Flow Monitoring Data Review

Maximizing data uptime by applying a scattergraph review process.

Flow Monitoring Data Review

FIGURE 6: A hydraulic jump occurs when flow transitions from supercritical to subcritical flow. Supercritical conditions are observed on the upstream side of the jump, and subcritical conditions are observed on the downstream side of the jump.

We started this series of articles to provide guidance based upon over 40 years of flow monitoring experience. It is also an aid to rainfall-dependent inflow and infiltration study practitioners so that they may avoid project stumbling blocks that have been observed over time.

In the last article, we discussed the importance of focusing basin sizes to provide maximum granularity in identifying problem areas in the collections system. By reducing basin footages, a targeted approach can be applied that has benefits in cost and remediation effectiveness. This article is focused on data quality and maximizing data uptime by applying a scattergraph review process throughout the project progression.  

It is a common approach to evaluate flow data by reviewing its hydrograph. A regular diurnal pattern suggests the meter is working OK. But it must be recognized that no flowmeter on the face of the earth actually measures flow; they all measure something else and calculate or derive a rate of flow. For the open-channel flowmeters used in I&I studies, water depth and water velocity are measured and flow is calculated by knowing the cross-sectional area of the flow. So the proper way to assess the quality of the meter data is to evaluate its measurements — depth and velocity — and the scattergraph is the tool for this evaluation.

There is a concept in the business community that key performance indicators can be used to guide the business. KPIs help organizations achieve organizational goals through the definition and measurement of progress. The key indicators must be measurable and must be key to the success of the task. KPIs can apply to any endeavor, and for driving a car, the speedometer and gas gauge are examples. For measuring before-and-after RDII reduction, the two KPIs are the depth-velocity scattergraph of open-channel flowmeter data and the Q vs. i (RDII vs. rainfall) plot. 

Analyzing a scattergraph is a mandatory first step in evaluating a flow monitor’s performance whenever flow monitors are utilized, and the user can apply two tests in sequence to evaluate data in a scattergraph: looking for a repeatable pattern and comparing data to manual readings. When this method is applied, you can count on reliable data forming the foundation of a successful I&I program.

Repeatable patterns

The Manning curve is the classic curve used in sewer hydraulics to define the depth-velocity relationship. If the flow monitor data lines up with the pipe curve, the user knows that the sewer is experiencing uniform flow conditions; as depth increases, the velocity increases. If the flow monitor data does not line up with a pipe curve, one of only two things is occurring:

A) The sewer is not experiencing uniform flow conditions.

B) The flow monitor is failing to make valid measurements. 

It is critical that the observer attempts to make this distinction before moving on to any other accuracy issues.

It is important to recognize that under nonuniform flow conditions the data may still be valid even though the data and the Manning pipe curve do not coincide. If the patterns indicate that valid hydraulic conditions exist, then the user should move on to manual confirmation of monitor readings.

The precision of a flowmeter can be assessed by the repeatability of the data pattern. Precision is determined by how tightly the data conform to a pipe curve. This concept is illustrated in Figure 1. Precise meters will produce a data pattern that adheres to the pipe curve much better than a low precision meter. 

Manual confirmation

If the flow monitor’s depth and velocity readings are coincident or very close to the manual readings, the user knows that the measurable components are correct or bias-free. It is important that only those data points immediately before and after the manual confirmation be used in this comparison. Apparent bias can occur in sewers with rapidly changing hydraulics such as an upstream pump station. Common sources of depth bias include pressure sensor drift, unstable hydraulics, large waves, noisy sites and fouled sensors.

To assess the quality of flow monitor data, three sets of data are plotted together on a scattergraph: (1) flow depth and velocity readings from the flow monitor, (2) manual depth and velocity confirmations collected with a ruler and a portable velocity meter, and (3) a pipe curve described by the Manning Equation. Figure 2 shows how these data should appear in a site under uniform flow conditions with a properly functioning flow monitor. All three data sets should be aligned with each other. 

It can be easily seen if the data points align well with the Manning pipe curve, but it is not clear how closely the manual measurements coincide with the flow monitor data. This is especially true if there are many data points. It is often instructive to look more closely at individual data points and the corresponding manual confirmations. Taking a manual confirmation at exactly the same moment as a monitor reading could distort sewer hydraulics and the monitor reading.

Using this technique

If the flow monitor data lines up with the pipe curve and can be confirmed manually, only then can the user consider the flow rates reported by the flow monitor. Several things, including the use of an inappropriate equation or an incorrect pipe diameter can affect the accuracy of the subsequent flow calculations. A very common source of error is using the nominal pipe diameter shown on the drawings instead of using a field measurement. This frequently occurs in small-diameter sewers where the actual diameter often does not equal the nominal diameter. 

