel flow . The graphic represents a pipe with flow that was always less than half pipe. (It is the first of eight types of hydraulic graphs in this article.). In addition to the flowmeter data points (red), the graph includes the manual depth and velocity calibrations (five points at lower part of graph), and the theoretical pipe curve (blue). (The sewer generating the data for Figure 3 contains around 2 cm. (.75 in.) of silt , causing the origin of the Manning pipe curve to be offset to the right.) The combination of the three data sets offers the analyst a powerful tool, determining if the flowmeter is "tuned in" properly.

Figure 3: Example of normal flow - Type 1.



The vast majority of sewers produce scattergraphs similar to one of the seven other types shown in this paper.

The case studies and examples that follow appear in the format of Figure 3,which is the standard display of the ADS software Quadrascan 5.05. In addition to depth-velocity displays, in some cases the depth and Q (flow) data are also presented in scattergraph format. First time viewers of scattergraph data should be sure to recognize that these are not time-series displays. In this respect, moving from left to right on a scattergraph is not the same as moving left to right through time as displayed by a hydrograph. Experience indicates that many first time viewers intuitively believe that a data point at the right of a scattergraph is the last data point generated.

About the Author: Patrick L. Stevens is with ADS Environmental Services, 6630 E. 75th Street, Suite 204, Indianapolis, IN 46250.