Iso-Q lines are another useful tool when evaluating flow during an I&I project. Iso-Q lines are lines indicating depth-velocity points that result in a constant flow rate. They are made up of all depth and velocity combinations that generate a given flow rate. They are interpreted similarly to contour lines on a map. 

When dealing with I&I, capacity is a high priority. Overflows occur when capacity is maximized during a wet-weather event. One way to prevent this from happening is to determine if all lines are operating at maximum capacity or if there is a blockage in the line that may be inhibiting the pipe from conveying its maximum potential. Iso-Q lines are an effective way to visualize the capacity at which the monitored line is operating.

Flow monitor data that do not lie on a pipe curve indicate either that the hydraulics are different or that the flow monitor is not working correctly. The scattergraph shown in Figure 3 displays data from a flow monitor with a drifting pressure depth sensor.  

Note that the reported flow depth drifts over a wide range without a corresponding change in flow velocity. In this case, a series of pipe curves are observed at multiple depths and deviate significantly above and below the manual confirmations. The data from this flow monitor is invalid and should be disregarded.

The scattergraph shown in Figure 4 displays data from a flow monitor with a drifting velocity sensor. Note that the reported flow velocity drifts over a wide range without a corresponding change in depth.

In this case, an electromagnetic velocity sensor was fouled by grease, and sensor performance deteriorated over time, eventually causing the flow monitor to record negative velocities. The data from a flow monitor exhibiting this behavior is invalid and should be disregarded.

 Common scattergraph patterns

The above examples indicated how scattergraphs can be used to identify equipment maintenance requirements. Scattergraphs also provide insight into site hydraulics and can describe relevant information pertaining to wet-weather performance.

Surcharge conditions are common in sewer systems, especially during wet-weather events. The flow monitoring data shown in Figure 5 indicates that this sewer operates as expected up to its rated capacity of 910 L/s (20 mgd). This value is shown using an Iso-Q line. 

Although surcharge conditions are common, it is uncommon to find a surcharged sewer that actually accommodates its rated capacity, as shown here. The minimum and maximum dry-weather flow rates are also shown using Iso-Q lines. The maximum dry-weather flow rate occurs at a depth-to-diameter (d/D) ratio of 0.77, a depth in excess of generally accepted design guidelines.

Most surcharge conditions result from downstream restrictions that reduce sewer capacity. The sewer will operate as expected up to a certain depth, but backwater conditions will be observed above this point and result in surcharge conditions at a lower capacity than expected. The impact on sewer capacity is readily identified using Iso-Q lines.

A hydraulic jump occurs when flow transitions from supercritical to subcritical flow. Supercritical conditions are observed on the upstream side of the jump, and subcritical conditions are observed on the downstream side of the jump, as shown in Figure 6.

This condition causes a wave to travel up and down the pipe. While the sensor can read well in either condition, the constant fluctuation causes two distinct flow patterns to be observed, which can create difficulty when evaluating an average flow rate. In the case of a hydraulic jump, the data takes a distinct shape and the relationship to Froude numbers can easily be observed.

As a data reviewer, this would indicate that the sensor position needs to be adjusted or an alternative site should be considered.

The signature of a sanitary sewer overflow in the scattergraph of flowmeter data depends on its position relative to a flow monitor. A SSO that occurs upstream from a flow monitor will be identified on a scattergraph by a cluster of surcharge data points at a constant flow depth and a constant velocity, as shown in Figure 7. The depth reported by the flow monitor during the SSO is controlled by the overflow elevation, and the velocity is controlled by the capacity of the downstream restriction. 

This SSO lasted for almost eight hours. However, since the SSO occurred upstream from the flow monitor, the overflow volume cannot be estimated.

The scattergraph signature of a downstream SSO can also be identified.

Both upstream and downstream SSOs are characterized by a constant flow depth during an overflow. However, the additional flow escaping the system during a downstream SSO is detected by the flow monitor as an increase in velocity during the event, as shown in Figure 8.

The maximum overflow rate is determined using the Iso-Q lines and is approximately 2.8 mgd (13.3 mgd minus 10.5 mgd). This SSO lasted for about six hours and discharged 331,000 gallons of untreated wastewater to the environment.

Data review performed while flow monitoring for an RDII study is absolutely critical to the success of the study. The scattergraph is the ideal tool for performing this analysis through its ability to visually represent flow monitoring equipment concerns, as well as offering added information on hydraulic conditions present at a monitoring location.


Comments on this site are submitted by users and are not endorsed by nor do they reflect the views or opinions of COLE Publishing, Inc. Comments are moderated before being posted.